The 2026 numbers
for brands and apps

The most common question I get, from a DTC founder or a Shopify app builder, is some version of "is this good?" Is a 2% conversion rate good. Is 5% monthly churn a problem. Is my CAC payback too slow. So I put the numbers I actually use in one place. This is the reference I reach for when I read a brand or an app, and it covers both, because I have run both.

Most benchmark posts pick one side. This one spans the whole Shopify ecosystem because my seat has: I helped build the Shopify partner ecosystem, co-founded and scaled a brand portfolio into nine figures, and founded and sold a software business. Brand economics and app economics get read side by side here on purpose, and by the end you will see how much they rhyme.

One rule up front: every number is labeled. Where a figure is a published industry median, I cite and link the source. Where it comes from the brands and apps I have operated and advised, I say so and you can weigh it accordingly. If you want to run your own numbers as you read, the free calculators cover most of what is below, from DTC profitability to app CAC payback.

A word on why this is one report and not two. The temptation across the ecosystem is to treat brands and apps as separate worlds with separate advisors. They are not separate. The same forces move both, the same mistakes sink both, and increasingly the same people move between them, brand operators building tools and app founders buying brands. Reading the two sets of numbers together is not a gimmick, it is how the sharpest operators in this ecosystem already think.

Use it however fits. If you run a brand, the app section is a window into your vendors' economics, worth understanding before you sign another annual contract. If you build an app, the brand section is your customer's profit and loss, and knowing where their money actually goes will make you a better partner and a better product. And if you are somewhere in between, which more people are every year, the whole thing is your map.

Two kinds of
numbers, labeled
honestly.

There are two kinds of benchmarks in this report, and I keep them separate so you always know what you are looking at. The first kind is a published median: conversion rate, add-to-cart rate, return rate, app churn, free-to-paid conversion, valuation multiples. These come from public research, and I link the source next to the number. The second kind is operator-derived: brand EBITDA bands, platform and software cost as a share of GMV, the cash conversion cycle, the valuation anchor I use in practice. Those come from the brands and apps I have run and advised, and I flag them as experience, not industry fact.

The second rule matters more than the first: a median is a starting line, not a target. If your conversion rate matches the median, you have not arrived, you are average. Benchmarks are good for one thing above all, finding your single binding constraint, the one number holding the business back right now. Read every table below asking "which of these is furthest from healthy for me," not "am I average." The rest is just knowing where the lines sit.

Operator numbers and public medians, and why both belong

A fair question: if some of these numbers come from published research and others from my own experience, why mix them? Because neither alone is enough. Public medians are broad but blunt. They tell you what the middle of a huge, messy sample looks like, but they average across business models that have little to do with each other, and they lag, sometimes by a year or more. Operator numbers are narrow but sharp. They come from a smaller set of businesses I have actually run the books on, which means they carry judgment the surveys cannot, but they are one person's field of view, not a census.

So I use them for different jobs. Where a public median exists and is current, it anchors the number and you get the link to check my work. Where the honest answer lives in operating experience, brand EBITDA being the clearest case since most DTC brands are private, I say so and give you the range I have actually seen rather than dressing it up as an industry statistic. The one thing I will not do is launder experience as data or present a median as a target. Both are common failures in benchmark content, and both are why so much of it is quietly useless.

The metrics in plain language

Before the tables, quick definitions so we are using the same words. If you know these cold, skip ahead to the brand benchmarks.

What good looks
like for a
DTC brand.

The brand-side reference table is below. Read the "healthy" column as the bar to clear, and the "watch or risk" column as the line where I start asking hard questions. The basis column tells you whether the number is a sourced public median or drawn from the brands I have operated and advised.

The table is the map. The rest of this section is the terrain: what each number means, what moves it, and where I have watched brands fool themselves. Take them in order.

Conversion rate, the median is a trap

The median Shopify store converts 1.4 to 1.8% of visitors, best-in-class stores clear 3.7%, and the top tenth run 4.7% or higher. For global context, Statista put worldwide ecommerce conversion near 1.6% in late 2025, so the Shopify median sits right on the world line. The trap is what happens when you hit it. You feel fine. Average is not a strategy, and the distance from 1.8% to 3.7% is worth more revenue than almost anything else you can do to the top of the funnel.

What closes that gap is rarely more traffic. It is site speed, offer clarity, and product-page quality. Mobile converts 30 to 40% below desktop, so if your traffic skews mobile, your blended number is being dragged down by an experience problem you can actually fix. The median also hides enormous category spread, with food and beverage clearing roughly twice what luxury does, so benchmark against your vertical rather than the global line. The conversion benchmark breakdown has the by-category numbers, and the conversion revenue leak calculator turns your gap into a dollar figure.

Add-to-cart rate, the upstream tell

Add-to-cart sits one step upstream of conversion and it is often the more honest signal of intent. The ecommerce median is about 6.3% of sessions. Under 4.5% and the leak is usually upstream of checkout entirely: a price objection, a weak product page, or traffic that was never going to buy. Fixing checkout when the real problem is at add-to-cart is a common and expensive misdiagnosis, because checkout optimization is visible and satisfying while the actual loss is happening two steps earlier. Category spread is wide here too, food and beverage near 10% and luxury under 2%, so read it against your own vertical (Dynamic Yield).

The reason add-to-cart is worth watching separately is that it splits two very different problems a bare conversion rate blends together. If add-to-cart is healthy but checkout conversion is weak, the problem is in the funnel: shipping surprises, a clunky checkout, a trust gap at the payment step. If add-to-cart itself is weak, the problem is earlier and more fundamental, the offer, the price, or the product page is not earning intent at all. Fixing the second with tactics meant for the first is how teams burn a quarter optimizing a checkout that was never the bottleneck.

EBITDA margin, where profitable-growth brands separate

This is the number I spend the most time on, because it is where profitable-growth brands separate from the rest. The bands here are operator figures, not a public dataset, since most DTC brands are private and do not publish. Across the seven-figure brands I have worked with, the EBITDA median lands near 4%. Eight-figure brands run closer to 7%. Healthy, the number I want a brand managing toward, is 10% and up, and mid-teens is genuinely top-tier.

The gap between a brand that is growing and a brand that is keeping money is almost always contribution margin discipline, not revenue. I have watched brands triple revenue and shrink profit in the same year, because every incremental order was sold at a contribution margin that could not carry the overhead the growth required. If you have never traced a real per-order profit and loss, start with contribution margin, three layers deep, then run the DTC profitability calculator on your own numbers.

I hold a strong view here, shaped by operating rather than theory: chasing revenue growth while margin erodes is the most common way good brands quietly become bad businesses. Growth is intoxicating and it photographs well, but a brand that grows 60% while EBITDA falls from 8% to 2% has made itself larger and more fragile at once. The discipline is to grow and hold the margin line together, which is slower, less glamorous, and the only version that compounds into something durable and worth owning.

LTV:CAC, the ratio everyone quotes

The 3:1 LTV:CAC floor is the most-quoted number in the room, and it is worth knowing where it came from: David Skok popularized it around 2010 at Matrix Partners, drawn from mature public software companies at steady state. It still holds as a floor, and elite operators push for 4:1 and better. Two cautions. First, compute LTV on a gross-margin basis, not on revenue, or the ratio flatters you by a factor of two. Second, the ratio says nothing about timing, and timing is where brands run out of cash. A 3:1 ratio that takes eighteen months to realize can bankrupt a brand that a 2.5:1 ratio paying back in four months would not. That is why I lean harder on payback, next.

The other thing the ratio hides is how fragile the LTV half of it is. Lifetime value is a forecast, and forecasts flatter the person making them. A small, optimistic assumption about repeat rate or average lifespan can double the LTV in a model, which doubles the ratio, which justifies spending you cannot actually afford. I have watched brands talk themselves into aggressive acquisition on an LTV:CAC that looked great and rested entirely on a retention assumption that had not happened yet. If you are going to trust the ratio, stress-test the LTV against your real, cleaned cohort data first.

CAC payback, the number I trust more than the ratio

Payback is the honest version of the acquisition question, because it is denominated in the thing that actually kills brands: cash and time. Healthy is recovering acquisition cost in under 6 months. Over 12 months is dangerous unless your retention is genuinely exceptional, because you are financing a year of someone else's growth out of your own working capital. The reason payback beats the ratio is that a DTC brand pays for inventory and ads now and collects contribution slowly, so a long payback quietly eats the cash you needed for the next inventory buy. Set your spend ceiling from contribution with the max allowable CAC formula and the calculator, then read payback by vertical since a healthy number for supplements is not a healthy number for furniture.

Return rate, the line most P&Ls ignore

Returns are the line most brands leave off the profit and loss entirely, and it is a big one. The NRF put the 2025 online return rate at 19.3%, and apparel runs higher, 20 to 40%, driven by bracketing, where shoppers order three sizes intending to send two back. The NRF also flags that roughly 9% of returns are fraudulent. The number itself matters less than whether you price it in. A brand that recognizes returns only when they arrive is overstating margin on every order that ships, then getting surprised by a lumpy cost it could have smoothed. Apply a blended return allocation to every order and the surprise disappears. The returns cost calculator shows what the line really costs once you price it honestly.

Returns also deserve a second look because not all of them cost the same. An item you can put straight back on the shelf costs you reverse shipping and a little handling. An item you have to discount, refurbish, or write off costs far more, sometimes the entire margin. So the number to manage is not the gross return rate, it is the net economic cost of returns, and two brands with the same 20% rate can take completely different P&L damage depending on how much of that inventory is recoverable. Model the net cost, not the headline rate.

Cash conversion cycle, the silent growth killer

This one is operator-derived and it is the metric that ends more good brands than any other, precisely because it is invisible on the profit and loss. The cash conversion cycle is how long your money is tied up between paying for inventory and collecting from the customer. Under 30 days is strong. Over 75 days is a real risk, especially for an inventory-led brand with long lead times. The cruel part is that a profitable, fast-growing brand can run out of cash faster than a slow one, because every new order requires an inventory buy before the last one has paid back. Growth consumes cash, and the cycle is where you see it coming. Model it before you place the next big purchase order with the inventory cash flow calculator.

The lever most brands miss on the cash cycle is terms. You can shorten it from either end, faster collection or slower payment, and the payment side is usually where the room is. Negotiating net-30 with suppliers, or a deposit-and-balance structure on production, can move the cycle by weeks without touching sales at all. This is unglamorous finance work, and it is often the difference between a growth year you can self-fund and one where you are borrowing at a bad rate to buy the inventory your own success requires.

Platform and software cost, the creeping tax

The last brand-side line is the one that creeps: the total cost of your platform plus every app and tool, as a share of GMV. Under 4% is healthy, over 6.5% is where I start auditing. It creeps because no single app feels expensive, but a stack of thirty of them, each taking a monthly fee or a slice of orders, adds up to a tax you stopped noticing. Worse, apps cost you twice, once on the invoice and once in the milliseconds they add to your page load, which drags the conversion rate from the top of this section. Audit the stack in dollars and in load time with the lens from the app bloat post, and if you are weighing the jump to enterprise pricing, the Shopify Plus cost calculator models the threshold.

Why brand EBITDA lands where it does

It is worth seeing why a brand with a healthy-looking 62% gross margin ends up at 4 to 7% EBITDA, because that walk is the whole point of the margin line. Start with the order at 62% gross margin. Payment fees take two to three points. Fulfilment and shipping, even well negotiated, take ten to fifteen points of the order value. A blended returns allocation on an apparel-skewed brand takes several more. Now you are at contribution margin and you have not spent a dollar on marketing yet. Subtract a realistic blended acquisition cost and contribution per order is often a thin single-digit share of revenue. Then the whole fixed overhead of the business, team, warehouse, software, comes out of that. What survives to EBITDA is the 4 to 7% the operator bands describe. Nothing went wrong in that story. That is simply what the math does, and it is why the brands that clear 10% are running every one of these lines on purpose, not by luck.

What good looks
like for a
Shopify app.

Now the app side. If you build on the Shopify App Store, these are the numbers that decide whether the business compounds or leaks. Same table logic: healthy is the bar, watch or risk is where I start worrying, and the basis column separates sourced medians from operator experience.

Same structure as the brand side: the table gives you the lines, and the walk-through below gives you what each one is really telling you.

Free-to-paid, two numbers with one name

Free-to-paid is really two different benchmarks, and the difference between them is a credit card. A billed trial that asks for a card up front converts at 25 to 35% when onboarding is good, and 50 to 60% when it is great. A no-card freemium model converts at 3 to 5%. The public data lines up: a 2026 ChartMogul study across roughly 200 products put opt-in trials near 8.9%, card-required trials at 31.4%, and freemium around 5.6% (First Page Sage puts organic opt-in near 18% and opt-out near 49%). Neither model is better in the abstract. The card filters out tire-kickers and lifts the conversion percentage, but it shrinks the top of the funnel, so fewer merchants ever start. Judge your number against the model you actually run, not the other one.

Whichever model you use, the lever is the same: time to value in the first session. The billed trials that hit 50%-plus almost always get a merchant to a visible win before the trial clock matters. The onboarding benchmarks post covers what that looks like, and the free-to-paid calculator models both trial types.

One more thing the free-to-paid number hides: the quality of the merchants it converts. A billed trial that converts at 30% but attracts price-shoppers who churn in month two is worse than a 22% trial that converts committed merchants who stay for years. Conversion and retention are the same funnel viewed at two points, and optimizing the first while ignoring the second is how apps end up with a strong headline trial number and a base that leaks. Judge the trial by who is still paying six months later, not by who signed up.

Monthly churn, the first thing I check

Churn is the first number I look at on any app, because it silently sets the ceiling on everything else. For SMB-focused Shopify apps, 3 to 5% monthly is normal, under 3% is strong, and over 6% is a fire to put out before you touch acquisition. The math is unforgiving: 6% monthly compounds to losing more than half your base in a year, so you are refilling a bucket faster than you can grow it. Shopify apps churn fast because merchants pay monthly and decide monthly, and because a big share of losses land in the first ninety days, before the merchant ever saw value. That makes churn and onboarding the same problem wearing two labels. The churn benchmark post and the churn cost calculator put a dollar figure on each point of churn.

When churn is the problem, resist the urge to fix it with discounts and win-back campaigns, which only treat the symptom. High early churn almost always traces to a value gap: the merchant installed for a reason, did not get there fast enough, and left. That makes onboarding, not retention marketing, the real lever. Every point of churn you prevent in the first ninety days is worth more than any amount you claw back later, because those merchants never build the habit that would have made them stick.

Net revenue retention, the metric behind the metric

Logo churn alone can lie to you. Net revenue retention tells you whether the accounts that stay are expanding fast enough to outrun the ones that leave. Above 100% means your existing book grows before you sell a single new seat, which is the quiet engine behind every app that compounds. An app losing 4% of logos a month with 110% net revenue retention is healthier than one losing 3% with 95%, because the first one grows on autopilot and the second one has to sprint just to stay flat. It is also the first thing a serious buyer checks, which is why it sits directly under churn in how I read an app.

The practical trap with net revenue retention is treating expansion as a monetization afterthought instead of designing it in from the start. The apps that clear 110% did not bolt on an upsell late, they built pricing that grows naturally as the merchant grows, through usage, seats, or tiers tied to the value delivered. If your pricing is a flat monthly fee with no path up, your net revenue retention is capped at whatever churn subtracts, and no amount of customer-success effort will push it over 100%. Expansion is a pricing-design decision first and a sales motion second.

CAC payback for apps

App payback runs longer than brand payback and that is fine, because software has no cost of goods on the next unit. Under 12 months is healthy, under 6 is fuel you can pour on with confidence, and over 18 is a warning. Worth knowing: the market has moved against you here. Private B2B software has seen median CAC payback stretch toward 18 to 20 months, and Bessemer's efficiency bar now pairs a sub-18-month payback with an LTV:CAC above 3:1 as the minimum for efficient growth. So a 12-month payback that felt merely fine a few years ago is now genuinely above average. Because SMB merchants pay monthly and churn monthly, a long payback is more dangerous for a Shopify app than for an enterprise tool, since you may never reach payback before the merchant leaves. The CAC payback calculator ties it to your own churn.

Payback and churn have to be read together, because they fight each other directly. If your payback is 14 months and your average merchant stays 17, you earn only three months of profit per customer before they leave, which is not a business, it is a treadmill with good branding. Lengthening payback is only safe when retention lengthens to match. Any time payback drifts up, the honest question is whether your customers are actually staying long enough to reach it, and if they are not, the fix is retention, not more efficient acquisition.

Shopify revenue share, model it before the cap

This is the number founders forget until it bites. Under the current Partner Program terms, Shopify takes 0% on your first $1,000,000 of lifetime App Store earnings, then 15% above that (Shopify dev docs). Read the details, because they changed: the exemption is lifetime now rather than the old annual reset, earnings before 2025 do not count toward the threshold, and the largest developers, those over $20M in prior-year App Store revenue or $100M in company revenue, pay 15% on everything. The trap is pricing your app as though the 15% does not exist, then watching your contribution margin step down the day you cross $1M lifetime. Price for the post-cap world from the start. The revenue share calculator shows exactly where the step lands.

Practically, pricing for the post-cap world means building your target margin around the 15% from day one, then treating the 0% band as a temporary tailwind that funds early growth rather than a permanent feature of the model. Founders who do the opposite, who set prices that only pencil at 0%, end up needing a painful increase right when they are also trying to scale past a million in revenue, which is the worst possible moment to ask every existing merchant to pay more.

Valuation, what actually moves the multiple

Valuation is where operator judgment earns its keep, because the public ranges are wide and the number you get is mostly about factors, not formulas. For small software, the rough public picture is 3 to 5x profit for anything under about $1M ARR, low-single-digit ARR multiples for micro apps, and private SaaS medians near 4.8x ARR (Aventis Advisors). Growth dominates: businesses growing above 40% command 7 to 10x while sub-20% growth gets 3 to 5x, and companies clearing a Rule of 40 score traded at a median above 10x revenue. My working anchor for a Shopify app is roughly 4x profit or 4.5x ARR, then adjusted up or down for the four things buyers actually pay for: growth rate, net revenue retention, how concentrated your merchant base is, and whether the business runs without you in the loop. Founder-dependent and concentrated discounts hard. The what Shopify apps sell for breakdown and the valuation calculator go deeper on each factor.

If a sale is anywhere on your horizon, the single highest-return thing you can do is reduce how much the business depends on you personally. Buyers pay a premium for a company that runs without its founder and discount hard for one where the founder is the product, the sales team, and the roadmap all at once. That means documenting, hiring ahead of the sale, and proving the machine works when you step back. It is the same durability that moves the multiple for a brand, and it is worth starting two years before you think you need to.

Why churn quietly caps everything

The compounding math is what makes churn the first thing I check on an app. At 3% monthly churn, the average customer stays about 33 months. At 6%, that halves to under 17 months. Double the churn and you have halved the lifetime value of every customer you will ever acquire, which halves what you can afford to spend to get them, which means your whole acquisition engine now has to run at half the cost or it stops working. Nothing else on the app table has that kind of leverage over everything else. It is also why an app with mediocre acquisition and strong retention beats a great-acquisition, leaky-retention app almost every time: the first compounds and the second treadmills. Retention is not a customer-success task you delegate and forget. It is the core economic engine, and it belongs on the founder's desk next to pricing.

Brands and apps
rhyme more than
they look.

Put the two tables next to each other and the same four questions show up on both sides, just under different names. This is the part most people miss, and it is why I keep both sets of numbers in one head. If you operate one and are curious about the other, this mapping is the bridge.

This is not a stretch to make the report longer. The overlap is real and structural. Both a brand and an app are subscription-shaped businesses whether or not they sell a subscription, because both depend on customers coming back rather than buying once. Both spend cash up front to acquire and earn it back over time. Both are priced at exit on the durability of that relationship. Once you see the shared skeleton, the specific numbers stop looking like two unrelated dashboards and start looking like the same business logic wearing two costumes.

Both businesses live or die on retention. A brand with a great first-order margin and no repeat purchases is the same story as an app with a great trial conversion and 8% monthly churn: a leaky bucket you keep refilling with paid acquisition. In both, the fix is not more top-of-funnel, it is keeping the customers you already paid for. That is why churn and returns sit in the same row above.

And both are read the same way at exit. When a buyer values a brand or an app, they are really pricing retention and how dependent the thing is on the founder. A concentrated, founder-run business trades at a discount on either side of the ecosystem. Durable retention and a team that runs without you is what moves the multiple, whether the asset is a brand or an app.

Why I keep both in one head

Most advisors specialize, and there is a reason: the surface details differ. A brand worries about 3PL rates and Meta CPMs, an app worries about API limits and App Store reviews. But under the surface, the operating questions are identical, and having lived on both sides changes how I read either one. When I look at a brand, I see the software vendors taking a slice of its GMV and I know what their businesses look like from the inside. When I look at an app, I see the merchant profit and loss its pricing lands on, and I know which line it is squeezing. That two-way visibility is the reason this report exists in one piece instead of two, and it is the lens I bring to the actual work.

If you track one number per side

If you forced me to pick a single leading indicator for each, I would take net revenue retention for an app and repeat rate for a brand. Both are the earliest honest signal that you have built something people keep choosing. Almost every other number, payback, valuation, EBITDA, is downstream of that. A brand with a rising repeat rate can survive a mediocre acquisition quarter. An app with net revenue retention over 100% can survive a churn spike. The reverse is not true: no amount of top-of-funnel brilliance saves a business people leave. When I have limited time to read a company, retention is the first place I look and the last thing I would trade away.

How the numbers move as you scale

One more thing the two sides share: the healthy line moves as you grow. A benchmark that is right for a $2M brand is wrong for a $50M one, and the same is true for an app at $200K ARR versus $5M. Reading yourself against the wrong stage is one of the most common mistakes I see, and it cuts both ways, a scaling brand judging itself by early-stage margins, or an early app holding itself to scale-stage retention it has not earned yet. Here is roughly how the targets shift.

The pattern is identical on both sides. Early, the game is proving the unit works at all. In the middle, it is holding the unit together while you pour fuel on it. At scale, it is durability, the boring compounding that makes a business worth buying. Match your benchmarks to your stage, and re-check them every time you cross a zero.

Don't manage
to the median.

The wrong way to use a benchmark table is to score yourself green across the board and feel good. Averages are a low bar, and being average is not a strategy. The right way is to find the one line where you are furthest below healthy, because that is almost always your binding constraint, the thing that, once fixed, unlocks the most.

A brand at a median 1.6% conversion with strong margin and retention does not have a margin problem, it has a conversion problem, and the effort should go there until it moves. An app with great free-to-paid and 7% churn does not have an acquisition problem, it has a retention problem. Benchmarks are a diagnostic, not a report card. Use them to pick the one fight worth having this quarter.

The reason the single-constraint approach works is that businesses do not improve on every axis at once, and pretending otherwise spreads your effort so thin that nothing moves. Pick one line, put real weight behind it for a quarter, and you will move it meaningfully. Try to nudge all seven and you will move none of them enough to matter. Focus is not a nice-to-have in operating, it is the mechanism by which anything actually changes.

A worked read: a $6M apparel brand

Here is the scan in practice. Say a $6M apparel brand comes to me with a conversion rate of 1.9%, gross margin of 62%, a return rate of 24%, CAC payback of nine months, and 3% EBITDA. Run it against the brand table. Conversion is just above median, fine. Gross margin looks healthy. But the return rate is high even for apparel, the payback is into the danger zone, and EBITDA is thin. The temptation is to go work on conversion, because that is the number everyone knows how to move and the wins feel good. That would be the wrong quarter, and it is the quarter most brands would pick.

The binding constraint here is the interaction of returns and payback. A 24% return rate on a nine-month payback means the brand is financing a lot of orders that come back, then paying to acquire many of those customers a second time to make up the lost revenue. The highest-leverage move is not more traffic, it is cutting the return rate through better sizing guidance and product-page fit content. Do that and effective margin rises, payback shortens because fewer orders reverse, and the saved cost flows straight to EBITDA. One root cause, three metrics fixed. That is what finding the binding constraint buys you, and it is invisible if you only look at the number that is easiest to move.

A worked read: a $1.2M ARR app

Now an app. A $1.2M ARR Shopify app arrives with a 32% billed-trial conversion, 5.5% monthly churn, 96% net revenue retention, and a 14-month CAC payback. Trial conversion is genuinely good, so the founder is proud of it and wants to pour budget into more trials. But read the other three together: churn is at the top of the normal band, net revenue retention is under 100%, and payback is stretched. Those three say the same thing in three ways. Merchants convert but do not stick, and the ones who stay are not expanding. Acquisition is not the problem. The bucket is leaking.

So the quarter goes to retention and expansion, not to more trials. Get first-90-day churn down by closing the activation gap that is losing merchants before they see value, and build one real expansion path so net revenue retention crosses 100%. Do both and payback fixes itself, because the same acquired merchant is now worth more over a longer life, with no extra acquisition spend. Pour budget on top of a leaky bucket and you just spend faster to stand still. Fix the bucket first. It is the same lesson as the apparel brand, told in a different language, which is the whole point of reading both tables side by side.

A worked read: a freemium app at $400K ARR

One more, because freemium reads differently. A $400K ARR app on a no-card freemium model shows a 4.2% free-to-paid, 4% monthly churn, 85% net revenue retention, and a huge free base. On paper only one number looks bad. But put them together and the story is monetization depth, not acquisition or even retention of paying users. The app is very good at getting merchants in the door for free and mediocre at turning that goodwill into expanding revenue, which is exactly what sub-90% net revenue retention on a big free base tells you.

So the quarter is not more free signups, the funnel is already full. It is the paid tiers and the expansion path: better reasons to upgrade, usage-based triggers, and packaging that grows with the merchant. Freemium businesses live or die on the depth of monetization behind the free tier, not the width of the free tier itself. Read against the wrong benchmark, this founder spends the quarter driving more free installs and wonders why revenue does not move. Read against the right one, they spend it building the ladder up from free, and it does.

When two benchmarks disagree

Sometimes the table sends mixed signals, and that is information, not noise. A brand with best-in-class conversion and thin margin is telling you it is buying conversion with discounts. An app with strong free-to-paid and weak retention is telling you the trial oversells what the product delivers. When a top-of-funnel number looks great and the downstream number looks bad, trust the downstream number, because it is measuring what actually happened after the promise was made. The healthiest businesses are not the ones with the best single metric. They are the ones with no glaring weak line, because a chain breaks at its weakest link and growth only exposes the break faster.

The mistakes I see most

Across both sides of the ecosystem, the same benchmarking mistakes come up again and again. Avoid these and you are already ahead of most operators.

• Scoring yourself green and stopping. Being average across the board is not a win, it is a to-do list you have not read yet. Hunt for the red, not the green.
• Benchmarking against the wrong stage. A $40M brand judging itself by early-stage margins, or a $300K app holding itself to scale-stage retention it has not earned. Match the line to your size before you judge it.
• Comparing to the global median instead of your vertical. Conversion, add-to-cart, and returns all swing two or three times across categories. The blended average is almost never your average.
• Trusting a ratio over a cash fact. A great LTV:CAC that takes eighteen months to realize can still put you out of business. Payback in months is the harder and truer number.
• Ignoring the invisible metrics. Cash conversion cycle for a brand, net revenue retention for an app. Neither shows up on the headline profit and loss, and both decide whether the business survives its own growth.
• Fixing the visible thing instead of the binding thing. Checkout is easy to work on, so brands optimize checkout when the leak is upstream at add-to-cart. Solve the constraint, not the symptom that happens to be convenient.

Turn it into a quarterly rhythm

Benchmarks earn their keep when they become a habit rather than a one-time audit. The rhythm I recommend is simple. Once a quarter, pull your own numbers for every line on your side of the table and mark each one green, watch, or risk. Pick the single reddest line and make it the quarter's focus, with one owner and one number to move. At the end of the quarter, re-run the whole scan. The binding constraint will usually have moved, because fixing one line changes the others, and the next quarter writes itself.

Do that four times and you have run the business by its constraints for a full year, which is how the brands and apps that compound actually operate. It is unglamorous and it works. The DTC Growth Scorecard automates the read for a brand, and the full calculator set covers the individual lines for both brands and apps, so the quarterly scan takes an afternoon, not a week.

One caution: benchmarks age. The conversion, churn, and payback medians in this report reflect 2026, and some are drifting. CAC payback in particular has stretched across the market, so a number that reads "below average" today might have been healthy two years ago. Treat the tables as a snapshot, re-check the sourced ones once a year, and weigh the direction of travel, not just the level.

Where these
numbers come
from.

Two sourcing rules, stated up front. The public-median figures are drawn from the research below, and I link each one at the point it appears in the tables. The operator figures come from the brands and apps I have run and advised directly: the brand portfolio I co-founded and scaled into nine-figure revenue, the software business I founded and sold, and the DTC and app clients I work with now. Where a number is operator-derived, I label it "operator" so you can weigh it as experience rather than a survey.

The public sources, and what each one informs:

Two honest limits on what you just read. First, published medians are averages of enormous, uneven samples, so treat them as a center of gravity rather than a precise line, and always weight your own category over the global figure. Second, the operator numbers are drawn from the brands and apps in my field of view, which is a real and hard-won sample but a sample nonetheless, weighted toward the DTC and Shopify-ecosystem businesses I work with rather than every corner of commerce. I would rather say that outright than pretend a range is a law. Where the two kinds of numbers agree, you can lean on them hard. Where they diverge, treat it as a prompt to dig into your own data, which is the only benchmark that ever truly matters.

This is also why the report will keep changing. As the sourced medians refresh and as I see more businesses up close, the bands will move, and I will update them here rather than leave a stale number standing. If a figure does not match what you are seeing in your own operation, that is worth a conversation, not a shrug.

The bottom line

Strip all of it down and the report says three things. Know your numbers, because you cannot manage what you have not measured, and most operators are managing to a headline figure that hides the truth. Read them against the right line, your stage and your vertical, not the global median, and against the honest bar of healthy rather than the low bar of average. And spend your effort on the one constraint that is furthest from healthy, because that is where the leverage is, on a brand or an app, at any stage. Everything else here is detail in service of those three moves.

That is the whole reference. Keep it next to you when you plan the quarter, and use it to argue with your own numbers, not to feel average. If you want a second read on where your brand or app really sits, and what to do about the one line that is furthest from healthy, the DTC benchmark card and the free calculators are the fastest way in, and my inbox is open.

Questions I get
about these
benchmarks.

Free tools: Run your own numbers against these bands. Brands: DTC profitability and max allowable CAC. Apps: CAC payback and app valuation.