Several ad themes were competing for the same clicks in a paid search program I was consulting on at one of the largest players in marketing software. The theme pulling the highest click-through rate had the lowest conversion rate of the group.
Every ad in that group pointed at the same landing page, and the page was written about the product. The winning theme made a promise the page never repeated.
That gap between what earns the click and what greets the click is the most expensive quiet failure in paid acquisition, and large software companies produce it reliably, because the page gets built from the inside out.
The page everyone in the building likes
Here is how a landing page usually gets made at scale. A product marketer writes down what the tool does. A copywriter turns that into a headline describing the tool. A designer arranges the capabilities underneath it in a tidy grid.
Then it goes to review, and review is where the page quietly dies. Every reviewer already knows the product, so a page written about the product reads as accurate to all of them, and accuracy feels like approval.
Accuracy is the wrong test. Everyone in that room has already bought the product. The person the page was built for had never heard of it thirty seconds earlier.
The visitor arrives mid-thought
The click did not start on your page. It started on an ad that named a problem the person was already carrying around, and the ad earned the click by naming it in words they recognized.
Then the page opens on your product, and the thread breaks. The visitor is handed the translation work, in the moment before they decide whether to stay.
A landing page has one job: continue the sentence the ad started, in the words of the pain that made the person click.
The paid search trade has a plain name for the fix. Message match is a measure, in Unbounce's wording, of how well your landing page copy matches the phrasing of the ad that brought the visitor there. Practitioners also call it scent: the visitor is following a trail of words, and the page either continues the trail or ends it.
Back to that account. Several themes ran against overlapping searches, all of them landing on one shared page.
The theme that won the most impressions and clicks came last on conversion. People clicking it were coming to see one specific thing, and the shared page never put that thing in front of them before the fold ended.
The ad had sold a promise the page never agreed to keep, and we were paying full price for every one of those clicks.
Three elements, assembled in that order
So we stopped starting with headlines. Every page in the program had to align three things before anyone wrote a word of copy.
First, the customer pain point, phrased the way the customer phrases it. This is the underlying why, the reason the search happened at all. Where that phrasing comes from when you have no direct line to customers is a method of its own, which I cover in voice-of-customer research without customers; this framework picks up the moment that list of named problems exists.
Second, the ad theme the page has to continue. Whatever the campaign promised, the page repeats it near the top, close to verbatim, before it says anything clever.
Third, the product feature that answers the pain, named specifically. A generic claim about powerful automation proves nothing to anybody. The specific version names the thing a real customer opens on a Tuesday morning and gets something out of.
Three is a constraint, and the constraint does most of the work. A page that cannot fill all three slots is a page with no argument yet, which is useful to discover in a brief rather than after the build.
The sentence that vetoes everything else
The glue holding the three elements together is a Single-Minded Proposition: one internal sentence naming the only argument the page is allowed to make. Copy, layout, imagery, and the call to action all ladder up to that sentence.
Advertising planners have been writing these for decades, and the craft's own advice is that the proposition should be no more than one sentence, and that it is the hardest group of words in the entire brief to write. Both halves of that are true in software too.
One page, one proposition. The page that makes three points wins three meetings and zero customers.
The proposition earns its keep in the review meeting. Every stakeholder shows up with one more true thing the page could mention, and every one of those additions is defensible on its own terms.
Without a proposition, all of them get in and the page becomes a list. With one, the question stops being whether an addition is true and becomes whether it serves the argument. Most additions fail that question, which is the correct outcome.
The belief you commit to before the build
Even a well-built page teaches you nothing if it launches with no stated belief attached. Most of the live pages I have inherited had no record anywhere of what the people who built them expected to happen.
So every brief carried a hypothesis, in a fixed shape. If we do this, then this happens, because of this belief about the customer. Then one measurement line: which money event judges it, over what window.
Write the hypothesis before the headline. If/Then/Because turns a page launch into an experiment you can lose usefully.
The because is the part that matters. If and then only describe the change you made and the result you were hoping for. Because states what you actually believe about the person clicking, and that belief is the thing a losing test kills.
A page that loses with a because attached retires a belief about your customer, which is worth considerably more than the page was. A page that loses without one gets replaced by whatever the next person's taste suggests.
Framed this way, a page launch is an experiment, which means it competes for a slot against every other experiment the quarter could run. Ranking those slots is the learning agenda, and pages that arrive with a hypothesis attached tend to win the argument for a slot.
A worked example, end to end
Take a neutral case. A scheduling tool sold to small clinics.
The pain, in the customer's words: patients do not show up, and the front desk finds out when the room sits empty. The ad theme running against it: fewer no-shows. The capability that answers it: automated reminder texts with one-tap confirmation.
The proposition: cut no-shows without adding work at the front desk. That one sentence decides everything else. Shared team calendars stay off the page. The integrations grid stays off the page. Both are real and neither serves the argument.
The hypothesis: if we build a page whose whole argument is cutting no-shows, then clicks from the no-show theme will convert above the general product page, because the visitor's reason for clicking is answered before they scroll. Measured by incremental signups and net revenue over four weeks.
Now the page has a job, a judge, and a losing condition. Design and copy can still argue about execution, and nobody has to argue about what the thing is for.
Name the page for its argument
One small discipline saved more time than it had any right to: name each page for what it is strategically about, in words a busy executive can read once. Cut no-shows for small clinics. Nothing more polished than that.
Marketing-flavored names hide the argument, and a portfolio of them is impossible to review. When a page name needs a paragraph of explanation, the page underneath it usually does too.
Named plainly, the whole set becomes legible: leadership reads a list of arguments the team is running, sees which pains are covered and which are orphaned, and stops asking what all those pages are for. Which concepts then deserve budget is a scoring problem, and I run that with the creative scoring system.
Run it on your next page
You do not need a research budget or a redesign to start. You need one page and the discipline to build it in this order:
- Write the pain first, lifted from something a real prospect said, rather than paraphrased from a positioning doc.
- Name the ad theme the page must continue, and repeat that promise near the top in words the visitor will recognize from the ad.
- Pick one capability that answers the pain, specific enough that a customer would recognize using it.
- Write the Single-Minded Proposition in one sentence, and use it to kill every addition that fails to serve it.
- Write If/Then/Because plus the money event that judges the page, before anyone drafts a headline.
- Name the page for its argument, in language leadership can read without a translator.
Do that and the page stops being a brochure with a form stapled to it. It becomes an argument you can win or lose on purpose, traceable back to a problem a real person reported. Once the argument lands, the friction that remains is mostly the form itself, and what order you ask its questions in is worth more than most teams assume.
Skip it and the arithmetic is unkind. Every click you pay for arrives at a page answering a question the visitor never asked, and the money leaves regardless of how much the room liked the design.
If your paid traffic lands on pages that describe your product and convert nobody, this is the kind of rebuild I run. Let's talk.