The manual way comes first, in Google Ads itself. Then the same problem as Goldbeater finds it, every day, without anyone asking.
Landing page experience is one of Quality Score's three parts.
Google rates each keyword’s Quality Score from three parts: expected clickthrough rate, ad relevance and landing page experience. Landing page experience is, in Google’s words, how relevant and useful your landing page is to people who click your ad. Each part reads Above average, Average or Below average.
The rating is a comparison with the other advertisers whose ads showed for exactly the same search over the last 90 days. So Below average doesn’t say the page is broken. It says Google rates it less relevant and useful to that keyword’s searchers than theirs. It’s given keyword by keyword, so one page can rate well for one ad group’s keywords and below average for another’s.
Google’s glossary says what the experience is made of: how useful and relevant the information on the page is, how easy the page is to navigate, how many links are on it, and what the ad led people to expect. Google also says information from its crawlers may be used to assess ad quality.
Quality Score is a diagnostic and isn’t used in the auction. The ad quality it summarizes is: Google says it affects whether your ad can show, where it appears and what a click costs, and that higher quality ads typically cost less per click. The low Quality Score guide covers the other two parts.
Where Google Ads shows the rating, keyword by keyword
- Go to Campaigns › Audiences, keywords, and content › Search keywords.
- Select the columns icon above the table, then Modify columns. Open the Quality score section, add Landing page exp. and Quality Score, and select Apply.
- Set the date range to the last 30 days and select the Cost column’s name to sort by it. Read down for the keywords that read Below average, and note each one’s ad group.
- The page Google rated is the one that ad group’s ads send people to, their final URL, unless the keyword has a final URL of its own.
A dash in the Quality Score column means too few searches match the keyword exactly for Google to rate it. To see when a rating changed, add the column’s (hist.) version and segment the table by day. Google offers the Day segment only for a date range of 16 days or less, so over a longer range segment by Week. The Quality Score drop guide walks through reading it.
Then read the ratings page by page. Where most of the keywords landing on a page read Below average, the page is the thing to mend. Where one ad group’s keywords do and the other ad groups sharing the page rate well, the page answers their searches and not this ad group’s. Mending it for the odd one would cost the others, so give that ad group a page about what its keywords search for.
How to improve landing page experience, in Google's own advice
Google’s help page lists five things to try when Landing page exp. reads Below average or Average:
- Give people what they’re looking for. Google’s example: someone who searches for flannel shirts and clicks your ad for them should land on a page that features your flannel shirts.
- Keep the message consistent from ad to page. The page follows through on the ad’s offer or call to action. If you can’t change your site, Google says to try your existing pages and find the most useful one.
- Use conversion rate as a proxy. Google says it doesn’t change the rating, and that it’s a good way to measure a page and improve it.
- Make the site work on a phone. Google says people value easy navigation even more on mobile sites.
- Improve loading speed. In Google’s words, it can be the difference between someone bouncing or buying.
Google adds that the search’s exact words needn’t be on the page. Its page on optimizing landing pages gives three more: where an ad offers a free tour, feature the sign-up form prominently; put the important information toward the top, where a visitor sees it without scrolling; and don’t clutter the page with too many ads or pop-ups.
The rating tells you a page falls short, not how. The rest of this guide is how to measure it: what each page converts, how fast it loads, and whether its tag and its address are sound.
Compare your landing pages by what they convert
- Go to Campaigns › Insights and reports › Landing pages. Each row is a page your ads sent clicks to. Google says the page covers Search, Display, Video and Shopping campaigns, and for Search it lists your sitelinks’ pages beside your ads’.
- Set the date range to the last 30 days, and add the Conversions, Conv. rate and Cost / conv. columns.
- Filter the table to one campaign. Google says this page filters and segments as its other reports do, and the filter button sits above the table. A campaign’s pages share its goal and its bidding, so they can be set against each other. Pages of two campaigns can’t: their clicks come from different searches and are bid toward different goals.
- For each page, add up the campaign’s other pages: their conversions divided by their clicks, and their cost divided by their conversions.
- Multiply the page’s clicks by the others’ conversion rate, and divide its cost by what they paid for a conversion. Each is what the page would have brought had it done as well as they did. A page that brought under half of both is the one to open.
| Landing page | Clicks | Cost | Conversions | Conv. rate | Cost / conv. |
|---|---|---|---|---|---|
| Total | 1,180 | $2,016 | 21 | 1.78% | $96.00 |
| /sofas | 400 | $660 | 1 | 0.25% | $660.00 |
| /sofas/velvet | 335 | $636 | 5 | 1.49% | $127.20 |
| /sofas/leather | 215 | $366 | 8 | 3.72% | $45.75 |
| /sofas/corner | 130 | $201 | 4 | 3.08% | $50.25 |
| /sofas/linen | 100 | $153 | 3 | 3.00% | $51.00 |
| Sum | Working | Result |
|---|---|---|
| The other pages' conversion rate | 20 ÷ 780 | 2.56% |
| What they paid for a conversion | $1,356 ÷ 20 | $67.80 |
| Its clicks, at their rate | 400 × 2.56% | 10.3 conversions |
| Its spend, at their price | $660 ÷ $67.80 | 9.7 conversions |
| So few by luck | BINOMDIST(1, 21, 32.7%, TRUE) × 8 pages | 2.2% |
The last row asks whether so few could be luck. A page with a few dozen clicks can bring nothing by chance, so test it before you rebuild anything. In Google Sheets, BINOMDIST gives the chance that a page taking 32.7% of a campaign’s spend would bring 1 or fewer of its 21 conversions: 0.27%. Do the same with its share of the clicks and keep the larger answer. Then multiply by the number of pages you compared, since each one was another chance for luck to show. Cedar & Pine compared 8, this campaign’s five and three in Rugs — Broad, and 2.2% is still under 5%.
A page that takes clicks from several campaigns gets the same sum in each, and the expected conversions are added. View expanded landing pages, under a page’s link, lists the addresses people reached once keywords or custom parameters were written into it.
Then open the page on a phone, as someone clicking the ad would. Check that it loads, that it says what the ad said, and that it asks for the conversion the campaign counts. If the page is sound, look at the searches reaching it: /sofas is where Cedar & Pine’s catch-all ad group lands, and the keywords per ad group guide shows what that ad group holds. The ad group guide sets it against the campaign’s other four ad groups over 13 weeks. Google reports a conversion on the day of its click, not the day it happened, so the latest clicks may not have theirs yet; the not-converting guide covers how long to allow.
Measure page speed the way Google does
Google says that in retail it has seen a one-second delay on mobile affect mobile conversions by up to 20 percent. Its help pages describe three columns on the Landing pages page for finding pages that do badly on phones: Mobile speed score, from 1 to 10 with 10 the fastest, for Search; Mobile-friendly click rate; and Valid AMP click rate. Google needs a minimum amount of data to score a page, so a less visited page may show none.
For a page with no score, or to see what holds a page up, use Google’s PageSpeed Insights:
- Open pagespeed.web.dev and enter the landing page’s address.
- Read the real-user figures first. They are what Chrome’s visitors experienced over the previous 28 days, at the 75th percentile: a quarter of visits were slower than the figure shown.
- Largest Contentful Paint is how long the largest image, block of text or video in view takes to show. Google’s line for good is 2.5 seconds or less, and past 4 seconds is poor.
- Interaction to Next Paint is how long the page takes to respond to a click, a tap or a key press, at its slowest in a visit. Good is 200 milliseconds or less, and past 500 is poor.
- Where the page has too few visitors, PageSpeed Insights shows the figures for your whole site, and where the site has too few, none. Then read the lab test’s performance score, a page load simulated on a mid-tier phone over a mobile network: 90 or above is good, and below 50 is poor.
The real-user figures need a page that is public, which Google defines as crawlable and indexable. A landing page you keep out of search results has none of its own, so read the site’s. A fast lab score doesn’t promise fast visits: Google says good lab data doesn’t necessarily mean real users’ experiences are good too.
The Google tag belongs on every page a click lands on
The Google tag is one tag for your whole site, either the gtag.js snippet or a Tag Manager container, that sends what happens there to Google Ads. Google asks for it on each page. On a landing page it has a job of its own: the click’s details arrive in the landing page’s address, and the tag stores them in a first-party cookie on your domain, so a conversion later on is tied to the click.
So before you blame a page that seems to convert nothing, check its tag. Google Ads reports untagged pages in two places:
- Go to Tools › Data manager. Under Google tag, find your tag and select Manage.
- In the Tag quality section, select View diagnostics. Tag Diagnostics shows Some of your landing pages are not tagged when landing pages of your ads or your Performance Max campaigns have no tag.
- Open the Tag coverage summary for the list. Each address reads Tagged, having loaded your tag in the last 30 days; No recent activity, not in the last 30 days but before; or Not tagged, never.
- Check each Not tagged page in Tag Assistant, which reads the page as it loads. Google says a page with little traffic can be flagged by mistake.
The conversion tracking guide covers the conversion actions the tag reports to.
Check each Google Ads final URL: https, and a page that loads
The final URL is the address of the page people reach after they click your ad. Each ad has one, with a separate field for a mobile page, and a keyword or a sitelink can carry its own. Google’s policy is that the landing page shares its domain with the ad’s display URL, and that a final URL doesn’t redirect to another domain.
Start it with https. Google Ads recommends, and in some cases requires, that every URL uses https. Where your site redirects http to https, Google still strongly recommends changing the final URLs themselves: the visitor’s first request isn’t encrypted, and the redirect slows the page, which Google says affects the landing page experience. There are two cases where Google Ads forwards an http click straight to https for you: your site uses HSTS, or the address answers with a 301 redirect to the same address on https.
Chrome is the other reason. Google’s Chrome team announced on October 28, 2025 that Chrome 154, in October 2026, would turn on Always Use Secure Connections by default: Chrome tries https for every connection, and asks the visitor’s permission before the first visit to a public site that has none.
- Open your site in Chrome with https:// in place of http://. If the page doesn’t load, your site doesn’t support https, and Google says to ask your web developer or whoever runs your site’s platform to turn it on.
- If it loads, change your final URLs and mobile final URLs to https. For many at once Google names bulk uploads, Google Ads Editor and the API.
- Test the address where you set it: on Campaigns › Ads, select an ad’s pencil icon, open Ad URL options and select Test. The result reads Page found, Page not found, Final URL mismatch or Unreachable. Google notes the test doesn’t check its policies.
Google’s own review checks the page too. An ad whose page it couldn’t load is disapproved for Destination not working, one whose page isn’t where the ad says it goes for Destination mismatch, and one it can’t reach in the places you target for Destination not accessible. Hover over the reason in an ad’s Status column to read it. An ad you edit and save is reviewed again automatically. Where you mended the page and left the ad as it was, Google’s page on Destination not working says to file an appeal. The destination guide has the fix for each, and for pages that redirect.
What it looks like when Goldbeater finds it
Every day Goldbeater sets each landing page beside the other pages of every campaign sending it clicks, over the last 30 days. It raises a page that spent at least $20 where its clicks bought under half the conversions they would have at those pages’ rate, and under half what its spend would have bought at their cost per conversion. Both have to be unlikely by chance, at 5% across every page it tested. A page with clicks from several campaigns is judged in each and the results pooled, so it’s never set against a campaign with another goal.
https://example.com/sofas converted 0.3% of its clicks at $660 each, against 2.6% at $68 each on the other pages of the campaigns sending it clicks
Sofas — Phrase
+8.7 conversionsEstimated, a month
At what each campaign's other pages paid for a conversion, its $660 would have brought about 9.7; it brought 1.
The same campaigns' other pages, last 30 days
Its 400 clicks from 1 campaign cost $660 and brought 1 conversion where 9.7 were expected at what each campaign's other pages paid for a conversion, and 10 at their rate a click: under a 2.2% chance of so few by luck, allowing for the 8 pages tested across the account.
- Clicks
- 400
- Cost
- $660
- Conv.
- 1
- Conv. not counted
- 1
- Expected at its campaigns' other pages' cost / conv.
- 9.7
The last 30 days
| Figure | This page | Its campaigns' other pages |
|---|---|---|
| Cost per click | $1.65 | $1.74 |
| Cost per conversion, beyond chance | $660 | $68 |
| Conversion rate | 0.3% | 2.6% |
This page
Its campaigns' other pages
Each page's conversions and those its spend would have bought, last 30 days
| Member | Conversions expected, last 30 days | Conversions, last 30 days | Against the conversions expected |
|---|---|---|---|
| example.com/sofas | 9.7 | 1 | below |
| example.com/sofas/velvet | 7.4 | 5 | below |
| example.com/sofas/leather | 2.9 | 8 | above |
| example.com/sofas/corner | 1.9 | 4 | above |
| example.com/sofas/linen | 1.5 | 3 | above |
example.com/sofasConversions expected 9.7 · Conversions 1 · below the conversions expected
example.com/sofasThe othersThe conversions expected
| Sofas — Phrase | 400 | $660 | 1 | $1.65 | $1.74 | $660 | $67.80 | 9.7 |
A page whose clicks bring under half the conversions the same clicks bring on the other pages of the campaigns sending them there, and whose conversions cost more than theirs, too far under to be chance. Each campaign is its own measure, so a page is never judged against a campaign with another goal. The usual causes are a conversion tag missing on the page, a page that doesn't say what the ad did, or searches it wasn't built for; the estimate is the conversions its spend would have bought at what the other pages paid for one.
Open the page as someone clicking these ads would, on a phone: check that it loads, says what the ads say, and asks for the conversion the campaigns count, and that the conversion tag fires on it (a page on another domain, or a new one, often has none). If the page is sound, compare the searches reaching it with those reaching the campaigns' other pages: the page may be getting searches it wasn't built for.
The steps under What to do are the fix. Goldbeater drafts no change for this one.
This is “A landing page converting far below the other pages of its campaigns” on a sample account. Goldbeater runs it on yours every 24 hours, and your AI analyst answers what you ask about any finding.
Get a free first lookThat is the page worked above: the finding’s expected conversions and its chance are the table’s last three rows. Its estimate is in conversions, what the page’s spend would have bought at the other pages’ cost per conversion less what it bought. It also counts what the page’s clicks recorded that the campaign doesn’t bid on, here one add to cart, and draws the campaign’s pages on one chart. The fix is the page or what reaches it, so Goldbeater drafts no change.
Five more checks read your landing pages every day:
- A page that suits the ad groups sharing it, and not this one. An ad group where keywords rated below average on landing page experience carry more than half its rated spend, at least $20 in 30 days, on a page other ad groups also land on and mostly rate average or above on, where so many below average is unlikely by chance, at 5% across every ad group tested. Keywords naming a competitor are left out. Where an ad group of the same name lands on a page Google rates well, the finding names that page.
- Pages that load slowly on phones. Goldbeater asks PageSpeed Insights about the ten pages your ads send the most spend to. It raises a site, once, where real visitors’ 75th percentile is past Google’s line for good, 2.5 seconds for Largest Contentful Paint or 200 milliseconds for Interaction to Next Paint, or, where there are no real-visitor figures, the lab score is under 50.
- Pages with no Google tag. Goldbeater opens every page your ads, keywords, asset groups, sitelinks, promotions and price assets name, and any page Google chose itself that took a click in the latest week, and reads its HTML. It raises the pages with no Google tag in it on a site where another page has one. A tag loaded only after consent doesn’t show in the HTML, so a site with no tag on any page isn’t judged.
- Final URLs on http. An enabled ad, a keyword with a final URL of its own, a Performance Max asset group, a sitelink, a promotion or a price asset whose final URL or mobile final URL starts with http://. It drafts no change, since the https address is right only where your site answers on it.
- Ads Google flagged for their page. An enabled ad Google’s policy review marks as having a page it couldn’t load, couldn’t reach, or that isn’t where the ad says it goes: high when the ad is disapproved, medium while it still serves.
A page most of whose spend comes from keywords rated below average is the low Quality Score guide’s finding, and a page that fails or redirects when Goldbeater opens it is the destination guide’s. The audit checklist has every landing page line with where to find it in Google Ads.
Questions
- What does landing page experience below average mean?
- Google rates the page your ad sends people to as less relevant and useful to them than the pages of other advertisers whose ads showed for the same search over the last 90 days. It’s one of Quality Score’s three parts, rated keyword by keyword, so the same page can rate well for one keyword and below average for another.
- How do I fix a below average landing page experience?
- Google’s advice is to give people what they searched for, keep the ad’s offer and call to action on the page, make the page easy to use on a phone, and make it load quickly. The search’s exact words needn’t be on the page. If the page rates well for other ad groups, send this ad group to a page of its own.
- Does landing page experience affect my cost per click?
- Quality Score itself isn’t used in the auction. The ad quality it summarizes is: Google says ad quality affects whether an ad can show, where it appears and how much a click costs, and that higher quality ads typically cost less per click.
- Where is the landing page report in Google Ads?
- Under Campaigns, Insights and reports, Landing pages. Each row is a page your ads sent clicks to, with its clicks, cost and conversion rate, and View expanded landing pages lists the addresses people reached once keywords or custom parameters were written into a page’s address.
- What is a final URL in Google Ads?
- The address of the page on your website that people reach after they click your ad, also called the landing page. You set it on each ad, with a separate field for a mobile page. Google’s policy is that it shares its domain with the ad’s display URL and doesn’t redirect to another domain.
- Does page speed affect Google Ads?
- Loading speed is one of the five things Google says to work on when landing page experience is rated below average, and Google says that in retail a one-second delay on mobile can affect mobile conversions by up to 20 percent. A redirect from http to https slows the page too.
- Does converting more raise my landing page experience rating?
- No. Google says conversion rate doesn’t affect the status, and that the conversions you report don’t affect ad quality. It suggests conversion rate as a proxy: a page that converts its clicks far worse than your campaign’s other pages is the one to look at first.