What a homepage is for
A homepage carries an unreasonable number of jobs. The existing customer looking for a phone number. The applicant looking for careers. The buyer comparing three suppliers. The journalist. The supplier. Someone who half-remembers your name.
To serve all of them it stays broad: full navigation, several calls to action, a summary of everything. That breadth is correct — and it is precisely what fails a stranger who arrived with one narrow question.
Someone who clicked an ad about a specific offer lands on a page that does not mention that offer, presents nine other options, and asks them to go find what they came for. A meaningful share leave rather than search. You paid for that click.
What a landing page does differently
One page, one audience, one action. Everything follows from that constraint.
- Message match — the promise the ad made is the first thing on the page, in words close to the ad's own. This single factor moves conversion rate more than any design change.
- Navigation removed. Not a design flourish; a decision. Every other link is a way to leave without converting.
- One call to action, repeated. Not four competing ones.
- The specific objection answered where it arises — price, trust, timing, delivery — rather than left for a phone call that will not happen.
- A form asking for the minimum needed to have a useful conversation. Every extra field costs completions.
When the homepage is genuinely right
This is not an absolute rule, and treating it as one wastes money in the other direction.
Brand and awareness campaigns. If you are buying recognition rather than a specific action, the homepage is a reasonable destination — the goal is that people learn who you are.
Retargeting people who already know you. Someone who visited twice and is deciding may genuinely want the full picture, and a stripped-down page can feel like a step backwards.
Single-offer businesses. If you sell one thing to one audience, your homepage effectively is a landing page. Building a duplicate would be work for no gain.
The test is whether the ad made a specific promise. Specific promise, specific page. Broad message, homepage is defensible.
The mistakes people make building the alternative
Deciding to build a landing page is the easy part. Four ways it still goes wrong.
Reusing one page for every campaign. If three ads with three different promises point at one page, you have rebuilt the homepage problem with fewer options. Message match is the whole point.
Writing the design first. Copy decides the outcome; the design amplifies it. A beautiful page carrying a vague promise converts worse than a plain page carrying a sharp one.
Leaving it uninstrumented. A page you cannot measure is a page you cannot improve. Conversion events, scroll depth and form-field analytics belong in the build, not in a follow-up task.
Testing too early or on too little traffic. Below a certain volume, tests never reach significance and you end up making decisions on noise. At low traffic, improve on form analytics and qualitative evidence instead, and be honest that it is not a test.
The arithmetic that makes this worth doing
Take a campaign at a two percent conversion rate. Getting it to three percent cuts your cost per acquisition by a third, on identical ad spend, permanently.
That is almost always a better return than increasing the budget, and it is available exactly once per page. It is also why we generally will not run ads to a page we have not looked at — the page decides the result more often than the targeting does, and a campaign pointed at a weak page mostly buys an expensive lesson.