SEM
Landing pages that convert Google Ads traffic
Google Ads landing pages convert when the query, the headline and the first screen tell the same offer, and when that screen is fast enough to tap. I am a Gold Coast web developer with 18 years of storefront work. I have watched spend hit homepages that were never written for the keyword, and I have watched beautiful PDPs lose because LCP sat on an unoptimised hero. In 2026 Smart Bidding will still send you the click. It will not rewrite your H1.
Google's landing page experience guidance still talks about relevance, transparency and ease of navigation. That language is older than Core Web Vitals and still right. My production version is: match the query, keep the promise, make add-to-cart the obvious next tap, and do not load a chat widget before the product.
This note is for developers and merchandisers who share a URL with a media buyer. If you only change bids, you are bidding on a page you have not read on a phone.
Why Google Ads landing pages still lose the click
Google Ads landing pages still lose the click when the ad promised a SKU and the URL dumped the shopper on a brand story. Performance Max is aggressive about URL expansion. Search is only as honest as the final URL you typed. Either way, the person arrived with a sentence in their head. If the H1 is a mood and the product is below a film, they bounce. Quality Score and conversion rate both notice.
The second loss is speed. LCP on category and product pages is usually the hero image. If Ads traffic is paid, a 4-second hero is you paying for a blank. I keep Core Web Vitals for online stores next to this piece because INP on variant pickers is how you lose the add-to-cart after a successful landing.
The third loss is a different offer. Ad says 20 percent off. Page shows full price until a popup. Ad says "in stock." Page says sold out. Ad says "CourtStrike 2.0." Page is a generic shoes collection with no way to see construction. You trained Google that this URL converts. It converts until people learn not to trust the brand.
The fourth loss is mobile layout that was designed at 1440px. Sticky bars, CMP overlays, and a header that eats half the viewport push the product below the fold. Paid traffic is impatient. Organic traffic is impatient too, but you paid for this impatience.
I do not treat a homepage as a default landing page for generic queries. Homepages exist to route. Ads exist to land. If the query is a product, land on the PDP. If the query is a category, land on a collection that is already filtered to that intent, not on a search results page with a query string the theme barely supports.
How Google Ads landing pages should match the query
Google Ads landing pages should match the query by repeating the words and the constraint the shopper used. If they typed a size, a colour, a model year or a dietary claim, that fact should be visible without a tap. Headlines in the ad and the H1 should be cousins, not strangers.
For Search, I write the final URL first, then the ad. If I cannot find a URL that deserves the keyword, I do not buy the keyword yet. That sounds slow. It is cheaper than a month of wasted CPC.
For Shopping and PMax, the feed link is the landing page. Feed quality and landing quality are one system. See Google Shopping feed quality for product ads and Performance Max vs Search campaigns for stores. A perfect title in Merchant Center that lands on a 302 chain through an old Magento layered URL will still feel broken.
On product pages I want, above the fold on a phone: product name, price, primary image that is the LCP element with width and fetchpriority, variant selection, add to cart, and the one proof that matches the query (allergen line, material, warranty, exploded construction). Reviews can wait a scroll if they are not the reason for the click. Chat can wait until idle.
On collections I want a real H1, a short intro that is not keyword soup, filters that work without reloading the world, and product cards with price and a readable image. Infinite scroll that breaks the back button is how paid users lose their place.
Campaign-specific landing pages are allowed when the offer is real and temporary. They are not allowed as thin copies of the PDP with a different H1 and no way to check out. Duplicate SKUs with parameterised URLs also split measurement. Use one canonical PDP and change the module above the fold if you must message a promo.
I built an interactive PDP for Selkirk CourtStrike 2.0 so the construction story and the buy path could share a page. That is the standard I hold ads to: if the query is about how the shoe is built, show the build, then take the size. If the query is a colourway, do not hide variants behind a story module.
Partake Foods needed a storefront that matched a new pack system: colour, lineup filters, PDPs that still shop. Paid traffic to a default theme after a packaging relaunch is how you pay to unteach the new identity. The landing page is part of the brand system, not a leftover template.
Production checklist for query, LCP and offer
A URL is fit for Google Ads when you can complete this list on a mid-range Android phone, not on your desktop with cache disabled in a way nobody else uses.
- Query-to-H1 match. Read the search term. Read the H1. They should share the product or category language.
- Offer match. Promo, stock and price match the ad and the feed.
- Primary CTA visible without scrolling on a phone. Add to cart or the real next step, not "learn more" to a journal.
- LCP element is the product image or a short heading, not a cookie banner or a carousel of three megabyte slides.
- Width, srcset, fetchpriority on the hero. No lazy-load on the LCP image.
- INP: variant clicks and filter clicks respond. No 300ms of tag soup before the handler.
- CMP does not cover the CTA. Consent can be required. It cannot steal the first tap forever.
- No URL expansion onto cart, account, search-internal, or out-of-date campaign pages.
- 200 status, HTTPS, same product as the ad. Redirects at most one hop.
- Conversion tag fires on the thank-you that this journey actually uses, including shop-pay and wallet returns.
I also print the page with JavaScript disabled once. If the product name and price disappear, you have an SEO problem and an ads problem.
What breaks after a theme or campaign launch
A theme launch that changes URLs without mapping old final URLs in Google Ads will send paid traffic into redirects or 404s. PMax remembers URLs. Update asset groups.
A hero video autoplay will destroy LCP and data plans. Ads users on 4G are still users.
A merchandiser swaps the collection template to a lookbook with no prices. Instagram likes it. CPC traffic does not.
App pixels and reviews widgets shift INP on the add-to-cart. The landing looks "rich." The tap feels drunk. I defer third-party tags on landing templates the same way I do on checkout, with more room for reviews if they are the proof the query asked for.
Personalisation that shows different prices to Ads bots versus shoppers is how you get policy problems and angry customers. One price.
A/B tools that hide the H1 until a variant loads will flash a blank to Google's crawler and to humans. Ship the default content in HTML.
How to measure landing pages that take Ads traffic
Measure conversion rate and bounce or engagement by landing page path in GA4, filtered to paid. Then measure Ads conversion value by final URL or by asset group. Then measure LCP and INP for those URLs in CrUX or in your RUM, not only in Lighthouse on a cable connection.
If a page has strong Ads CTR and weak conversion, read the first screen. If a page has weak CTR, the ad and the sitelinks may be the issue. If both are weak, you bought the wrong query.
Do not average all landing pages. Homepages, collections and PDPs are different jobs. A homepage conversion rate will always look worse. That is not a reason to land generic product queries there.
On Their Nibs the published lift was 31 percent conversions and 48 percent orders after a Shopify rebuild that put the purchase journey first. That is landing-page work in the broad sense: navigation, merchandising, theme. I will not pretend a headline tweak on an old theme would have equalled it. On Selkirk and Partake the proof is live pages that still shop while they tell the product story. Use them as the bar for what an ads URL should be able to do.
Related work on this site
Selkirk CourtStrike 2.0 is an interactive Shopify product page: exploded construction, colourways, and a path to size and cart. That is query-matched landing behaviour for a launch. Partake Foods is a custom Shopify storefront after a packaging redesign, with lineup filters and PDPs that still buy. For campaign structure that dumps traffic onto the wrong URL, read Performance Max vs Search. For the speed floor, read Core Web Vitals for online stores.
FAQ
Should Google Ads landing pages be separate from organic PDPs?
Only when the offer is truly different and the extra page can still check out. Most stores should improve the real PDP or collection. A microsite that cannot share cart state with the store is how you lose mobile users.
Do Google Ads landing pages need unique templates per keyword?
No. They need the right URL per intent. One PDP can serve many queries if the first screen shows the matching fact. Unique templates per keyword do not scale and they rot.
How fast should Google Ads landing pages be?
Fast enough that LCP is not the reason they leave. I use the Core Web Vitals thresholds as the floor, then I watch field data for the paid landing URLs. A lab-only 100 on an empty template does not count.
Can I use the homepage for Google Ads landing pages?
For brand queries, sometimes. For product and category queries, rarely. Homepages negotiate. Ads clicks already chose a sentence. Meet them at the product or the collection that sentence names.
Keep reading
Contact if you want this kind of work on a live store.