Every conversion-rate-optimization course will teach you the same set of dark patterns: countdown timers that reset when you reload the page, fake stock counters, urgency banners that have been there for three months, social proof notifications that pop up regardless of whether anyone actually bought anything. They work, in the narrow sense that they lift conversion in the short term. They also poison the well for everyone, including you, by training visitors to distrust everything they read on a landing page.
The frustrating thing is that the manipulation is unnecessary. Honest landing pages can convert just as well, and usually better, once you account for the back-end effects: lower refund rates, fewer chargebacks, better word of mouth, customers who come back. The trick is that honest pages have to be built more carefully, because they cannot lean on tricks to compensate for vague messaging.
What a landing page is for
A landing page has exactly one job: take a person who has just clicked something, understand what they were promised on the click, and either confirm that they are in the right place and what to do next, or politely send them somewhere more useful. That is it. Everything else is in service of that.
The most common mistake is treating a landing page like a homepage. Homepages serve everyone — investors, journalists, candidates, customers, current users. Landing pages serve one person at one moment with one decision in front of them. Mixing the two is a fast way to lose the conversion.
The structure that almost always works
You can build a strong landing page from six elements, in roughly this order. Not because it is magic, but because it matches the order in which a visitor's questions arrive.
- A specific headline that names the outcome. Not a clever phrase. A plain statement of what changes if they say yes.
- A subhead that names the audience and the mechanism. Who is this for, and how does it actually work? One sentence each.
- A primary call to action, visible without scrolling. The visitor should be able to act immediately if they are already convinced.
- Evidence. Customer quotes, real numbers, screenshots, demonstrations. Specific evidence beats general evidence; one detailed case study beats five vague testimonials.
- An honest answer to the biggest objection. Whatever the most common reason for not buying is, address it directly. Pretending the objection does not exist does not make it go away; it just lets the visitor leave with it unresolved.
- A second call to action, with a small reassurance next to it. Money-back guarantee, free trial terms, what happens after they click — whatever lowers the perceived risk of saying yes.
That is enough for most pages. Adding more sections rarely improves conversion. It usually dilutes attention.
The copy is most of the work
Designers will tell you the design is what converts. Designers are wrong, in a friendly way. The design has to be clean enough not to interfere, and beyond that, the words are doing the heavy lifting.
Good landing page copy has a few properties.
- It is specific. "Save time" is generic. "Send invoices in two minutes instead of fifteen" is specific. Specific is more believable, even when it sounds smaller.
- It uses the customer's words. If you have done customer interviews, you have a list of phrases people actually use to describe their situation. Use those phrases. They convert because they recognize the reader.
- It does not oversell. Claims that are obviously inflated cause the entire page to lose credibility. Underselling slightly often works better than overselling slightly.
- It assumes intelligence. Visitors can read. They can think. Treating them like they need to be tricked into the next step is both rude and ineffective.
What to test, and what not to
Most A/B tests on landing pages are theater. They run on too little traffic to produce statistically meaningful results, they test trivial details like button color, and they are stopped early when one variant looks like it is winning. Real testing needs real volume and real patience, neither of which most pages have.
If you do test, test things that could plausibly change behavior: the headline, the offer itself, the structure of the proof. Skip the button-color tests. Skip the comma-vs-period tests. They will not move the number meaningfully even when they "win."
For most pages, the higher-leverage move is not testing variations of the same page. It is rewriting the page from scratch, after another round of customer interviews, every six months. The best landing pages get rebuilt periodically as the team learns more, not optimized into local maxima.
The honest finish
A landing page that converts well, ethically, looks deceptively simple. It says exactly what it offers, to exactly the right person, with exactly the right proof. That clarity is hard to produce and easy to admire after the fact. The trick is that there is no trick. The work is the work, and the work is mostly thinking carefully about who is reading and what they actually need to know to say yes.