The site went live, the product is real, and nothing much happens. The instinct is to blame the design, so the founder commissions a redesign, and six weeks later the new site converts about as well as the old one did. What was wrong wasn’t visual, so making it prettier didn’t fix it.
None of the five below is about how the site looks on a phone. That argument was settled years ago, and every current template handles it. These are decisions the site is making on your behalf, mostly by accident.
The site explains the product before it says who it is for.
A stranger arrives with one question, and it is not, “What does this do?” It is whether this is for them. A homepage that opens with capability answers a question nobody asked and leaves the real one open, so the reader settles it by leaving.
Naming the buyer costs nothing you were going to get anyway. The visitors it turns away were not going to buy. The ones it keeps now have a reason to read the second paragraph, which is the only thing the first paragraph has to achieve.
This is the same decision underneath everything else on the page, which is why a site that has not made it cannot be fixed by rewriting. The sentence has to be settled first.
Every page asks for something different.
Book a demo. Join the waitlist. Get the newsletter. Follow us on LinkedIn. Read the blog. Each was reasonable when it was added, and together they mean the site has no ask at all. A reader who is offered five next steps takes none, because choosing between them is work and leaving is not.
Pick the one action the site exists to produce, and point every page at that one. The rest can stay as links. They stop competing the moment only one of them is a button.
The site is written for investors.
Deck language migrates onto the homepage without anyone deciding to put it there. Platform. Infrastructure. The category the company hopes to define, named as though it already exists. It reads as ambition to an investor and as fog to a buyer, and the buyer is the one on the page.
Customers describe their problem in the words they used before your product existed. Those are the words that belong at the top. Category language isn’t wasted; it is just early, and it belongs on the about page where the people who care about it are already looking.
The test takes a minute. Show the homepage to someone who has the problem and does not know the company, and ask them what it sells. If they answer with your category rather than their problem, the page is talking to the wrong room.
Nobody can tell what happens after the form.
On most startup sites, the form is the whole conversion path, and the button next to it says Submit. A stranger is being asked to hand over an email address in exchange for an unspecified event at an unspecified time, from an unspecified person.
Say what happens and when. Who replies, roughly how quickly, and what the first exchange is for. It costs one sentence, it requires no design work, and it is the cheapest change on this list.
The same applies to what you ask for. Every additional field is a reason to stop, and the fields founders most want are the ones a stranger is least willing to give before anyone has spoken to them.
The site was launched and then never touched again.
A first site gets written in the two weeks before a launch, and then the company changes, and the site does not. Eighteen months later, it describes a product that has moved, a buyer the company stopped chasing, and a plan abandoned in the second quarter. It is the only employee still repeating the old strategy.
Nobody reads their own homepage. That is not carelessness; it is familiarity, and it is why this one goes unnoticed longer than the other four combined.
Put it on the calendar. Once a quarter, read the site as a stranger would, out loud, and cut every sentence that is no longer true. That quarterly pass is the least glamorous part of Brand, Positioning, Website, Demand, Analysis. Then Repeat.
What we would do first on a site like this is not mentioned above. We would read it out loud as if we had never heard of the company, and mark every sentence that survives only because somebody wrote it once. That afternoon usually settles what the redesign was intended to accomplish.
Tell us what you’re building.