Why your best-converting page is not your homepage
A stranger who lands on your homepage and a person who just clicked your ad arrive with different information in their heads. The ad made one promise. The homepage has to introduce a company, which asks the visitor to work out where they fit.
That gap is where paid clicks leak. Google rates landing page experience as one of three parts of Quality Score, and a page that ignores the ad's claim makes the click feel like a wrong turn.
I'm Sterre van Dijk, founder of Seven Purple Mangos. In this post I walk through what the visitor already did before your page loads, why message match keeps the promise, and where the click leaks when the page stops continuing the ad.
You will see how the homepage routes while the campaign page receives, and how one promise and one action beat five. The first screen can look like the ad that produced it. Then we get practical, with one angle, its ad and its page side by side, how to build a page per angle without doubling the work, what I cannot promise you about the numbers, and where to start this week.
What the visitor already did before your page loads
Someone saw one ad, in one place, at one moment. They tapped it because the claim in that ad was interesting enough to spend a click on. That claim is the only thing they agreed to, and your page is the first test of it.
They arrive with one question. It matches the ad sitting in front of them. They are not curious about your company as a whole and they did not come to browse.
Your homepage has to answer for everyone who lands on it, from a brand search to a referral to a typed-in URL. That is exactly why it sounds vague to someone holding one specific promise in their head. The first screen either continues the ad or asks them to start over, and starting over costs you the click you already paid for.
One small reason to hesitate and the tab closes. A first screen that loads slowly, or one extra choice, is enough.
So the page has one job before anything else: keep the agreement the ad just made.
Message match keeps the promise the ad made
Message match is easy to describe. The words, the offer and the image on your page echo the ad that produced the click. Stackmatix, a vendor blog, frames it the same way, with the ad headline as the reference point.
Google's own help page on optimizing ads takes the idea upstream. Group keywords by theme, it says, so each ad can be tailored to that specific theme. Closely paired ads and keywords are more likely to earn interest. That grouping is where a page per angle starts.
A homepage cannot do that. Someone who clicked an ad about one service lands on a page listing all of them and has to work out which one the ad meant. Most people will not do that work, so they leave.
Landing page experience is one of the inputs in Quality Score. Higher scores are reported to mean lower costs per click, with no measured figures given.
I write the page headline from the ad headline, in the same words, before I open Figma. The layout comes after the promise, never before it.
Where the click leaks when the page does not continue the ad
Message match is the promise. These are the places it breaks.
The breaks I look for are the same five: a page that does not match the ad, a form that asks too much, a slow first screen, traffic that should never have clicked, and a call to action buried below the fold. Each one is a break in the same chain.
A practitioner thread on Quora argues the ad itself can be the problem when it promises what the page cannot deliver. One anonymous account there reports a bounce rate near 90% from a Google text ad. That is one person's unverified experience, not a benchmark for yours.
The thread running through every cause is the same break between what was promised and what the visitor found after the click.
My own check takes a minute, and it is simple. Read the ad out loud, then read the first screen out loud. If they do not sound like the same sentence, the page is where the click leaks.
The homepage routes, the campaign page receives
Once you hear that break, you are looking at two pages with two jobs. The homepage is the router: it serves everyone who arrives and points each person toward the right path.
A campaign page serves one kind of visitor. Its job is to finish the argument the ad started. Treat the two as versions of the same page and paid clicks leak.
My approach keeps the homepage as the router, with clear lines to your main paths, and builds a separate page for every campaign angle you put real spend behind. The homepage keeps doing its job. The campaign page gets to do one thing well.
One promise and one action on the campaign page
With the routing job sorted, the campaign page can hold one promise and ask for one action.
Start by taking the site menu off it. The ad already did the sorting, so the visitor should not have to choose again. That work belongs on the homepage, and the principles behind creating intuitive navigation apply there, not here.
Decide the single action before you open a design tool. One form, one button, one next step.
If the button sits below three screens of copy, it is hidden. Move it up instead of restyling it.
Every extra field gives the visitor a reason to leave. Name and email usually carry the whole job.
Cut the section that lists everything else you do. That belongs on the homepage, where a visitor is still working out what they need. Here, the promise and the action are the layout.
The first screen should look like the ad that produced it
With the extra sections gone, the first screen has room for its real job: looking like the ad that produced the click. Carry the ad's image, colors and type into it. People recognize what they clicked before they read a word, and that recognition does quiet work.
Order matters as much as the pieces. Headline first, image second, button third, all inside the first screen on a phone. That is the order the click arrived in.
I keep the ad file and the page file in one Figma document so I can hold them side by side while I design. Distance between two files is how they drift apart.
Match the promise word for word where the sentence allows it. Google's help pages make the same case for pairing each ad tightly with one theme.
Watch type sizes across both. A headline that reads well in a feed often needs to be shorter on a page, so trim the copy instead of shrinking the type. That same care about where a piece lands goes into every marketing material I design.
How to build a page per angle without doubling the work
The frame itself does not need building five times. Start by listing your campaign angles. An angle is one reason someone buys, such as speed, price, or a service built for agencies with live campaigns. Google's help page on ad groups leans the same way, grouping keywords by theme so one ad can speak to one theme.
Then build one page per angle you are actively spending on. Two pages that cover your two biggest angles beat six half-built ones nobody visits.
Build the shared frame once, then swap the promise, the image and the headline for each angle. Framer suits this well because there is no code to maintain, so the frame holds while the words change.
My structure keeps this planned. I work on fixed delivery days, we run a weekly check-in, and you get a fixed price before you decide anything. Extra angles become scheduled work, never a surprise invoice.
Retire any page that gets no traffic. A page with no clicks gives you nothing to measure.
Next, let me walk you through one angle, its ad and its page, side by side.
One angle, its ad and its page, side by side
With your angles listed, put one angle, its ad and its page side by side.
Take speed for agencies working against a client deadline. The ad headline reads: ad files delivered in 48 hours for agencies with a live campaign. The page headline repeats that claim word for word. The subhead names the reader: for paid media agencies with campaigns running now. Then one button: send me the brief.
The first screen uses the same image and color as the ad. The visitor sees the thing they clicked before reading anything else. There is no site menu, no services list and no pricing table. The only thing there is the next step.
A second angle gets its own page, its own ad and its own headline. The two are never mixed. This walkthrough is the structure I use, and it is an illustration rather than a client result.
What I cannot promise you about the numbers
What I showed you is a structure, and I have no number to attach to it. No source behind this piece measured how much a matched page lifts conversion, so I will not quote one. Anyone who hands you a percentage here is guessing.
Quality Score is a 1 to 10 rating per keyword, and it reads as a scale rather than a result. Read it as a signal, never as a target.
The near 90% bounce rate from that Quora thread is one anonymous, unverified account. It shows a symptom worth recognizing. It tells you nothing about what your page will do.
You can measure this inside your own ad account. Pull cost per click by ad, conversion rate by page, and bounce rate by page. Compare each page against the version you ran before it.
One honest note. The page sits alongside the ad copy, the offer and the quality of the traffic you bought. A better page will not rescue a weak offer.
Where to start this week
Start with what you already have running. Open your ad account and list the angles you are actively spending on in two columns: angle and spend. That list is your build order, and the top line gets a page first.
Write that page's headline straight from the ad headline, in the same words. Same promise, same phrasing, no rewriting for polish.
Then build it sparse, the way the earlier sections describe. Leave the homepage alone. It stays the router for everyone else.
Check both the ad and the page on a phone before you publish.
Give it two weeks. Then compare that page's bounce rate and conversion rate against what it did before. Two weeks is short of a controlled test, so read it as a direction rather than proof.
Conclusion
Your homepage has a different job. A campaign page receives one visitor with one promise and one action. When the ad and page match, the click no longer leaks.
I would start with the angle you are already paying for. Build that page this week.
Frequently Asked Questions
Why does message match matter for paid clicks?
Message match means the words, offer and image on the landing page echo the ad that produced the click. The visitor clicked because of one specific claim, so the page must continue that promise. If it doesn't, the click feels like a wrong turn and the visitor is likely to leave.
Why shouldn't a campaign page just be the homepage?
A homepage has to serve everyone who arrives, from brand searches to referrals to typed-in URLs, so it sounds vague to someone holding one specific promise. The post recommends keeping the homepage as a router that points visitors to the right path, and building a separate campaign page for each angle you actively spend on.
What is the test for whether a page is leaking clicks?
Read the ad out loud, then read the first screen out loud. If they do not sound like the same sentence, the page is where the click leaks. The post lists five common breaks: a page that does not match the ad, a form that asks too much, a slow first screen, traffic that should never have clicked, and a call to action buried below the fold.
What should a campaign page include?
A campaign page should hold one promise and ask for one action. Remove the site menu and any section listing all services, decide the single action first, and keep the form to the fewest fields possible, with name and email usually carrying the whole job. The first screen should carry the ad's image, colors and type, with the headline first, image second and button third.
How can I build a page per angle without doubling the work?
List your campaign angles, meaning one reason someone buys, then build one page for each angle you are actively spending on. Build a shared frame once and swap the promise, image and headline for each angle. Retire any page that gets no traffic, and start with the top angle on your spend list.
