Across 216 rated ad and landing page pairs captured in 2026, 55 keep the ad's promise intact, 110 weaken it and 51 lose it outright. The 161 failures fall into six recurring patterns, and the most common one is not the most expensive one.
- The page drops the argument that earned the click is the most common failure at 43 pairs, but 37 of those stay Partial: the product is still buyable, so the ad limps rather than dies.
- The ad sold one thing, the page shows the whole shop is second at 38 pairs and owns 24 of the 51 outright breaks, nearly half. Routing, not copy, is where the money actually dies.
- A popup is the landing page on 36 pairs, of which 32 are recoverable by dismissing the overlay.
- Every ad in the library built for a subscription objective fails the handoff. 17 of 17, with no Continuous rating anywhere in that objective.
- The rarest failure is the worst: 8 pairs where the handoff resolves perfectly and then the item is sold out, and 6 of those 8 are outright breaks.
Source: The Handoff, 243 published teardowns, 216 rated pairs, captured February to August 2026
What happens to an ad's promise at the click?
In 2026, across 216 ad and landing page pairs captured together and rated, 55 keep the ad's promise intact, 110 weaken it and 51 lose it outright (The Handoff, 243 published teardowns, August 2026). That is a 75% failure rate on the single handover that every paid dollar has to survive.
This is not a survey and it is not scraped. Every pair in it was captured first-hand: the creative as it ran in feed, and the page it actually clicked through to, opened signed out from a US IP. Both artefacts are dated. The rating reads the first view of that destination before anything is dismissed, which matters more than it sounds, because a page you can only judge after closing two overlays is not the page the visitor got.
Three ratings. Continuous means the promise arrives intact: same product, same offer, same register. Partial means it arrives, but the visitor has to reconcile something the ad did not warn them about. Break means it does not arrive at all. Where exactly that line falls, including the two cases that decide most verdicts, is written down rather than left to judgement.
What are the six ways a handoff breaks?
The 161 Partial and Break pairs sort into six patterns, and each pair is filed under the one that costs it the click. The largest, at 43 pairs, is a page that keeps the product and drops the argument. The one that produces the most damage is second by count and first by breaks.
| Pattern | Pairs | Partial | Break |
|---|---|---|---|
The page drops the argument that earned the click | 43 | 37 | 6 |
The ad sold one thing, the page shows the whole shop | 38 | 14 | 24 |
A popup is the landing page | 36 | 32 | 4 |
The offer on the page is not the offer in the ad | 23 | 15 | 8 |
The right page opens on the wrong version | 13 | 10 | 3 |
They clicked to buy and there was nothing to buy | 8 | 2 | 6 |
Read the Break column rather than the total. The most frequent pattern is 86% Partial, which means the shopper can still buy the thing; the ad is simply working harder than it needed to. The second pattern owns 24 of the library's 51 outright breaks on its own. That is the difference between a leak and a hole.
The failure that happens most often is not the failure that costs the most. Tightening your copy match optimises the cheap one.
Why does routing cost more than copy?
38 pairs send a single-product ad to a homepage, category or collection, and 24 of them are outright breaks. It is the only pattern where spending more makes it worse, because every extra impression buys another visitor a page that does not contain the thing they clicked for.
The tell is blunt. Bataleon ran a boot and the destination was a sale banner: "Not one boot appears anywhere on the page, and the hero copy on the sale banner talks about grabbing your new favourite snowboard gear rather than the boot the ad zoomed in on." Ridge advertised a chain and landed on a page headed "MEN'S RINGS & WEDDING BANDS" whose 28 results are rings. Zesty Paws put its advertised probiotic powder at $33.97 on screen as the fourteenth tile, five rows down.
None of those flags in reporting. The destination is on-brand, it loads, it is shoppable, and the campaign keeps spending. That is what makes routing the expensive one: it fails silently in exactly the systems built to catch failure.
For every creative that names or shows one product, set the destination to that product's own URL. If it genuinely has to be a collection, use a pre-filtered or hand-sorted one where the advertised item is the first tile, not the fourteenth.
This is a link change, not a redesign, and it is the highest-value hour in this entire piece.
What happens when the page keeps the product and drops the argument?
43 pairs land the visitor on the right product with the reason for the click missing. The claim, the audience, the occasion or the comparison that did the work in the creative appears nowhere on the destination, so the page has to re-sell someone it should only have had to close.
Cozy Earth ran a menopause text thread and the page title is the single word "Sheets", with nothing above the 18 items saying cooling, temperature, night sweats or menopause. Hydrant built an entire fabricated Reddit thread to reach GLP-1 users, and that audience is dropped at the click, which is what turns it from a soft landing into a Break. Jones Road sells eye shape in the ad; the page sells glide, staying power and water resistance.
The mechanism is organisational rather than creative. The ad is written by whoever owns paid social, in the week of the campaign. The product page was written once, by someone else, for everyone. Nothing in either workflow requires them to agree.
Take the single strongest line in the creative and put those words, verbatim, above the fold on the destination, as an eyebrow, a badge or the first bullet, before you launch the ad.
Verbatim matters. A paraphrase reads as a different claim to a visitor who has just held the original in their head for four seconds.
Is a popup really a landing page problem?
On 36 pairs the destination is correct and covered: an email modal, a spin-to-win, a scratch card, a quiz, a cookie bar or an auto-opened cart drawer sits over the thing the ad promised at first paint. 32 of the 36 are Partial, because dismissing the overlay recovers the promise.
Parachute is the ordinary version: "The title, the price, the review count and the size selector all sit behind it." Craighill is the loud one, a full-screen YOU'VE GOT A MYSTERY OFFER panel over the hero image, the price and the buy button. SKIMS is the version that stops being recoverable, where the four brushes that are the entire ad have no counterpart on screen at all behind a NEVER MISS A DROP capture.
Every one of those overlays was added by someone optimising a different number: list growth, consent compliance, cart recovery. None of them were tested against a paid arrival. They stack because nothing suppresses them for traffic that has already converted on an ad.
Which kinds of campaign fail most often?
Failure rate tracks what the ad was built to do far more tightly than what it looks like. Every subscription ad in the library fails the handoff, 17 of 17, with no Continuous rating anywhere in that objective. Conversion ads fail at 72%, and the spread across ad formats is barely half the spread across destination types.
| Objective | Partial or Break | Rate |
|---|---|---|
Subscription | 17 of 17 | 100% |
Traffic / Clicks | 9 of 10 | 90% |
Conversion / Purchase | 133 of 184 | 72% |
Lead Gen | 2 of 2 | too few |
Awareness | 0 of 3 | too few |
The subscription number is the one worth stopping on. A perfect record of zero, on 17 attempts, is not variance. The recurring shape is a buy box that preselects a delivery cadence the ad never raised: OneSkin defaulting to "Delivered every 2 months" flagged Recommended, biom on AutoShip, Eight Sleep opening with Pay now selected while Rent monthly, the entire offer in the ad, sits beside it unclicked.
By category, Home and Household is worst at 84% (31 of 37), ahead of Footwear at 83% (10 of 12), Beauty and Skincare at 78% (21 of 27) and Supplements at 74% (23 of 31). Apparel is mid-pack at 70% despite carrying 63 of the 216 pairs, which is the largest sample in the set.
The store audit runs the same read on your own funnel: what the ad promises, what the page delivers, and what falls out in between.
What should you fix first?
Fix in the order the failures cost money, not the order they occur. Routing produces 24 of 51 breaks and is usually a link change. Overlays produce 32 Partials and are a suppression rule. The dropped argument is the biggest count and the slowest fix, because it is a copy job across every product page you advertise.
- Point every single-product creative at that product's own URL. If it must be a collection, pre-filter it so the advertised item is the first tile.
- Suppress welcome modals, games and cart drawers for paid arrivals. If consent law requires a bar, dock it so it cannot cover the buy box.
- Put the creative's strongest line, verbatim, above the fold on every destination you advertise into.
- Append the variant to the link so the advertised colour, size or count is preselected. 13 pairs open on the wrong version of the right page.
- Wire an inventory and URL health check to the ad set so a creative pointing at a sold-out or dead destination pauses itself. This is the rarest failure and the most severe: 6 of 8 are outright breaks.
One thing that is not on the list: raising the discount. Ads that carry an offer in the creative fail at 82.6% against 72.4% for ads that carry none, and depth does not rescue it. If the handoff is broken, a bigger number arriving at a page that contradicts it makes the contradiction louder.
If you want the underlying arithmetic on whether a discount can pay for itself at all, that is a separate and unforgiving calculation.
Common questions
What is a handoff between an ad and a landing page?
The handoff is the moment an ad passes a visitor to a page. It is rated on whether the page keeps the promise the ad made: the same product, the same offer, the same register. Across 216 rated pairs captured in 2026, only 55 keep it intact.
How common is poor message match in real DTC advertising?
In this library it is the norm rather than the exception. 161 of 216 rated pairs lose or weaken the promise at the click, a 75% failure rate, and 51 of them lose it outright. Every pair was captured first-hand rather than surveyed.
Should I send ad traffic to a product page or a collection?
Send single-product creative to that product's own URL. Sending it to a collection is the single most expensive pattern in the data: 38 pairs do it and 24 of those are outright breaks, because the advertised item is absent, buried or replaced at the top.
Do popups on a landing page actually hurt paid traffic?
They cost the rating as much of the promise as they hide. 36 pairs arrive to an overlay over the product, and 32 stay recoverable because dismissing it restores the page. A consent bar across the foot is rarely fatal. An email capture over the buy box frequently is.
Why do subscription ads fail the handoff so often?
Every one of the 17 subscription ads in this library fails, with no Continuous rating in the objective. The recurring cause is a buy box that preselects a delivery cadence the ad never mentioned, so the visitor arrives expecting one purchase and is offered a standing order.
Does your own handoff survive the click?
This is the same read I run on live funnels: the ad, the page, and the gap between them. Scored, free, no call required.
Audit my store free