Why the click and the page must agree
A paid click is the end of one decision and the start of another. The person decided to click because the ad said something that fitted what they wanted. On landing they decide, in a second or two, whether to stay, and they decide it by looking for what the ad said. If the headline repeats it, the search continues on the page; if the headline says something else, the visitor has to work out whether this is still the right place, and many will not bother.
That is the whole mechanism, and its effects show up in two places. On the page, conversion rate (CVRShare of visits or clicks that convert.) falls as mismatch rises, because visitors who leave in the first seconds never see the offer. In the account, Google Ads documents landing page experience as one of the components of Quality Score (Quality ScoreGoogle's 1–10 keyword diagnostic.), alongside expected click-through rate and ad relevance, and describes it in terms of how relevant and useful the page is to people who click the ad. The Quality Score guide covers how the diagnostic works and what it does and does not affect; the practical point for this page is that a mismatched landing page is penalised twice, once by the visitor and once by the auction.
Message match is not the same as repeating the keyword. The keyword is a proxy for an intent, and the page has to match the intent, which the search intent guide explains from the organic side. The ad is the bridge between the two: it translates the intent into a promise, and the page has to keep that promise. A landing page brief records each traffic source and the promise it makes before design starts, so the page can be built to keep them.
The comparison of click-through rate and conversion rate shows how reading the two rates together exposes a broken promise between ad and page.
Match at three levels
| Level | What has to agree | How it fails | Where to fix it |
|---|---|---|---|
| Keyword to ad | The ad's headline reflects what the keyword's searches want | A generic ad across an ad group of mixed intent | Ad group structure and ad copy |
| Ad to page | The page's headline and first screen restate the ad's promise, offer and any price or claim | One page for every ad group; the ad promises a discount the page does not mention | The landing page, or a variant per ad group |
| Page to form | The form, the button and the confirmation continue the same promise | A "free audit" ad leads to a "contact us" form | The form and its labels |
| Tone and format | A question-style ad lands on a page that answers; a comparison ad lands on a comparison | An ad in plain language lands on a page in corporate language | Copy on both sides |
| Audience and device | A mobile ad lands on a page usable on mobile; an ad for a segment lands on that segment's page | One desktop page for every placement | Device-specific layout; segment variants |
The first level belongs to the ads account, and the responsive search ads guide covers how headlines are assembled and tested; this guide assumes the ad is right and concentrates on the second and third levels, where the page is the variable. The last two rows are the ones most audits forget: a page can restate the ad's words and still break the match by changing register, or by being unusable on the device the ad was shown on. When the landing page is a product listing rather than a single offer, the guide to merchandising the store for a campaign keeps the match by opening the collection with the products the ad showed.
Keyword intent to page
The unit of message match is the intent behind a group of keywords, not the individual keyword. Searches that want the same thing in different words can share a page; searches that use the same word for different things cannot. "Accountant for small business", "small business accounting services" and "bookkeeper for limited company" are one intent with one page; "accounting software" is a different intent and needs a different page, however similar the words.
The method is to read the search terms report rather than the keyword list, because the terms that actually bring traffic are the ones the page has to match. Group the terms by what the searcher wants to happen next: get a quote, see a price, compare options, learn how something works, find a local provider. Each group is a page or a variant. Within a group, the most common phrasing becomes the headline, so that most visitors see their own words.
Where the account is structured by product rather than by intent, the mapping can reveal that one ad group serves two intents and should be split, or that two ad groups serve one intent and could share a page. That is a structural change in the account rather than on the page, and the landing pages hub sits beside the Google Ads search guides for that reason.
Mismatch can also run the other way: a page that is right for the intent but an ad that over-promises. A page cannot fix an ad that says "free" when the page says "free trial", or "from" a price the page never shows. Reading the ad and the page together, as the visitor does, is the only way to catch this.
Dynamic text and page variants
There are two ways to make a page agree with many ads: change the words on one page, or make several pages.
Dynamic text replacement reads a parameter from the landing URL and swaps a headline, a subhead or a line of copy accordingly. The ads account passes the parameter in the final URL, and the page reads it and shows the matching text, with a default when the parameter is missing or unexpected. Used well, it lets one page restate dozens of ad headlines. Used badly, it produces headlines that are grammatically wrong, that repeat a search term containing a competitor's name or a typo, or that promise what the page does not deliver. Rules that keep it safe: a fixed list of allowed values rather than free text from the URL; a human-written headline for each value; a sensible default; and nothing below the headline that contradicts any of the values.
Google Ads Help documents keyword insertion as a feature that updates ad text with the keyword that triggered the ad; the landing-page equivalent is a separate mechanism built on the site, and the two should be designed together so that the ad and the page swap in the same way and never swap in different directions.
Page variants are the other route: a copy of the page per intent group, with its own headline, hero, proof and form. They cost more to build and maintain, and in exchange they can differ in more than words: a variant for a comparison intent can lead with a comparison table, a variant for a local intent can show the nearest branch. The practical split is to use variants between intent groups and dynamic text within them.
Both routes create a maintenance obligation. A price change, a new guarantee or a retired offer has to reach every variant and every dynamic value, which is why an inventory of landing pages and their parameters belongs in the account's documentation.
Scent from click to form
Message match does not end at the headline. The visitor who recognised the page then reads down to the action, and each section either continues the thread or breaks it. Information scent is the term for this: the cues that tell someone they are getting closer to what they want.
The scent breaks in predictable places. The ad promises a specific product and the page's benefits describe the company. The ad names a price and the page's call to action says "request a quote". The ad says "free consultation" and the form's button says "Submit". The ad speaks to a segment and the proof on the page is from a different one. Each break is a moment where the visitor has to re-decide whether to continue.
Continuing the scent means carrying the ad's promise through the subhead, the benefits, the proof and the form. The proof should come from people like the visitor the ad targeted. The form should ask for what the promised next step needs and the button should name that step. The confirmation page or message should restate the promise once more and say when it will be kept. The anatomy guide describes each of those sections; message match is the thread that runs through them.
Scent also runs through the URL and the page title. A URL that contains the campaign's subject and a title tag that repeats the headline reassure the visitor glancing at the browser, and they help the people reviewing the account find the page later.
One campaign, three intents, three results
A furniture retailer runs one "sofas" search campaign with 40 keywords sending all traffic to the sofa category page. The search terms report shows three clusters: "corner sofa" and variants (around 55% of clicks), "sofa bed" (25%) and "leather sofa" (20%). The category page's headline reads "Sofas"; its first screen shows a fabric two-seater. Conversion to purchase runs at 0.9% across the campaign. The team splits the ad groups by cluster, writes a headline per group, and builds three landing variants: each leads with the matching product type, a filtered product grid and reviews mentioning that type. Over the following six weeks the corner-sofa variant converts at 2.1%, the sofa-bed variant at 1.7% and the leather variant at 1.5%; the blended figure rises to about 1.9% on the same spend, and the landing page experience component moves from "below average" to "average" or better on most keywords.
Auditing paid landing pages
An audit answers one question for every ad group: does the page a visitor lands on say what the ad said? The method is mechanical enough to repeat monthly.
- Export the pairs. From the ads account, list every ad group with its final URLs, its top search terms by clicks and its ad headlines. One row per ad group and URL.
- Open each pair as a visitor. Read the top search term, read the ad, open the page on a phone. Note whether the headline restates the ad, whether the offer and any claim in the ad appear in the first screen, and whether the first call to action continues the promise.
- Score the match. A simple three-point scale works: matched, partly matched, mismatched. Record the reason for anything below matched.
- Join to performance. Put conversion rate and cost per acquisition (CPACost of each conversion.) next to each row, by device. Mismatched rows with high spend are the priority; matched rows with poor conversion have a different problem and belong to the forms, proof or speed guides.
- Decide the fix per row. A new headline, a dynamic value, a variant, a split ad group, or a changed ad. The fix is not always on the page.
- Re-audit after the change. Message match decays as ads are rewritten and pages are edited by different people; the audit is a routine, not a project.
Click-through rate (CTRShare of impressions that result in a click.) belongs in the audit too, read alongside conversion: an ad group with a high click-through and a low conversion rate is the signature of a promise the page does not keep. On LinkedIn, where each click is costly, the diagnosis of LinkedIn ads with clicks but no conversions treats a page that does not repeat the ad's offer as one cause among several to rule out in order.
Message match is one of several checks worth running together, and the landing page audit gives you a structured way to review the promise, proof and friction on each page.
When paid clicks are bought for reading rather than converting, the Google Ads setup for website traffic shows how matched landing pages and engagement checks keep cheap visits worth having.
Visitors from a short video arrive with the video's promise in mind, and the TikTok setup for website traffic shows how to land them on exactly that subject inside the in-app browserThe browser built into a social, video or messaging app that opens a tapped link inside the app instead of handing it to the phone's own browser..
When the ad opens an app store rather than a web page, the same principle applies to the listing's first screenshots, and the tutorial on custom store listings for app campaigns shows how to give each campaign theme its own matching page.
Where visitors leave a page within seconds and the match looks sound, the diagnosis for a landing page with a high bounce rate checks speed, the first screen, traffic intent, overlays and the bounce definition itself.
Common mistakes
- One page for the whole account. The homepage or a category page cannot restate every ad.
- Matching the keyword, not the intent. Repeating a word is not the same as answering the need behind it.
- Dynamic text from raw URL input. Headlines built from unvalidated parameters produce nonsense and worse.
- Over-promising ads. A page cannot keep a promise the ad should not have made.
- Scent that stops at the headline. The benefits, proof, form and confirmation have to continue the thread.
- Variants nobody maintains. A retired offer that lives on in one variant is a mismatch the account created.
- Auditing on desktop. Most paid visitors arrive on a phone; the match has to hold on the first screen there.
- Treating Quality Score as the goal. The diagnostic improves when the visitor's experience does, not the other way round.
