Landing pages: where paid media is actually won or lost.

Consider two improvements. The first is a 10% reduction in CPC on one campaign, won through bid and structure work. The second is a 10% improvement in landing page conversion rate. The first helps one campaign on one channel until the auction shifts again. The second makes every click worth more: every campaign, every channel, paid social and paid search alike, today and next quarter. They are not the same size of win, yet most teams spend almost all their time chasing the first kind.

The reason is ownership. The ad account belongs to the marketing team; the website belongs to someone else. So optimization stops at the click, and the most expensive real estate in the funnel, the seconds after a paid visitor lands, goes unmanaged. Treating the landing page as part of the media buy rather than an adjacent department is one of the clearest differences between teams that scale efficiently and teams that plateau.

The arithmetic: why the page wins

Cost per acquisition is cost per click divided by conversion rate. That one line explains the whole argument. A cost per click gain changes the numerator for the campaign where you won it. A conversion rate gain changes the denominator for every visit that reaches the page.

Two details are worth noticing. Within a single campaign, the two gains are almost the same size: a 10% cheaper click lowers cost per acquisition by 10%, and a 10% higher conversion rate lowers it by about 9.1%. The difference in the table comes entirely from reach. The page gain touches every channel; the bid gain touches one campaign. And the two compound, which is why the order of work matters more than which one you prefer.

The same arithmetic runs in reverse. A page change that quietly drops conversion rate by 10% raises cost per acquisition on every channel at once, and nothing inside the ad account will explain it. That is why a site release belongs in the first check of any performance drop diagnostic. To see what a given cost per acquisition does to profit, run it through the ROAS and profit calculator.

Message match: keep the promise the ad made

The single most common conversion killer is a broken promise. The ad says one thing; the page says something more generic. Someone clicks an ad for waterproof trail running shoes and lands on a category page for all footwear. The scent goes cold, and the visitor concludes within seconds that they are in the wrong place.

• →The headline on the page should restate the specific promise of the ad that was clicked
• →The offer in the ad must be immediately visible on the page, not three scrolls down
• →Ad groups with distinct intents deserve distinct pages, not one generic destination
• →If you run the ad in five languages or five cities, the page should know which one the visitor came from

What changed in 2026: the platform may choose the page

Message match used to be something you controlled by choosing a final URL. That is no longer fully true. Google describes final URL expansion in Performance Max as letting it match queries to relevant landing pages on your site and generate dynamic assets, with the advertiser still responsible for the accuracy of that page content. AI Max for Shopping, in beta, extends the same idea to Shopping, pairing final URL expansion with rewritten product titles and dynamic landing pages.

The practical consequence: any page Google can crawl is now a potential paid landing page. A thin collection page, an out-of-date promotion, or a blog post with no path to purchase can receive paid traffic you did not plan for. Three habits follow.

1. Review the landing page report for every campaign with expansion on, and look for pages you never briefed
2. Use URL exclusions for pages that should never take paid traffic, such as policy pages, careers, and expired offers
3. Raise the floor on every template rather than polishing one hero page, because the platform may route around the hero

Whether to turn expansion on at all is a separate decision, covered in AI Max for Search. If it is on, the landing page work described here applies to your whole site, not a single page.

Speed is a bidding decision

Every second of load time sheds paid visitors you already paid for. Google also names landing page experience as one of the three components of Quality Score in Search, alongside expected click-through rate and ad relevance, so a poor page can feed back into what you pay per click. We treat page speed as part of the media plan: set a speed budget, test on a mid-range phone over a mediocre connection because that is closer to the typical paid click than your office connection, and challenge anything that blows the budget, however pretty it is.

Use the standard thresholds rather than inventing your own. Google’s Core Web Vitals define a good experience as Largest Contentful Paint within 2.5 seconds, Interaction to Next Paint of 200 milliseconds or less, and Cumulative Layout Shift of 0.1 or less, measured at the 75th percentile of page loads and split by mobile and desktop. Interaction to Next Paint replaced First Input Delay as the responsiveness metric in March 2024, so older audits that report First Input Delay are measuring the wrong thing.

Shopify stores do not need a separate tool to start. The web performance reports in the Shopify admin show all three metrics at the 75th percentile, over time and by device type. Check them after every theme change and every new app, because apps and scripts are the usual source of a sudden slowdown.

One page, one job

Pages fail when they try to serve everyone. A landing page for a paid campaign has one job: move this specific visitor, who clicked this specific promise, one step forward. That means one primary call to action, not four competing ones. It means cutting the global navigation when the traffic is bought for a single purpose. It means the form asks for what the next step needs, not everything sales would like to know eventually. Every additional option on the page is a tax on the decision you actually want.

How we brief a landing page

The discipline that makes this repeatable is the brief. Before design starts, we write down: who arrives on this page and from which ad, what they were promised, the one action we want, the objections that stand in the way, and the proof that answers each objection. The page then gets built as an argument: promise restated, objections handled in order, proof next to claim, one clear ask. Design serves that argument. When a page underperforms, the diagnosis starts with the argument, not the aesthetics.

How to know a page is leaking

Look at paid traffic by landing page, not by campaign. A page that takes meaningful paid sessions and produces no purchases over a period long enough to expect several is the clearest leak there is, and it hides easily inside a campaign whose average looks acceptable. We monitor for exactly this pattern, paid landing pages that take traffic and convert nobody, because it is invisible in the ad platforms’ own campaign views.

Beyond that, compare conversion rate by page for the same traffic source, and compare mobile with desktop for the same page. A large mobile gap on a page with poor Core Web Vitals is usually a speed problem. A large gap between two pages fed by the same campaign is usually a message match problem.

The order of operations matters. Fix the page and every ad account improvement you make afterward is amplified. Fix the account first and you are optimizing the delivery of traffic into a leaking container. Conversion rate is one of the few levers that raises the return on every dollar at once, which is why it sits among the core levers in how to scale an ecommerce business. Media buying and conversion rate optimization are not separate crafts. They are two halves of the same price of a customer.