The spreadsheet ad
and the death of
please believe us.

Spend enough years buying media against fast-growing consumer brands and you develop an instinct for when the ground shifts under the feed. Not a new tactic, a new grammar. I have been watching one of those shifts play out across the ad accounts I work in since roughly the back half of 2024, and it has only accelerated since. The ads that win now have almost nothing in common with the ads that won five years ago, and the brands that have internalized that are quietly pulling away from the ones that have not.

Here is the clearest symptom of it, and it is a strange one. A lot of the best-performing direct-to-consumer ads right now look more like spreadsheets than ads. Brands you already know, Liquid Death, David, AG1, Eight Sleep, are running creative that is almost boring in its simplicity. Put the products side by side. Compare the things customers actually care about. Make the decision feel obvious. I have been buying against brands like these for years, first as an operator co-founding WIN Brands Group and scaling individual brands into eight figures, then as an advisor sitting in the ad accounts of companies trying to get from six figures of revenue to nine. So when I say this is not a gimmick, I mean it is the surface symptom of something much larger that has been reshaping how paid acquisition works.

The spreadsheet ad is the canary. What it is telling us is that the entire grammar of the feed has changed, and the ads that win now are the ones that stopped trying to look like ads at all.

Here is the line I keep coming back to, and it is the spine of everything in this piece. Most ads say "please believe us." The spreadsheet ad says "look at the data yourself." One of those is a brand begging for trust. The other is a brand handing you the evidence and stepping back. In a feed where people have been trained for fifteen years to flinch at anything that pattern-matches to advertising, the second one wins, and it is not close.

I want to be upfront about why I am the one writing this, because the internet is full of people with takes on ad creative who have never spent a dollar of their own money on it. My vantage point is specific. I was an early employee at Shopify and built the Partner Program, which means I watched the entire ecosystem of brands and tools grow up around the platform. Then I went operator, co-founding WIN Brands Group, where we scaled a portfolio into nine figures of revenue and I personally lived inside ad accounts that were spending real money every single day, watching creative win and die in real time with my own capital on the line. Then I sold a software company and started advising, which put me in the accounts of dozens of other brands and Shopify app founders, all wrestling with the same question: what actually works in the feed right now, and why. So when I tell you the spreadsheet ad matters, it is not theory. It is because I have been watching this specific shift happen across a lot of accounts, from a lot of angles, for a while, with real budgets riding on which way it broke.

The stakes here are higher than they look, too, which is part of why I am giving this the full treatment. Creative is now the single biggest lever in paid acquisition. For about a decade, the edge in performance marketing came from targeting and bidding sophistication, knowing how to find and price the right audience. The platforms have automated almost all of that away. Targeting is largely the machine's job now. What is left, the one thing you still control and the one thing that still separates a brand that scales profitably from one that bleeds out, is the creative. In 2026, if your acquisition is broken, it is almost always a creative problem wearing a media-buying costume. That makes understanding these formats not a nice-to-have for the marketing team, but a core competency for anyone trying to grow a consumer brand or a software business through paid channels.

This is the longest thing I have written for this site, on purpose. I want to give you the full map. Not just the spreadsheet ad, but the whole family of formats it belongs to, the mechanics underneath them, the economics that make them win on Meta specifically right now, and the operator's playbook for actually building and testing them. I have run almost all of these. I will tell you where they work, where they quietly fail, and where I have personally lit money on fire learning the difference. By the end you should be able to look at any ad in your feed and read it the way a media buyer does, and more usefully, brief your team to make the kind that performs.

Let us start with why the old ad died, because you cannot understand the new formats without understanding the wreckage they grew out of.

Why the feed
broke the
old ad.

There is a number that should end every debate about whether advertising has fundamentally changed, and it is the click-through rate on a banner ad. The first one ever served, in 1994, earned a 44 percent click-through rate. People clicked it because they had never seen anything like it. By 2026 the average banner ad click-through rate has collapsed to roughly 0.05 percent. That is a decline of more than 99 percent from the peak. Studies put the share of consumers who suffer from banner blindness at around 86 percent, and when researchers ask people to recall the last display ad they saw and what it was for, only about 14 percent can.

Banner ads are not social ads, but the psychology is identical and it has migrated. Banner blindness is not laziness. It is a trained reflex. The human brain builds what researchers call an ad schema, a pattern recognition system assembled from years of repeated exposure to the standard shapes, positions, and visual rhythms of advertising. Rectangular borders. Predictable placement. A logo in the corner, a product shot in the center, a call to action button in a slightly-too-saturated color. Your brain learned that schema, and now it filters for it before conscious attention even engages. You scroll past the ad before you have decided to scroll past the ad.

The feed made this worse, not better. On a website you at least knew where the ad slots were. In an infinite social feed, the ad sits between your friend's vacation photos and a video your cousin posted, which means the contrast is brutal. An obviously produced ad in a feed of organic content is like a billboard in a living room. It announces itself. And the moment it announces itself, the schema fires, the thumb keeps moving, and you have paid for an impression that landed on a brain that was actively not looking.

So the first reason the old ad died is on the demand side. People got immune. But there is a second reason, and it is the one most brand teams do not understand, because it lives inside the machine.

In 2025 Meta rolled out a new core ad delivery system called Andromeda. The technical details are not the point. The behavioral consequence is. Andromeda organizes creative into a hierarchical index based on similarity. It looks at your ads, evaluates their visual composition, their audio, their copy, their narrative framing, and it clusters the ones that are conceptually alike. Ads that look and feel the same get grouped together and forced to compete for the same audiences. The more genuinely distinct your creative is, the more pockets of audience the system can match it to.

Sit with what that means. The platform is now actively penalizing sameness. If you run five variations of the same lifestyle photo with different headlines, the machine sees one concept and serves it to one slice of people. If you run a founder story, a customer testimonial, and a comparison ad, the machine sees three concepts, reaches three different audiences, and generates three independent streams of learning signal. Meta's own testing has put numbers on this. Creative diversification, in an internal Performance 5 framework test, drove a 32 percent increase in efficiency and 8 percent incremental reach compared to running similar creative.

That is the structural shift. For about a decade the winning move was to find one creative that worked and pour budget into it. Now the winning move is range. The brands that scale in 2026 are not running more of the same ad. They are running more kinds of ad. And once range becomes the requirement, you need a deep bench of formats to draw from, which is exactly why this whole catalog of strange new ad styles exploded into the feed at once. The spreadsheet ad, the notes-app ad, the founder letter, the ugly ad. They are not separate fads. They are a brand's answer to a machine that now rewards being different.

Put the two forces together and you have the whole story. Demand-side, audiences have a reflex that filters out anything ad-shaped. Supply-side, the algorithm rewards creative that does not pattern-match to other ads. Both pressures point in the same direction, toward the same conclusion. The ad that wins is the one that does not look or behave like an ad. Everything that follows is a variation on that single move.

The spreadsheet ad,
deconstructed.

Let us go back to the format that started this, because it is the cleanest expression of the whole shift and it deserves a proper teardown.

The spreadsheet ad is, at its simplest, a comparison laid out as a grid. Your product runs down one column. The alternatives run down the others. The rows are the attributes the customer actually weighs. Protein per serving. Grams of sugar. Price. Number of ingredients you cannot pronounce. Whether it is third-party tested. The cells get filled in, and the visual does the arguing. Your column is full of green checks and good numbers. Theirs is full of dashes and worse ones. You did not say you were better. You showed a table, and the table said it for you.

Why does this work so well that brands pacing toward nine figures lean on it? Three reasons, and they stack.

The first is cognitive, and it is the one most people get wrong. Humans do not evaluate products in isolation. We understand value comparatively. "More fiber" means nothing as a standalone claim, because more than what? But "more protein than the competitor, less sugar than the competitor, lower price than the competitor" suddenly feels like information rather than marketing. The comparison gives the number a frame, and the frame is what makes it meaningful. You are not asking the customer to assess your product against some abstract internal standard they do not have. You are handing them the standard and the scorecard in the same frame.

The second reason is about the locus of trust, and this is the part most brands miss. When a normal ad makes a claim, the customer has to decide whether to trust the brand. That is a high bar, because the brand obviously wants the sale and everyone knows it. A spreadsheet ad performs a quiet sleight of hand. It does not ask you to trust the brand. It asks you to trust the table. And tables feel neutral. A grid of numbers carries the visual authority of something a researcher would build, not something a marketer would. The customer reads it and feels like they are doing their own research, reaching their own conclusion, rather than being sold to. They walk away believing the decision was theirs. That feeling of self-directed discovery is worth more than any superlative you could have written, because people defend conclusions they believe they reached on their own.

The third reason is that it survives the schema. A spreadsheet does not look like an ad. It looks like a tool. It looks like the thing you would have built for yourself in a Google Sheet if you had the time to research every option in the category. So it slips past the banner-blindness reflex that kills produced creative, and it does so while delivering more decision-relevant information than a beautiful lifestyle shot ever could. It is native in format and dense in substance at the same time, which is a rare combination.

This is why the brands I mentioned all reach for it. David built an entire creative engine around comparison. One of its best-known ads lines up five familiar foods, pizza, chips, chocolate, peanut butter, and a David bar, and shows that for the same calories you get four times the protein. That is a spreadsheet ad in spirit even when it is not literally a grid. It frames the product against the things the customer already eats and lets the contrast carry the pitch. The brand is pacing toward roughly 100 million dollars in first-year sales, which almost never happens in DTC, and a repeatable comparison format that does not fatigue is a meaningful part of how you sustain spend at that velocity.

But I want to be honest about where this format fails, because the LinkedIn version of any trend skips the failure modes and the failure modes are where operators actually live.

The spreadsheet ad fails when you do not actually win the comparison. The format is ruthless because it is specific. The moment you put yourself in a grid against competitors, you are inviting a literal line-by-line evaluation, and if your product is only better on one row and worse or equal on the rest, you have just built a beautiful, high-trust ad that argues for buying someone else. I have watched brands get excited about the format, build the table, and quietly discover that their honest grid is not flattering. The fix is not a better table. It is choosing the rows where you genuinely win and the comparison set where you genuinely lead, which is a positioning decision, not a design one.

It also fails when the category does not turn on comparable attributes. Spreadsheet ads thrive in supplements, food, sleep, and anything where the buying decision is a spec sheet in disguise, which tends to track with how the category's unit economics work by vertical. They struggle in categories where the value is emotional, aesthetic, or experiential, because you cannot put "makes me feel like the person I want to be" in a cell with a checkmark. A candle brand running a spreadsheet ad looks like it is trying too hard. Know which kind of category you are in before you reach for the grid.

And it fails, slowly, when everyone does it. Part of why the spreadsheet ad works today is that it still reads as novel and neutral. As the format saturates, the schema will adapt, the way it adapts to everything. The grid will eventually become just another ad shape the brain learns to flinch at. That does not make it a bad bet now. It makes it a bet with a clock on it, which is true of every creative trend and is the single most important thing to internalize about all of this.

Us versus them,
and us versus
nothing.

The spreadsheet ad is the most literal member of a larger family, and the family is worth understanding as a whole, because the underlying move generalizes far beyond a grid.

Every comparison ad does the same fundamental thing. It changes the question. A standard ad asks the viewer to evaluate a product in isolation: here is a thing, decide if you want it. A comparison ad reframes the entire decision: here is what you are doing now, and here is what is different. That is a dramatically easier cognitive task, and crucially, it meets the viewer exactly where they already are. They are not a blank slate. They are aware of the problem, weighing options, and not yet decided. The comparison ad speaks to that state directly instead of pretending the customer just arrived in the category.

There are two flavors, and the second one is more powerful than most brands realize.

The first is us versus them. You name or imply a competitor and show why you win. This is the classic comparison, and it works best when you have a genuine, demonstrable edge and the category has a clear incumbent for people to measure against. The challenger brands love this because the incumbent has done the expensive work of making the customer aware of the category, and the challenger gets to position against that awareness for free. When IM8 went after AG1, it was running this playbook: pick the leader, build the rivalry, ride the leader's recognition. The risk is obvious. If you pick a fight you cannot win on the merits, you have spent your budget advertising the strength of your competitor.

The second flavor is us versus nothing, and this is the one I push clients toward more often, because in most categories the real competitor is not a rival brand. It is inertia. The customer is not actively choosing someone else. They are just not doing anything. They are still using the drugstore version, the thing they have always used, the duct-taped status quo. A comparison ad aimed at the status quo does not attack a competitor. It makes the customer's current situation feel costly, inconvenient, or quietly inferior, and it gives them a reason to switch that they had not actually articulated to themselves. The best comparison ads do not beat a rival. They reframe doing nothing as the expensive option.

I learned the power of this the hard way. At WIN we had brands in categories where we kept building comparison ads against direct competitors, and they were fine, not great. It clicked when we stopped comparing ourselves to other brands and started comparing the product to the life the customer was currently tolerating without it. The before-and-after of the problem, not the head-to-head of the logos. Conversion moved, because we stopped fighting for share inside a small pool of category-aware shoppers and started expanding the pool by showing people the cost of their own inaction.

The comparison family also includes a quieter cousin worth naming: the implicit comparison, where you never show the other column at all but you build the entire ad around a number that only means something relative to a norm. "Twice the protein." "Half the sugar." "Thirty percent cheaper." The comparison is happening in the customer's head against a baseline they carry around, and you are simply supplying the multiplier. This is the most flexible version, because it works in formats where a literal grid would be clumsy, like a five-second video hook or a single static image. The discipline is the same as the spreadsheet ad: the number has to be true, it has to be checkable, and it has to land on a row the customer actually cares about.

The thread running through all of it is the same one from the spreadsheet ad. You are not asserting superiority. You are constructing a frame in which the customer concludes it. That is the move. Everything else is format.

Ads wearing
the feed's
clothes.

Now we get to the formats that take the "do not look like an ad" principle to its logical extreme, by literally dressing the ad up as a piece of the platform's own interface.

You have seen these even if you have not consciously clocked them. The ad that looks like a screenshot of someone's iPhone Notes app, a raw paragraph of text on that familiar cream background. The ad that looks like a text message conversation, two people in a blue-and-gray bubble exchange where one of them is raving about a product. The ad built to look like a screenshotted tweet, a Reddit thread, or a five-star review pasted as an image. The whole category is sometimes called platform-native creative, and the design brief is exactly what it sounds like: mimic the native interface so precisely that the eye reads it as organic before the brain catches up.

The mechanism is the schema again, but exploited from the inside. Banner blindness fires on things that look like ads. These formats look like the content the user came to consume. A Notes-app screenshot does not pattern-match to advertising. It pattern-matches to the dozens of times the viewer has seen a friend or a creator share a thought as a screenshot, because that is a genuinely common organic behavior. So the thumb does not flinch. The viewer is already reading before the ad-detection reflex would have engaged, and by then you have their attention.

Let me break down the main ones, because the craft details matter and most brands execute them badly.

The Notes-app ad is best for a certain kind of message: a personal, slightly raw, confessional statement. A founder's unpolished thought. A "I almost did not launch this" admission. A list of reasons the brand exists. The format signals intimacy and authenticity precisely because the Notes app is where people draft the things they have not polished yet. It is the visual language of "this is the unedited version." The mistake brands make is putting marketing copy into the Notes format. The instant the text reads like it was written by a brand, the disguise drops and the format actually works against you, because now it feels like a brand trying to fake authenticity, which is worse than an honest ad. The copy has to actually sound like a person typed it on their phone.

The text-message ad is best for social proof and reaction. The form is a conversation, so it naturally carries a recommendation: one person tells another about the product, the other reacts. It works because eavesdropping on a recommendation between two people feels more credible than being told directly, the same reason word of mouth has always outperformed advertising. The craft detail here is restraint. Real text exchanges are short, full of lowercase, occasionally misspelled. A text-message ad with perfect grammar and a clean call to action reads as fake immediately. The believability is the entire asset.

The fake-screenshot ad, the tweet or the review or the Reddit comment rendered as an image, is best for borrowing the credibility of the platform it imitates. A glowing review screenshotted from a third-party site carries more weight than the same words in your own ad copy, because the customer assumes you cannot fully control a third party. This is also where you have to be most careful, and I will come back to the ethics of it later, because fabricating a review or a social post that never existed is not a creative choice, it is a fraud, and it will eventually cost you more than it earns.

The reason all of these are surging at once is the same algorithmic pressure from earlier. They are maximally distinct from produced brand creative, so the machine loves them, and they are maximally native to the feed, so the audience does not filter them. They sit in the exact sweet spot the current environment rewards. They are also cheap and fast to produce, which matters more than almost anything else in the new economics of paid social, and I will get to why in a few sections.

But there is a tax on these formats, and it is rising. As they saturate, audiences are learning the new schema. People are starting to recognize the Notes-app ad as an ad. The disguise has a half-life, and every brand that runs a lazy version of it shortens that half-life for everyone. The brands that will keep winning with native-mimicry formats are the ones whose underlying message is actually true to the form: a real founder thought in the Notes app, a real customer reaction in the text thread. The disguise buys you the first second of attention. Only the substance keeps it.

When a person
outperforms
a brand.

If the native-mimicry family is about disguising the ad as content, the founder family is about replacing the brand with a human, and it is one of the most durable performers in the entire catalog.

The format is simple to describe. The founder, or a senior person at the company, talks. Sometimes to camera, sometimes as a written letter, sometimes as a Notes-app screenshot that bleeds into the previous category. They tell the story of why the product exists. The problem they hit. The gap they saw. The thing they could not buy, so they built it. There is no script polish, no agency gloss. It is a person making a personal case.

The data on this is genuinely striking. One analysis of more than 1,200 ad campaigns found founder-led creative outperforming traditional brand content by an average of 2.7 times across click-through, engagement, and conversion. That is not a rounding-error improvement. That is a different tier of performance, and it comes from two distinct mechanisms working at once.

The first is trust transfer at the source. Audiences are more receptive to a person than to a logo, full stop. A brand making a claim is a marketing message. A founder making the same claim is a personal account, and we are wired to extend more credit to people than to institutions. The founder format also answers the silent question every cold customer has and never asks out loud: why should I trust this company at all? A founder explaining why they built the thing answers that question in a way no amount of brand polish can, because provenance is the answer to trust, and only a person can provide provenance.

The second mechanism is pattern interruption. A founder talking to camera with imperfect lighting and real energy looks nothing like the sea of stock imagery and influencer aesthetics surrounding it in the feed. It stops the scroll precisely because it is not optimized to stop the scroll. The roughness is the signal. It reads as a human who had something to say, not a department that had a quota to fill.

I have a strong personal bias here, and I will own it. I have spent my whole career around founders, including being one, and I think the founder format is the most under-exploited high-performance creative available to early and mid-stage brands. The reason brands under-use it is not that it does not work. It is that founders are uncomfortable on camera, or they think it does not scale, or someone on the team decided the brand should feel bigger than one person. All three of those are mistakes at the stage where most brands are fighting for their first real traction. When you are small, the founder is your single most differentiated asset, because no competitor can copy your specific story. Spend that asset.

The format does have a ceiling, and it is worth naming. Founder-led creative is strongest top of funnel, when a cold audience is meeting the brand and deciding whether to care. It is weaker at the bottom, where a warm customer who already trusts you just needs a reason to act today, which an offer does better. And it has a real-world constraint: it depends on a founder who is at least passably watchable, and not every great operator is. If your founder genuinely cannot carry it, do not force it. A wooden founder ad is worse than no founder ad, because discomfort reads on camera and it transfers to the brand. But if your founder can string two honest sentences together about why the company exists, you are leaving money on the table by not running it.

There is also a B2B and SaaS version of this that I will return to later, because the founder letter works just as well for a software company selling to other businesses as it does for a supplement brand selling to consumers. Trust transfer does not care what you sell.

When bad design
is good
strategy.

Now for the format that makes brand designers physically uncomfortable, and that is exactly why it works.

The ugly ad is what it sounds like. Raw, text-heavy, intentionally rough. Mismatched fonts. A screenshot with the status bar still showing. A product photo that looks like it was taken on a kitchen counter under bad light, because it was. The broader category is the anti-ad: creative that deliberately subverts the conventions of advertising, that is self-aware or weird or so plainly unproduced that it earns attention precisely by refusing to look like the thing it is.

The strategic logic is the cleanest in the whole catalog, and it follows directly from everything we have covered. The schema fires on polish. Years of glossy advertising taught the brain that high production value equals "this is an ad," and the reflex follows. So a deliberately unpolished ad short-circuits the reflex. The brain does not categorize it as advertising fast enough to scroll past it, and you get the attention you paid for. Ugliness is camouflage.

The numbers back the instinct. An analysis of over 100 million dollars in top-spending DTC ads found that 42 percent were lo-fi, shot on an iPhone with minimal editing. Not 4 percent. Forty-two. Among the highest-spending advertisers, the kind pouring real money into channels where every dollar is measured, nearly half the winning creative looked like it cost nothing to make. That should reframe how you think about your creative budget, because it means production value and performance are not the same axis, and at the top of the market they are often inversely correlated.

There is a sibling to the ugly ad that deserves its own mention: humor. The same dataset found that 25 percent of the highest-spending advertisers used humor in their creative, compared to just 14 percent across the market as a whole. The biggest spenders are nearly twice as likely to be funny. Liquid Death built an entire nine-figure water brand on this, turning canned water into a comedy and lifestyle property by refusing to advertise water the way water is advertised. Funny and weird works at scale. Most brands are simply too cautious to try it, and that caution is the opportunity, because the formats your competitors are too nervous to run are the ones with the least competition.

I want to be precise about why premium brands resist this, because it is the most common objection I hear and it is usually wrong. The instinct is that an ugly or irreverent ad will damage a carefully built brand aesthetic. That fear is real and occasionally valid, but it is overweighted almost everywhere. Distinctiveness in a feed full of sameness is an advantage, not a risk. The household-essentials brand Who Gives A Crap is the clearest proof I know of: working across the UK, US, and Australia, the brand tested more than 375 concepts spanning everything from polished assets to humor-driven experimental styles, hit a 21 percent win rate, grew the UK market 70 percent, and increased global Meta spend 72 percent. The concepts that leaned hardest into the brand's humor and irreverence, the ones that looked nothing like conventional toilet-paper advertising, consistently outperformed the safe executions. The brand did not scale despite its personality. It scaled because of it.

The trap with ugly ads is thinking ugly is the strategy. It is not. The strategy is native and distinct. Ugly is just one cheap, reliable way to get there, and an ugly ad with nothing to say is still an ad with nothing to say, now also ugly. The roughness has to wrap around a real hook, a real claim, a real reason to stop. Get the substance right and the lack of polish is a feature. Get it wrong and you have just made a bad ad that also looks bad, which is two problems instead of one.

The format that
quietly became
the baseline.

No honest catalog of trending ad styles can skip the format that has quietly become the baseline for high-performing paid social, even though it is the least novel thing on this list: user-generated content and creator demos.

UGC-style video has become the default creative format for high-performing Facebook and Instagram advertisers in 2026, dominating across nearly every vertical when you filter the spy tools for the ads with the longest active lifespans, and active lifespan is the truest proxy for performance there is, because brands do not keep paying to run ads that lose. The format is a real or apparently-real person, in a real setting, talking about or demonstrating a product. Phone-shot, vertical, lightly edited or not edited at all.

Why has it become the default? Two reasons, and you have seen both before in this piece because the whole catalog rhymes.

The first is native feel. Creator and UGC content looks like the rest of the feed. It does not pattern-match to advertising, so it bypasses the scroll reflex. A viewer who would have flicked past a produced brand ad will stop for a person talking directly to camera in their kitchen, because that person looks like everyone else they follow. This is the same mechanism as the ugly ad and the native-mimicry formats, just delivered through a human instead of a screenshot.

The second is trust transfer, the same engine as the founder format, pointed at a different person. When the talent is a creator the viewer already follows, something specific happens: the viewer has already decided that this person has good taste, good judgment, or relevant experience. When that person recommends a product, some of that accumulated trust moves to the brand. A polished brand ad cannot manufacture that transfer. It has to be borrowed from someone who already earned it with their audience, and that is precisely what creator partnerships buy. You are not renting reach. You are renting credibility.

There is a third, more operational reason UGC became the default, and it is the one that actually drives the spend: it is the most efficient format to produce at volume. And volume, as I will argue in the next two sections, is the entire game now. You can commission dozens of UGC variations for the cost of one studio shoot, which means you can test more angles, more hooks, and more personas, which means you find winners faster. The format wins partly on performance and partly on the brutal math of how many at-bats it gives you.

The discipline with UGC is resisting the urge to over-produce it. The moment a UGC ad starts to look like a commercial, with a brand-supplied script the creator clearly did not write, color grading, and a clean call to action card at the end, it loses the native quality that made it work and becomes an expensive ad wearing a cheap costume. The best UGC keeps the rough edges on purpose. The worst is a TV spot in disguise, and audiences can tell the difference faster than any brand wants to believe.

The one move
underneath
all of it.

Step back from the individual formats and look at the shape they make together, because the shape is the actual insight and it is more useful than any single tactic.

The spreadsheet ad hands you a table and lets you conclude. The comparison ad hands you a frame and lets you weigh. The founder format hands you a person and lets you judge their sincerity. The UGC ad hands you a peer and lets you borrow their experience. The native-mimicry formats hand you something that looks like organic content and let you read it before you defend against it. The ugly ad hands you something unpolished and lets you decide it must be real because no one would fake something that rough.

Every single one of these is doing the same thing. It is moving the act of persuasion out of the brand's mouth and into the customer's head. The old ad asserted: we are the best, buy us, please believe us. The new ad assembles a situation in which the customer concludes it themselves: look at the data, weigh the options, hear the person, trust the peer, reach your own verdict. The brand stops being the one making the claim and becomes the one arranging the evidence.

This is the thesis of the entire piece, and once you see it you cannot unsee it. The winning ad styles of this era are not united by a look. They are united by a transfer of authorship. The customer is invited to do the persuading, on the brand's behalf, using materials the brand carefully provided. And it works because a conclusion you reach yourself is infinitely more durable than a conclusion you were handed. We defend our own conclusions. We resist other people's. Every one of these formats is engineered around that single fact about how humans hold beliefs.

There is a reason this shift is happening now specifically, and it is not just banner blindness and Andromeda, though those are the proximate causes. It is that the customer has more access to information than any customer in history. They can check your claim in three seconds, increasingly by asking an AI engine rather than a search box, which is becoming its own discipline I cover in answer engine optimization for commerce brands. They can read the reviews you did not write. They can find the Reddit thread where people argue about your category. In that environment, an ad that asserts superiority is competing against the customer's own ability to verify, and it loses, because the customer trusts their own research over your adjectives. So the smart brands stopped competing with the customer's research and started feeding it. They make the ad the first step of the customer's own investigation rather than a substitute for it. Show the data, because they are going to go look at the data anyway, and you would rather be the one who framed it.

This is also why I am confident the underlying pattern outlasts the specific formats. The spreadsheet ad will saturate. The Notes-app ad will get recognized. The ugly ad will become its own cliche. Formats always die. But the principle underneath them, that you win by transferring the conclusion to the customer rather than asserting it, is not a format. It is a permanent adjustment to a world where customers can check your work. The formats are just the current technology for executing the principle. When they wear out, the brands that understood the principle will simply invent the next technology for it, and the brands that only copied the format will be left running a Notes-app ad into an audience that learned to flinch at Notes-app ads.

Understand the principle. Rent the formats.

Four brands,
four costumes,
one move.

Theory lands harder with examples, so let me walk through the four brands I mentioned at the top, because each one runs the evidence-over-persuasion principle through a different format, and seeing the principle wear four different costumes is the fastest way to internalize it.

Start with David, the protein company, because it is the purest comparison brand on the list and it is doing something genuinely instructive. David did not invent a new protein bar and then advertise it the normal way. It built its entire creative identity around a single comparison move: same calories, more protein. Its best-known ad lines up familiar foods the customer already eats, pizza, chips, chocolate, peanut butter, and puts the David bar at the end of the row, holding calories constant and letting the protein number do the talking. That is a spreadsheet ad even when it is not a literal grid, because it frames the product against the customer's existing reality and lets the contrast carry the argument. The brand is pacing toward roughly 100 million dollars in first-year sales, a number that almost never happens in DTC, and a comparison format that does not fatigue is part of how you sustain that velocity, because you can refill the same frame with new foods, new claims, and new rows indefinitely without the audience tiring of it. David also leaned on credible scientific backers early, which is its own form of evidence-over-persuasion: it is not the brand asserting the protein matters, it is the brand pointing at people the customer already trusts on the subject.

AG1 runs a different version, and it is more about authority than comparison. AG1's creative plays heavily on routine and scientific credibility: clean visuals, expert voiceovers, simple demonstrations of the product fitting into a high-performer's day. The evidence it transfers is not a comparison grid, it is the implied endorsement of the experts and athletes around the product, plus the visual argument that serious people have made this part of their routine. AG1 is also a cautionary tale about the cost of this strategy, because when you build a category-defining brand on comparison-friendly attributes, you invite challengers to run spreadsheet ads against you. IM8 did exactly that, picking AG1 as the foil and building a rivalry to borrow AG1's recognition. That is the comparison family pointed back at the category leader, and it is the structural risk every successful brand takes on: your own success makes you the column everyone else wants to compare against.

Liquid Death is the anti-ad and humor brand on the list, and it is the proof that the principle does not require a grid at all. Liquid Death sells water, the single most undifferentiated product imaginable, and it built a nine-figure brand by refusing to advertise water the way water is advertised. No mountain streams, no hydration claims, no wellness aesthetic. Instead, comedy, heavy-metal branding, and deliberately absurd creative that looks nothing like beverage advertising. The evidence it transfers is cultural rather than factual: the ad does not argue the water is better, it demonstrates that the brand is funnier and more self-aware than everything around it, and it lets the customer conclude that a brand this entertaining is one they want to be associated with. This is the humor stat made flesh. Remember that 25 percent of the highest-spending advertisers use humor versus 14 percent of the market. Liquid Death is what the high end of that distribution looks like, and its whole existence is an argument that distinctiveness beats category convention.

Eight Sleep is the high-consideration, spec-comparison brand, and it shows how the spreadsheet format works for a premium, expensive product. When you are asking someone to spend thousands on a smart mattress cover, the customer is going to research exhaustively, so the smart move is to feed that research rather than fight it. Eight Sleep leans on comparison and data: temperature regulation, sleep tracking, the measurable difference in how you sleep. It hands the high-intent buyer the specs and the proof points they were going to go hunting for anyway, which positions the brand as the framer of the customer's own evaluation. This is the spreadsheet principle applied at the expensive, considered end of the market, and it works because the more expensive and researched the purchase, the more the customer wants evidence and the less they trust assertion.

Four brands, four costumes, one move. David transfers a comparison. AG1 transfers authority. Liquid Death transfers cultural credibility through humor. Eight Sleep transfers data to a high-intent researcher. None of them is asking you to please believe them. All of them are handing you something and letting you conclude. That is the pattern, and once you can see it operating across four completely different aesthetics, you can run it through whatever costume fits your own brand.

The hook is
the whole
ballgame.

There is a piece of all this that sits underneath every format and decides whether any of it works, and it gets far less attention than it deserves: the hook. The first frame, the first line, the first second. Get it wrong and the most beautifully constructed spreadsheet ad in the world dies unseen, because nobody scrolled far enough to see the grid.

Here is the uncomfortable reality of the feed. You are not competing for a decision. You are competing for a thumb that is already in motion. The default state of every viewer is scrolling past, and your ad has a fraction of a second to interrupt that default before it loses them forever. Every format in this piece is, at some level, a hook strategy. The spreadsheet ad hooks by looking like a useful tool. The Notes-app ad hooks by looking like organic content. The ugly ad hooks by looking unproduced. The founder ad hooks by looking like a real person with something urgent to say. The format is not separate from the hook. For static ads especially, the format is the hook.

For video, the hook is more explicit and more brutal, because you can measure exactly when people leave. The first two seconds carry an outsized share of the work, and the strongest creators front-load the entire value proposition before the three-second mark, because they know the audience is deciding to stay or go in that window. A hook that reads like organic content, a question, a provocation, a surprising claim, a pattern interrupt, will outperform a polished brand open at a fraction of the cost. The hook is also the cheapest thing to test, which is why smart accounts test hooks relentlessly, running the same body of an ad behind a dozen different openings to find the one that holds attention.

Let me give you the practical taxonomy of hooks that actually work in this environment, because briefing a team to "make a good hook" is useless and briefing them to make a specific kind of hook is not.

The contrarian claim hook opens by saying something that contradicts what the viewer assumes, which creates a small itch they have to scratch by watching. "Everything you have been told about protein is wrong" makes you stay to find out what. The discipline is that the contrarian claim has to be true and the ad has to actually deliver on it, or you have trained the audience to distrust your hooks.

The specific-number hook opens with a concrete, slightly odd number, because specificity reads as truth. "I tested 14 greens powders" is more arresting than "I tested a bunch of greens powders," because the precise number signals that real work happened. This pairs naturally with the spreadsheet and comparison formats.

The callout hook opens by naming exactly who the ad is for, which both filters and grabs. "If you are over 40 and lifting weights" makes the right person lean in precisely because it excludes everyone else. Relevance is attention, and the fastest way to be relevant is to name the person.

The in-progress hook opens mid-action, as though you caught a real moment rather than the start of a produced ad. The creator is already mid-sentence, already using the product, already reacting. This kills the ad-detection reflex because produced ads have clean openings and real moments do not.

The problem-agitation hook opens on the pain, not the product, naming the frustrating thing the customer lives with so precisely that they feel seen. This is the engine of the comparison-to-status-quo move from earlier, and it works because a customer who feels understood will stay to hear the solution.

The thing all of these share is that none of them lead with the product. They lead with the viewer, the problem, or a provocation, and they earn the right to mention the product by first holding attention. Brands that open on their logo and their product shot are opening on the two things the viewer cares about least, which is why they lose the thumb in the first half-second. Lead with the hook. Earn the rest.

A field guide
to the
formats.

I have covered a lot of formats, so let me consolidate them into something you can actually reference when you sit down to brief creative, because the value of a catalog is being able to pick the right tool deliberately rather than reaching for whatever is trendy.

Think of the formats along two axes. The first is what kind of evidence they transfer: data, social proof, authority, or cultural credibility. The second is where in the funnel they belong: awareness, consideration, or conversion. Every format has a natural home on both axes, and most of the expensive mistakes I see come from running a format outside its home, like firing a conversion format at a cold audience or using an awareness format to try to close.

The spreadsheet and comparison formats transfer data, and they live in consideration and conversion. They are for the customer who is already aware of the problem and weighing options. Do not run them cold. Run them at warm and mid-funnel audiences who are ready to evaluate.

The UGC demo and creator content transfer social proof and borrowed trust, and they span awareness and consideration. They are your volume workhorse, the format you produce the most of because it is the cheapest to make at scale and the most native to the feed. This is where most of your testing at-bats should come from.

The founder formats transfer authority and provenance, and they live at the top of the funnel, where a cold audience is deciding whether to care about you at all. They are your differentiation play, the thing no competitor can copy, strongest exactly when you are smallest and least known.

The native-mimicry formats, Notes app, text message, fake screenshot, transfer social proof and intimacy, and they span awareness and consideration. They are your attention-capture specialists, the formats that win the first second. Use them to get the scroll to stop, then make sure there is real substance behind the disguise.

The ugly ad and anti-ad transfer authenticity and distinctiveness, and they live at the top of the funnel. They are your pattern-interrupt and your cheap-volume engine, and they are the formats your cautious competitors are least likely to run, which is exactly why they have room to win.

Premium lifestyle production transfers aspiration and brand equity, and it spans awareness and consideration. It is the one format on this list that is not about looking un-ad-like, and it earns its place by building the brand credibility that makes everything else more believable. Do not abandon it. Just stop letting it be your whole library.

The way to use this field guide is not to run every format at once. It is to make sure that across your active creative, you have range on both axes: multiple kinds of evidence, multiple funnel stages, no single format carrying more than its share. When an account is struggling, nine times out of ten the diagnosis is not a bad ad. It is a library that is all one kind of evidence aimed at one stage of the funnel, which the algorithm reads as one concept and the audience exhausts in days. Range is the cure, and this catalog is how you build it deliberately.

Why volume
beat polish.

I have hinted at this several times, so let me make it explicit, because the economics of paid social changed underneath the creative trends and the two are inseparable. You cannot understand why these formats win without understanding the math of how creative gets tested now.

Start with the brutal baseline. Motion's 2026 creative benchmark report analyzed an anonymized dataset of more than 550,000 ads from over 6,000 advertisers, representing roughly 1.3 billion dollars of spend across Facebook and Instagram. The headline finding is one every operator should tape to their monitor: most ads fail, and that is normal. Depending on account tier, only about 4 to 8 percent of creatives are genuine winners. Somewhere around half are outright losers. The rest sit in the forgettable middle. Roughly 55 percent of total spend ends up flowing to the winning ads, and the share of spend concentrated on winners rises sharply with account sophistication, from around 23 percent at the smallest accounts to as high as 64 percent at the enterprise tier.

If you want concrete targets to grade your own creative against, the same body of data gives you two. The first is the thumbstop rate, sometimes called the hook rate, which is the share of impressions that turn into a three-second view, your hook doing its one job. On Reels and Stories you want 30 percent or higher, and feed placements typically land between 20 and 30 percent. The second is the hold rate, the share of those three-second viewers who stay to fifteen seconds, which tells you whether the body of the ad earned the attention the hook won. Average hold sits around 40 to 50 percent, above 60 is strong, and under 30 means the ad stopped the scroll and then wasted it. These two numbers diagnose almost every failing video ad. A weak thumbstop is a hook problem. A strong thumbstop with a weak hold is a body problem. Knowing which one you have is the difference between fixing the right thing and reshooting the whole ad for no reason.

Read that carefully and the strategy writes itself. If only 4 to 8 percent of your ads will win, and the winners carry the majority of your results, then your entire job is to find more winners. And the only reliable way to find more winners is to take more swings, because you cannot predict which creative will win in advance. Nobody can. The people who tell you they can are selling something. Even the best creative strategists are wrong most of the time, which is why the discipline is structured around volume of attempts rather than confidence in any single attempt.

This is the hidden engine behind every format in this piece. Top advertisers now test somewhere between 20 and 50 new creatives a week. The brands that test 60 or more creatives a month see roughly 2.8 times higher return on ad spend than brands testing fewer than 20. Read that again: the single biggest lever on ROAS is not targeting, not bidding, not even creative quality in the abstract. It is creative volume, the number of distinct shots you take, because more shots means more chances to find the rare winner that carries the account. Volume only pays, though, if the unit economics underneath it hold, which is a separate discipline: what you can actually afford to spend acquiring a customer comes down to your max allowable CAC and contribution margin, and you should know both before you scale a single winner.

Now connect it to the formats. Why did the spreadsheet ad, the Notes-app ad, the UGC clip, the ugly ad, and the founder note all explode at once? Because they are all cheap and fast to produce, and in a world where the winning strategy is volume, cheap and fast is everything. A studio shoot gives you a handful of expensive variations. A UGC pipeline, a comparison-template system, and a founder with a phone give you dozens of variations a week at a fraction of the cost. The formats that won are not just the ones audiences like. They are the ones a brand can manufacture in enough quantity to actually feed the testing machine. Production efficiency is now a creative strategy, not just a budget line.

And this is where Andromeda closes the loop. The algorithm rewards diversity of concept, the audience rewards native formats, and the economics reward cheap volume. All three forces point at the same operating model: produce a high volume of cheap, distinct, native-feeling creative, ship it in quantity, kill the losers fast, and pour budget into the small handful that win. Every trend in this piece is what that operating model looks like when it shows up in your feed. The spreadsheet ad is not winning because comparison is magic. It is winning because it is a cheap, repeatable, distinct, native-feeling format that lets a brand take a lot of swings, and taking a lot of swings is the whole game.

If you remember one thing from this section, make it this: stop trying to make the perfect ad. The perfect ad is a myth that costs you the volume you need to actually find winners. Make a system that produces a lot of good-enough, genuinely-distinct ads, and let the market tell you which ones are great. The market is a better creative director than you are, and it works for free.

Where AI fits,
and where
it does not.

I cannot write a piece about the economics of creative volume in 2026 without addressing AI, because the cheap-volume operating model I just described is exactly the thing generative tools are reshaping, and the brands that get the role of AI right have a real edge while the ones that get it wrong are quietly poisoning their own accounts.

Here is the honest version. AI is genuinely powerful for the production half of the equation and genuinely dangerous for the substance half, and telling the two apart is the whole skill. It is the same real-versus-theater line I drew for software founders in AI for Shopify app founders, just pointed at creative instead of product.

On the production side, AI is a volume multiplier, and volume is the game. The single hardest constraint on the operating model I described, produce a lot of cheap, distinct creative and let the market find the winners, has always been production capacity. You can only shoot so much, write so much, design so much. AI loosens that constraint dramatically. You can generate dozens of headline and hook variations in minutes. You can produce background variations, localize creative for different markets, and spin up static comparison layouts at a speed that was impossible two years ago. For the spreadsheet ad specifically, AI is a natural fit, because a comparison grid is structured, templated, and easy to regenerate with new rows and new framings. For hook testing, it is close to ideal, because testing twenty openings against one body is exactly the kind of high-volume, low-stakes variation generation that AI does well. Used this way, AI does not replace your creative strategy. It removes the production bottleneck that was preventing you from executing the strategy at the volume the algorithm rewards.

On the substance side, AI is a trap, and the trap is subtle because the output looks fine. The formats in this piece work because they feel real. The founder note works because a real founder had a real thought. The UGC demo works because a real person had a real reaction. The Notes-app ad works because it reads like something a human actually typed on their phone. AI is very good at producing things that look like these and have none of the underlying reality, and audiences are getting frighteningly good at detecting the difference. A slightly-too-smooth AI voice, a face that is almost right, copy that is grammatically perfect in a format that should be rough. Each of these snaps the viewer out of the native illusion, and the moment the illusion breaks, you have the worst of both worlds: an ad that was pretending to be authentic and got caught, which is less trustworthy than an honest ad would have been. There is also a real-imagery premium emerging precisely as a backlash. As feeds fill with AI creative, genuinely real, human, imperfect content stands out more, not less, because it is becoming the scarce thing.

The survey data on this is worth knowing, because there is a dangerous gap between what marketers believe and what customers actually feel. In a 2026 study, 82 percent of ad executives believed younger consumers felt positive about AI-generated ads, while only 45 percent of those consumers actually did, a perception gap of 37 points that has widened from 32 points in 2024. It gets sharper by generation: roughly 39 percent of Gen Z feel negative about AI ads, nearly double the 20 percent of Millennials. And when people are explicitly told an image is AI-generated, they engage less and view the brand more skeptically than when the same image is presented as human-made. The signal cuts the other way too. A January 2026 survey found 63 percent of US adults would trust AI search results less if those results contained ads, and a separate study found 75 percent would trust a recommendation less if they knew brand dollars influenced it. The lesson is not that AI is radioactive. It is that the realness has to be real, and the moment a customer suspects you faked it, the trust you were trying to borrow inverts into suspicion. That is the whole risk in one sentence: AI lets you fake authenticity at scale, and faked authenticity, once detected, is worse than honest advertising.

So the rule I give clients is simple. Use AI to multiply the variations of something real. Do not use it to manufacture the realness itself. Shoot the founder once, then use AI to generate twenty caption and hook variants. Capture the genuine customer reaction, then use AI to localize and reformat it. Build the comparison from true, verifiable data, then use AI to lay out and restyle the grid a dozen ways. The realness has to be sourced from the world. AI just helps you express it in enough variations to feed the machine. Get that division of labor right and AI is the best thing that has happened to high-volume creative. Get it wrong and you are using the most powerful tool ever built to make your ads less believable, faster, which is a remarkable way to lose money.

The operator's
playbook.

Theory is cheap. Here is how I would actually operationalize all of this if you handed me an ad account on Monday, written as the field guide I wish someone had handed me before I learned it the expensive way.

Start with format range, not creative volume. This is the counterintuitive part. Everyone hears "test more creative" and goes off to make 40 variations of the same concept, which Andromeda reads as one concept and serves to one audience. That is volume without range, and it is the most common expensive mistake I see. The first job is to build a portfolio of genuinely distinct formats: a spreadsheet or comparison concept, a founder note, a UGC demo, a native-mimicry execution, an ugly-ad hook, a premium lifestyle piece for brand equity. Six concepts that are actually different beats sixty that are secretly the same. Once you have range, then you add volume within each format.

Match the format to the funnel stage, because running the wrong format at the wrong stage wastes budget and, worse, gives you misleading data. Top of funnel, where cold audiences meet you, is for founder stories, POV hooks, behind-the-scenes content, lifestyle, and anti-ad creative, the formats that build awareness and stop scrolls. The middle is for testimonials, explainers, creator demos, and social-proof compilations, the formats that build consideration. The bottom, for warm audiences who already know you, is for comparison ads, spreadsheet ads, and offer-led urgency, the formats that close. The spreadsheet ad in particular is a consideration-to-conversion weapon, not a cold-traffic awareness play, because it rewards a viewer who is already weighing options. Fire it at a cold audience and it underperforms, not because the format is weak but because you aimed it at the wrong state of mind.

Match the format to the persona, not just the funnel. The same product legitimately speaks to four or five different people, and they need different creative. The customer who found you through a friend's recommendation needs a different ad than the one who found you through a problem-aware search. Mapping specific formats to specific personas rather than treating all your audiences as one interchangeable blob is where the real gains hide, and it is the part most brands skip because it is more work than spraying one message at everyone.

Build a production system, not a campaign. The brands that sustain this are not doing heroic one-off creative pushes. They have a repeatable pipeline: a comparison-ad template they can refill with new rows, a UGC brief they send to a roster of creators every week, a standing slot where the founder records two phone videos, a designer who can turn a customer review into a native-mimicry static in twenty minutes. The Seasalt Cornwall example is instructive: a brand with one of the tightest visual identities in UK fashion found that working within its existing approved photoshoot content, it could produce 50-plus distinct paid social ads per shoot batch without a single brand compromise. The format range did not require abandoning the brand. It required a system for extracting more distinct creative from assets they already had.

Kill losers fast and without ego. If only 4 to 8 percent of ads win, then most of your creative is going to die, and the speed at which you cut the losers determines how much budget survives to find the winners. This is emotionally hard, because someone on your team made each of those ads and is rooting for it. Build the discipline anyway. Set a clear threshold, give each ad a fair but finite test, and pull the ones that miss without relitigating them. The point of volume is not that every ad works. It is that you stop paying for the ones that do not, quickly, so the winners get the budget.

Measure the right thing. Active lifespan, how long an ad keeps getting spend before it fatigues, is one of the most honest performance signals there is, because brands do not keep funding ads that lose. Watch your winners' lifespans, because they tell you when fatigue is coming and when you need the next concept ready. And do not over-index on vanity engagement. A funny ad that gets a thousand laughing reactions and zero purchases is entertainment, not advertising. The job is conversion, and the metric that matters is the one tied to money.

Do not neglect lo-fi, and do not neglect hi-fi either. The data does not support an all-lo-fi library any more than it supports an all-polished one. The 42 percent lo-fi figure means the best libraries have both, in real proportion. Premium production still builds the brand equity that makes everything else more credible. Lo-fi still does the cheap-volume testing work. The mistake is treating it as a binary. Run both, weighted toward whatever your testing tells you, and let the proportion shift as you learn.

That is the playbook. Range first, then volume. Match format to funnel and persona. Build a system, not a campaign. Kill losers fast. Measure money, not applause. None of it is glamorous. All of it works.

The first
ninety days,
concretely.

Playbooks are easy to nod along to and hard to start. So here is the specific 90-day sequence I would run if you were a brand doing somewhere between a few hundred thousand and a few million in revenue and you wanted to rebuild your creative engine around everything in this piece. Adjust the pace to your resources, but keep the order, because the order is the part that matters.

The first thirty days are about establishing range and a production baseline, not about winning. The mistake everyone makes is trying to find a winner in week one. You will not, and chasing one will warp your decisions. Instead, the first month's job is to get one genuinely distinct concept live in each of the major format families. One comparison or spreadsheet concept built from real, verifiable data. One founder note or founder-to-camera piece, shot on a phone, no production. One UGC demo, either from an existing happy customer or a single commissioned creator. One native-mimicry static, a real review or a real founder thought in a native layout. One deliberately rough, distinct hook-led piece. That is five concepts across five families, which is more range than most brands in this revenue band have ever run at once. Get them live, let them gather a fair amount of data, and resist the urge to make twenty variations of any of them yet. You are mapping the territory, finding which families your audience responds to at all.

While those run, spend the first thirty days building the production system in parallel, because the system is what makes everything after day thirty possible. Set up a standing creator pipeline, even if it is just two or three people you can brief weekly. Build a comparison template you can refill with new rows in an hour, not a week. Schedule a recurring slot, fifteen minutes, where the founder records two phone videos against prompts you prepare. Create a simple intake for turning customer reviews and support messages into native-mimicry statics. The goal by day thirty is not a winning ad. It is a machine that can produce a high volume of distinct creative cheaply and repeatedly, because that machine is the actual asset.

Days thirty to sixty are about doubling down on what showed signal and adding volume within the winning families. By now the first five concepts have told you something. Maybe the founder format is over-performing and the premium lifestyle piece is dead. Maybe the comparison ad is holding attention but not converting, which tells you the format works but the rows are wrong. Take the two or three families that showed the most life and start producing volume within them: not the same ad twenty times, but twenty genuinely different executions of the formats that are working. New hooks on the UGC, new rows and new comparison sets on the spreadsheet, new founder prompts on the founder note. This is where the production system you built in month one earns its keep, because now you are asking it to produce range and volume at the same time, which is impossible without a system and trivial with one. Keep killing the losers fast. Most of what you make this month will still die. That is the math working as intended.

Days sixty to ninety are about scaling spend behind the proven winners while keeping the testing engine running underneath. By now you should have a small handful of genuine winners, the 4 to 8 percent the data predicts, and the job changes. You pour budget into the winners and watch their active lifespan like a hawk, because a winner at scale fatigues faster than a winner at low spend, and the moment you see the first decay you need the next concept ready. Meanwhile the testing engine keeps running at a steady cadence underneath the scaled spend, because the winners you are scaling today will fatigue, and the only insurance against that is a steady stream of new tests producing the next generation of winners before you need them. This is the steady state you are building toward: a layer of scaled, proven creative carrying the results, and a constant churn of cheap, distinct tests underneath it refreshing the pipeline. That two-layer system is what a mature creative operation actually looks like, and ninety days is enough to stand up the first version of it.

The thing I want you to take from this sequence is that none of it depends on a creative genius having a brilliant idea. It depends on building a system that takes a lot of disciplined swings at distinct formats, measures honestly, kills ruthlessly, and scales the rare winners. The genius, if there is one, is in the system and the discipline, not in any single ad. That is good news, because systems and discipline are buildable, and waiting for a creative genius is not a strategy. I have watched brands with mediocre individual ads dramatically outperform brands with beautiful ones, purely because the first brand ran a real testing system and the second fell in love with its own taste. Build the system. The winners come from the system.

Where it
goes wrong.

I would be doing you a disservice if I only sold you the upside, because every one of these formats has a failure mode, and some of them carry real risk that the LinkedIn version of this conversation never mentions. Here is where I have seen brands get hurt.

The first trap is faking the native format. The native-mimicry family works because it borrows the credibility of organic content. The instant you fabricate that credibility, you have crossed from clever into dishonest, and the line is not blurry. A Notes-app ad that contains a real founder thought is fine. A fabricated text-message exchange presenting a made-up customer as real is a lie. A screenshotted five-star review that the customer actually wrote is powerful. A fabricated review attributed to a person who does not exist is now flatly illegal in the US. The FTC's Consumer Reviews and Testimonials Rule took effect in October 2024, and it explicitly bans fake or AI-generated reviews, undisclosed insider testimonials, and the suppression of negative ones, with civil penalties of up to roughly 51,000 dollars per violation. Per violation, not per campaign. The agency has already run its first enforcement sweep, warning companies of possible violations, which tells you this is being actively policed, not theoretically prohibited. Stack that on top of the reputational damage when a brand gets caught faking social proof, which outlasts any short-term lift, and the math is obvious. Use the native formats to present real things in a native way. Never use them to manufacture things that did not happen. The format is a costume, not a license to invent.

The second trap is the comparison you cannot defend. I said this earlier and it is worth repeating because it is the most common own-goal with the spreadsheet format. The moment you make a specific, checkable comparison, you have invited scrutiny, and you have also invited your competitor's lawyers. Comparative advertising is legal in most places, but it has to be truthful and substantiated. If you claim more protein, you had better have the lab result. If you imply a competitor is worse, you had better be right and able to prove it. Brands get excited about the format and forget that a spreadsheet ad is a set of factual claims rendered as design, and factual claims have to be true and provable. The format raises your burden of proof, it does not lower it.

The third trap is mistaking native for unaccountable. Some brands read "ads should not look like ads" as permission to hide that an ad is an ad, and that is both an ethics problem and increasingly a disclosure problem. There is a difference between creative that does not pattern-match to advertising and creative that actively deceives the viewer about being a paid promotion. The first is smart. The second draws regulators. Creator content needs proper disclosure. Paid partnerships need to be labeled. You can be native and honest at the same time, and the brands that will survive the inevitable tightening of disclosure rules are the ones building that habit now.

The fourth trap is format-chasing without principle. This is the subtle one, and it is the trap most likely to catch a smart team. You read a piece like this, or you see a handful of spreadsheet ads go viral, and you go run one because spreadsheet ads are hot. But you ran it because it was trendy, not because your category turns on comparison or your product wins the grid. The format is a tool for executing a principle, and if you grab the tool without the principle, you get a technically-on-trend ad that does not actually fit your business. The brands that win are not the ones running the most fashionable format. They are the ones who understood why a format works and only used it where that why applied to them.

The last trap is forgetting the clock. I keep coming back to this because it is the thing operators most want to ignore. Every format in this piece has a half-life. They work partly because they are still novel and the audience has not fully learned them. As they saturate, performance decays, the schema adapts, and the edge erodes. This is not a reason to avoid them. It is a reason to stay in motion. The brands that get hurt are the ones that find a winning format, declare victory, and ride it until it is dead, then act surprised when the thing that worked for a year suddenly stops. Treat every format as a depreciating asset. Use it hard while it works, watch the performance for the first signs of fatigue, and always have the next concept in production before you need it.

The clock is also faster than it used to be, and that is not a vibe, it is measurable. Most concepts on Meta now follow roughly a ten-day decay curve: peak in the first three days, warning signs by days four to seven, and a 30 to 50 percent drop in click-through by days eight to ten. Andromeda made this worse, because it weights creative signals harder than the old system, so a concept that might have lasted six weeks two years ago now burns through its addressable audience in two or three. Frequency is the leading indicator to watch: in prospecting, performance starts sliding once weekly frequency passes about 2.5 and falls off a cliff past 4.0. And the cost of ignoring all this is real. Average Meta CPMs climbed past 13 dollars in 2026, up about 20 percent year over year, and fatigue compounds that fast. I have seen an account whose CPM ran from 22 dollars to 48 over three months purely from creative fatigue, then drop back to 25 within 48 hours of launching five fresh statics. That is the entire argument for the production system in one anecdote: the fix for rising costs is almost never a new audience or a new bid strategy. It is new creative, ready before you need it.

If you sell
software, not
supplements.

Everything so far has been framed through DTC, because that is where these formats are most visible and where the spreadsheet ad was born. But I advise a lot of Shopify app and SaaS founders, and I want to be clear that almost none of this is consumer-specific. The principle, evidence over persuasion, is, if anything, more powerful in B2B, because business buyers are even more allergic to being sold to and even more reflexively self-directed in their research.

The spreadsheet ad is a native B2B format, and most software companies are sleeping on it. The comparison grid is how software gets evaluated anyway. Every buyer of every tool builds, mentally or literally, a comparison of features, pricing, integrations, and support before they buy. So an ad that hands them that grid is not a gimmick, it is a shortcut to the exact mental model they were going to construct themselves. A Shopify app showing its feature set against the two incumbents it competes with, on the rows merchants actually care about, is the spreadsheet ad applied to software, and it works for the same reason: it transfers the conclusion to the buyer and lets the table do the arguing. The discipline is identical too. Win the rows you put in the grid, or do not run the grid.

The founder format might be even stronger in SaaS than in DTC. Software buyers want to know there is a serious person behind the product who understands their problem. A founder explaining why they built the tool, what was broken about the alternatives, and who it is and is not for does more to build trust with a skeptical B2B buyer than any feature list. The best early-stage software marketing I see is often just a founder being specific and honest about the problem on camera or in a written note. It does not scale forever, but at the stage where most app founders are fighting for their first thousand customers, it is the strongest creative play they have, and they almost always under-use it for the same reason DTC founders do: discomfort. If you are building one, this sits alongside the harder product and distribution questions I cover in how to build a Shopify app in 2026.

The native-mimicry and UGC formats translate too, just with different native references. The B2B equivalent of the UGC demo is the practitioner walkthrough, a real user showing how they actually use the tool in their real workflow. The B2B equivalent of the text-message ad is the screenshotted Slack message or the internal note where one colleague tells another about a tool that saved them. The B2B equivalent of the review screenshot is the G2 or app-store review rendered as creative. The references change. The mechanism, borrowing the credibility of organic content and transferring the conclusion to the buyer, does not.

The one genuine difference is the funnel length. A consumer can see a spreadsheet ad and buy a 40-dollar supplement an hour later. A business buyer sees the same ad and begins a six-week evaluation involving three other people. So in B2B, these formats are less about driving an immediate purchase and more about entering and winning the consideration set, planting the comparison frame early so that when the formal evaluation happens weeks later, your grid is the one the buyer remembers building. The formats work. You just have to be honest that they are starting a longer conversation, not closing a fast one, and measure them accordingly. Judging a B2B founder ad by next-day conversions will make you kill the thing that was actually working.

Static versus
video, and why
the grid is still.

A practical question I get constantly: should these be static images or video? It matters, because it changes how you produce them and how you spend, and the answer has more to it than the "video is king" mantra that everyone repeats.

Here is the operating reality at the top of the market. The highest-performing DTC accounts run static and video in parallel, and they use them for different jobs. Statics are the cheap, fast way to find the winning message. Video is the way you amplify a message you have already proven. You discover with statics and you scale with video, and brands that skip the discovery step and pour money straight into expensive video are guessing about the message with the most expensive format to produce, which is exactly backward.

Notice which formats in this piece are naturally static. The spreadsheet ad is almost always a static image, because a comparison grid is a thing you read, not a thing you watch, and forcing it into video usually makes it worse. The native-mimicry formats, the Notes-app screenshot, the review screenshot, the fake text thread, are static by nature, because they imitate things that are static. The ugly ad is often static. These formats are static partly because their substance suits a still image and partly because static is radically cheaper and faster to produce, which feeds the volume game. You can make and ship a comparison static in an afternoon. A polished video takes a week and a budget.

The video-native formats are the founder-to-camera, the UGC demo, and the creator content, because those depend on a human delivering something with energy and timing that a still cannot capture. Trust transfer through a person mostly needs motion and voice. So the rough division is: data and social-proof formats lean static, human and demonstration formats lean video, and a healthy library has both because they do different jobs at different stages.

The strategic implication is that you should be running far more statics than most brands do, specifically for message discovery, because statics let you take the most at-bats for the least money. Use them to find which hook, which comparison, which claim actually moves people. Then take the winning message and build the more expensive video around it, confident you are amplifying something that already works rather than gambling production budget on an unproven idea. The spreadsheet ad fits this perfectly: it is a cheap static that often tells you, fast, which comparison frame the market responds to, and that learning then informs every other format you build. Treat your statics as the research lab and your video as the factory that scales what the lab proved.

The channel
changes the
creative.

Almost everything I have said so far is framed through Meta, because that is where the spreadsheet ad was born and where most DTC spend still lives. But a lot of brands take their Meta creative, reupload it to TikTok unchanged, and then wonder why it dies. So I want to be explicit about how the channel changes the execution, because the principle holds everywhere while the format does not travel as cleanly as people assume.

TikTok punishes ad-shaped creative even harder than Meta does. The platform's audience and algorithm both actively reject content that does not look like it belongs, and users have developed an extremely fast reflex for "brand content that does not fit here." A polished spot that might scrape by in the Meta feed gets scrolled past on TikTok in well under a second. The flip side is that TikTok rewards watch time and interaction over production quality more aggressively than any other platform, which means the lo-fi, native, founder-and-creator formats from this piece are not just viable there, they are mandatory. If Meta tolerates a produced ad, TikTok demands a native one. And if you sell through TikTok Shop on top of running ads there, the platform-risk question is worth weighing separately, which I worked through in is TikTok Shop safe for sellers in 2026.

The format that encodes this best is the Spark Ad, which boosts an existing organic post with paid distribution while keeping the original post's engagement and social proof intact. A clip that already has fifty thousand organic views runs as an ad with that view count and those comments still attached, which is social proof a standalone ad simply cannot manufacture. The numbers are convincing: Spark Ads have been shown to drive around 142 percent higher engagement than standard ads when used well, and Spark Ads built from posts that already had more than 500 organic engagements saw roughly 52 percent lower cost-per-click and 45 percent higher landing-page visit rates than ones built from freshly published posts. That is the evidence-over-persuasion principle again, expressed through the channel's own mechanics: the ad arrives wearing proof it earned organically rather than proof the brand asserted.

The founder format is even more potent on TikTok than on Meta. Founder and origin-story content has been reported to deliver three to four times the engagement and as much as 60 percent lower acquisition cost on TikTok, because raw authenticity and a real person telling a real story is exactly what the platform's culture rewards. The same founder note that works as a static on Meta wants to be a talking-to-camera clip on TikTok, shot on a phone, no polish. The substance is identical. The packaging is native to where it runs.

The practical rule I give clients is to stop thinking of a creative concept as a single asset and start thinking of it as a message that gets re-clothed per channel. The comparison you built for a Meta spreadsheet static becomes a creator talking through the same comparison on TikTok. The founder letter becomes a founder clip. The customer review becomes a duet or a screen-recorded reaction. You are not making different ads for each platform. You are making one argument and dressing it in each platform's native costume, because the argument is portable and the costume is not. Brands that grasp this run the same strategy across channels at a fraction of the effort. Brands that miss it either reupload Meta creative to TikTok and watch it flop, or build everything from scratch per channel and burn out their team. The middle path, one message and many native costumes, is the whole game. The prior question, how much budget each platform should even get, is its own decision, and I laid out how I think about it in your channel mix is your strategy.

Where the
feed goes
next.

Let me put my neck out, because anyone writing about trends owes you a view on where this is heading, not just a snapshot of where it is. Predictions are how you find out whether someone actually understands the mechanism or is just describing the present.

The native-mimicry formats will be the first to decay, and the decay is already starting. The Notes-app ad, the fake text thread, the screenshot styles, these work on a disguise, and disguises have the shortest half-lives because the audience learns them fastest. Within a year or two, the average viewer will pattern-match a Notes-app ad as an ad the same way they now pattern-match a banner. The brands running lazy versions of these formats are accelerating that decay for everyone. They will keep working in the hands of brands whose underlying message is genuinely native to the form, a real founder thought, a real customer message, but the free ride of the format alone is ending.

The comparison and spreadsheet formats will last longer, because they are not really a disguise, they are a genuinely useful information structure. A comparison grid keeps working even after the audience knows it is an ad, because it still does the cognitive work of framing the decision, and usefulness does not decay the way novelty does. What will happen instead is an escalation: as more brands run comparison ads, the comparisons themselves will have to get more honest and more specific to stand out, because a vague or self-serving grid will read as exactly the kind of marketing the format was supposed to escape. The bar for what counts as credible evidence will rise. That is good for customers and hard for brands with weak substance, which is the right outcome.

The founder format will keep working roughly forever, because it is not a trend, it is a permanent feature of how humans extend trust. The specific executions will evolve, but a credible person explaining why they built something will outperform a faceless brand for as long as people are people. The constraint is just the supply of founders willing and able to be on camera, which is why this format is most available to the brands with the courage to use their most differentiated asset.

The biggest shift coming is the one I flagged in the AI section: a growing premium on the genuinely real. As feeds saturate with AI-generated and AI-assisted creative, the scarce and valuable thing becomes content that is unmistakably, verifiably human. I think we are heading toward a market where "proof of real" becomes a competitive advantage, where brands that can demonstrate their creative features actual people, actual data, and actual customer experiences will outperform brands running polished synthetic content, simply because the audience will have learned to distrust anything too smooth. The evidence-over-persuasion principle gets sharper here, because the most powerful evidence in an AI-saturated feed is evidence that is obviously, checkably real. The same AI shift is also changing where buyers discover products in the first place, moving the first touch from the feed toward AI assistants, which I dig into in AI shopping assistants are becoming the new storefront.

If I am right about all of that, the meta-lesson holds. Do not bet on a format. Bet on the principle, source genuine substance from the real world, and stay in motion as the specific costumes wear out. The brands that do that will keep winning through a dozen format cycles. The brands that chase the costume will win for a few months at a time, forever one trend behind, wondering why the thing that worked last quarter suddenly does not.

The least seductive
object in the world
is winning.

Let me bring it back to where we started, because the spreadsheet ad is a perfect ending as well as a beginning.

A spreadsheet is the least seductive object in the world. It is a grid. It has no music, no aspirational lighting, no celebrity, no clever tagline. And right now it is one of the highest-performing ad formats in direct-to-consumer, run by brands pacing toward nine figures. That fact should permanently change how you think about advertising, because it tells you the thing that wins is not the thing that dazzles. It is the thing that respects the customer's intelligence and hands them the evidence to decide for themselves.

That is the whole shift, compressed into a single object. The era of "please believe us" is over, not because brands got more honest, but because customers got more capable. They can check everything. They have a reflex that filters out anything that looks like a pitch. And they trust their own conclusions far more than they will ever trust yours. The brands winning right now are the ones that stopped fighting those facts and started building around them, making ads that look like content, that hand over data, that feature real people, that invite the customer to do the persuading themselves.

The specific formats will come and go. The spreadsheet ad will have its moment and saturate. The Notes-app ad will get recognized and decay. Something new will replace them, and then something will replace that. If you take the formats as the lesson, you will spend the next five years chasing a feed that has already moved on. If you take the principle as the lesson, you will be the one inventing the formats everyone else copies. Evidence over persuasion. Native over produced. The customer's conclusion over the brand's assertion. Build around that and the formats take care of themselves.

I have spent fifteen years on every side of this, as an operator buying the media, an investor pricing the brands, and an advisor fixing the accounts. The single most reliable pattern I have seen is that the brands who treat their customers as smart, capable people who can be trusted with the evidence outperform the brands who treat them as marks to be convinced. The spreadsheet ad is just the current, unusually literal proof of that. Make ads that hand people the data and step back. Then watch what they conclude.

What exactly is a spreadsheet ad?

A spreadsheet ad is a comparison rendered as a grid or table, usually static. Your product runs down one column, the alternatives down the others, and the rows are the attributes customers actually weigh, like price, key ingredients, or a headline spec. The visual does the persuading: your column reads well, theirs reads worse, and the customer reaches the conclusion themselves. It works because humans evaluate value comparatively and because a table feels like neutral research rather than a sales pitch.

Why are spreadsheet ads suddenly everywhere in 2026?

Three forces converged. Audiences have a trained reflex that filters out anything that looks like an ad, so native, non-ad-shaped formats win attention. Meta's Andromeda delivery system now rewards creative that is distinct from other ads, and a spreadsheet looks nothing like standard creative. And the economics of paid social reward cheap, high-volume, repeatable formats, which a comparison template is. The spreadsheet ad sits in the sweet spot of all three.

Do these formats only work for DTC brands?

No. The underlying principle, transferring the conclusion to the customer instead of asserting it, works just as well in B2B and SaaS, and arguably better, because business buyers are even more research-driven and ad-averse. The spreadsheet ad maps directly to how software gets evaluated, and the founder format is one of the strongest early-stage SaaS plays there is. The main difference is the longer funnel, so you measure these as consideration-set entry rather than immediate conversion.

Is it legal to compare my product to a competitor in an ad?

Comparative advertising is legal in most jurisdictions, but every claim has to be truthful and substantiated. If you state a spec, you need the data to back it. If you imply a competitor is worse, you need to be able to prove it. The spreadsheet format raises your burden of proof because it renders specific factual claims as design. Treat every cell as a claim you might have to defend, because you might.

What is the biggest mistake brands make with these formats?

Chasing the format without the principle. Running a spreadsheet ad because spreadsheet ads are trendy, rather than because your category turns on comparison and your product genuinely wins the grid. The format is a tool for executing a strategy. Grab the tool without the strategy and you get an on-trend ad that does not fit your business. The second biggest mistake is faking native formats, like fabricating reviews or text exchanges, which is both dishonest and increasingly illegal.

How many creatives should I actually be testing?

More than you are. Top advertisers test 20 to 50 new creatives a week, and brands testing 60-plus a month see roughly 2.8 times the return on ad spend of brands testing fewer than 20. But test range, not just volume: six genuinely distinct concepts beat sixty variations of one, because the algorithm clusters lookalike creative and serves it to a single audience. Build format diversity first, then add volume within each format.

Do these formats work the same on TikTok as on Meta?

The principle travels, the packaging does not. TikTok rejects ad-shaped creative even harder than Meta, and it rewards watch time and raw authenticity over production, so the native and founder formats are mandatory there rather than optional. The strongest-performing TikTok format is the Spark Ad, which boosts an organic post with its real engagement and social proof intact, and founder and creator content tends to outperform by a wide margin. Do not reupload Meta creative to TikTok unchanged. Take the same message and re-clothe it in the platform's native style.

What hook rate and hold rate should I be aiming for?

On video, aim for a thumbstop or hook rate of 30 percent or higher on Reels and Stories, and 20 to 30 percent in feed placements. For hold rate, the share of three-second viewers who stay to fifteen seconds, average is 40 to 50 percent and strong is above 60. These two numbers diagnose most failing ads: a weak thumbstop is a hook problem, while a strong thumbstop with a weak hold is a body problem. Fix the one you actually have.

Will these ad styles stop working?

Yes, eventually, individually. Every format has a half-life. They work partly because they are still novel, and as they saturate, audiences learn them and performance decays. That is not a reason to avoid them, it is a reason to stay in motion: use each format hard while it works, watch for the first signs of fatigue, and always have the next concept in production. The formats are depreciating assets. The principle underneath them, evidence over persuasion, does not depreciate.

If you want to go deeper on the numbers that decide whether any of this creative is affordable in the first place, start with the DTC profitability teardown and CAC payback by vertical. Then tell me what your ad account looks like and I will tell you where the creative is leaking.