Why do shoppers drop off after clicking a Meta ad?
Short answer
Often because the page makes them work. On one natural hair colour account, ads sent shoppers to a collection page, so comparing shades and pack sizes meant opening product pages one at a time. After ads moved to a single product page showing every shade and pack, the share of visitors adding to cart rose from 10.7% to 15.9%.
The brand sells natural hair colour in several shades and in packs of two or three. Its Meta ads sent people to a collection page. To buy, a shopper had to open a product page for one shade, go back, open another to compare, and then choose a pack size. Every extra page is a point where someone on a phone gives up.
Meta’s numbers already said the ads were doing their job: people clicked. The drop happened after the click, so buying more traffic would only have bought more drop-off. We studied how customers moved through the site and how competitors laid out shade selection, and the fix was to stop sending ads to the collection page at all.
Before: ads to a collection page
1. Ad
2. Collection page
3. Product page for one shade
4. Back to compare other shades
5. Choose a pack size
6. Standard checkout
After: ads to one product page
1. Ad
2. One page with every shade and Buy 2 or Buy 3 packs
3. One-click checkout
The new page showed every colour in one place and let shoppers mix shades inside a Buy 2 or Buy 3 pack. Calls to action became clearer and more prominent, the layout was simplified, and the checkout moved from Shopify’s default to GoKwik, an Indian one-click checkout.
How do you find where a landing page loses buyers?
Short answer
Split conversion rate into the stages between the click and the sale: page load, add to cart, checkout started and purchase. Each stage points at a different part of the site. On this account, add to cart gained the most, up by about half, while the share of started checkouts that finished fell from 44.6% (May) to 36.9% (1–27 July).
May against July 2026, Meta Ads Manager account totals
* From the 1–27 July campaign view, the only July view that reports these two stages. All other figures are full months.
The biggest single gain is the page itself: the share of visitors adding to cart rose by about half once they could see every shade and pack on one screen. More clicks also turned into page views, which coincided with the speed fixes (more on that below). More carts moved into checkout after the switch to one-click checkout.
The last line is the one to watch. A smaller share of started checkouts finished, down from 44.6% in May to 36.9% over 1–27 July. The data does not show why; easier checkout entry may be pulling in less committed shoppers. Either way, it is now the weakest stage in the funnel and the next thing to work on.
Four stages, one owner each
What is the Adbuffs Post-Click Ledger?
The Adbuffs Post-Click Ledger is how we read every account’s funnel after the click: four stages, each with an owner, checked every month. Conversion rate is the four stages multiplied together, so a gain in one stage and a loss in another can hide each other in the total.
Page
Page view → add to cart
Cart
Add to cart → checkout started
Checkout
Checkout started → purchase
Find the weakest stage, fix that one, and read the ledger again next month.
How does landing page CRO offset rising Meta CPMs?
Short answer
By getting more sales from every click you already pay for. From May to July 2026, cost per 1,000 impressions rose 45% and cost per click 65%, while conversion rate rose 62%. ROAS went from 2.11 to 2.38 instead of falling. At May’s conversion rate, July’s ROAS would have been about 1.5.
Costs against conversion, May to July 2026
ROAS moved only from 2.11 to 2.38 because the auction got more expensive at the same time, and most of the conversion gain went into paying for dearer clicks. Run the arithmetic the other way: if July’s clicks had converted at May’s rate, at July’s prices and July’s order value, ROAS would have been about 1.5 instead of 2.38. That is the gap that closed while the page, checkout and speed changes went live.
It is also why we judge growth on blended efficiency rather than one platform’s ROAS, as we explain in The one number that tells you if you’re actually growing.
Cost per click and conversion rate together would predict July’s ROAS slightly below May’s. It rose about 13% instead, which suggests larger orders. That fits a page built around Buy 2 and Buy 3 packs, though the data here does not show order value directly.
Can a slower landing page cancel out conversion gains?
Yes. The first version of the new page had speed problems on some days: slower loads and Core Web Vitals issues, Google’s measures of how fast and stable a page feels. A page that converts better but loads slower can give back everything it gained, because people leave before they see it.
We sent the brand’s team continuous feedback on image sizes, load speed and Core Web Vitals until the page kept its features and loaded properly. Over the same period, the share of clicks that became page views rose from 69.5% to 75.3%: about six more people in every hundred who clicked saw the page.
Did the landing page alone cause the conversion gain?
The data cannot tell. It does not split the gain between the page, the checkout and the speed work, because all three changed together. Spend dropped in June and creative testing continued throughout, and both affect who reaches the page. Conversion figures are Meta-reported. The case study quotes conversion rate as 2.4% to 4.06% (+69%): 2.4% is May rounded and 4.06% comes from the 1–27 July view. We use the full-month figures, 2.47% to 4.01% (+62%), so every month comes from one report. A controlled version would split traffic between the old and new page over the same dates.
What founders ask us about landing pages