Audit Your Website Like a Customer Who Doesn't Care About You

Summary

Everyone who runs the website, the product, and the inbox knows the ritual: traffic looks healthy, trials stay flat, so you open the homepage and move a button. This is for the evening when that stops working. You'll run a six-step audit built around one question—what would a busy stranger actually understand in five seconds—and each pass can be completed in a single sitting, no design tools or budget required. Along the way, the article leans on concrete examples like Calendly's plain-language headline and Linear's quiet hero, while pushing back on the sacred cows of conversion culture: the single CTA, the above-the-fold fetish, and the instinct to add more sections. The deliverable is not a redesign plan but a short list of edits with reasons attached. That's the point: a homepage you can read like a customer, not like its parent.

It's eleven p.m., and you've just reordered the hero section for the fourth time this month. The traffic numbers look respectable—visits are up, people are clicking through—and you could count this week's signups without taking off your socks. You know the page too well to know if it's good. Every word was a decision. Every margin has a reason. That's the entire problem. The visitor on the other end has no context, no patience, and no reason to believe you. They'll decide whether to stay in the time it takes to close a tab. So here's an audit built for that reality: six passes over your site, each designed to be done in a single evening by the person who is also designer, copywriter, and support team. No analytics overhaul. No "growth loops." Just a way to catch the pages you'd scrap if you weren't emotionally married to them.

1. Name the one decision you want this page to cause

The principle: a page doesn't convert because it's pretty; it converts because it makes one next action feel obvious. Before you touch a color, write the sentence that best completes the thought: "After viewing this page, the visitor should ___." If your answer includes words like "understand," "learn," or "feel," you've described a vibe, not a conversion goal. (If that's genuinely the goal—a portfolio that exists to be remembered—fine. But admit it, and skip the "we help brands" nonsense.)

The canonical example here is Calendly. Their headline is essentially "schedule meetings without the back-and-forth emails." Notice what it doesn't do: no mention of calendar integrations, no "scheduling platform for modern teams," no clever abstraction. It names the enemy (the email back-and-forth) and the outcome (scheduled meetings). That's the whole trick. For a software landing page, the sentence might be "Start a free trial" or "Book a 15-minute demo." For an agency or portfolio site, the decision might be "Read the case study" rather than "Hire us"—because a stranger who just met you is not ready to hire you.

Write that sentence. Then read your page and ask: does every section serve it, or is the page a museum of things you're proud of? Most of the time it's a museum. If you're torn between building a software-style page, an agency page, or a portfolio, the format matters more than any one section—our guide to choosing the right website format gives you the short version.

2. Run the five-second verdict

The principle: comprehension is upstream of persuasion. A visitor should be able to answer "what does this company do?" after a five-second look, before reading a single full sentence. If they can't, no amount of "above-the-fold clarity" saves you—because they've already left.

We all nod along to the "above-the-fold" lecture until we remember that on a phone, the fold is roughly one sentence and a button. The useful version of the advice isn't "pack everything into the first screen." It's "make the first thing visible meaningful." The favorite example in every roundup is Linear's hero: a clean headline, a short subheading, and a single call-to-action. The aesthetic is minimal, sure, but what actually matters is that you can parse what the product is in the time it takes to blink.

How to test it honestly: open your site on a phone, hand it to someone who has never seen it, and ask two questions: "What does this company do?" and "What would happen if you clicked that button?" If the first answer is vague ("something with analytics") and the second is "I don't know," you've found your edit. It's not a design problem. It's a decoding problem. Fix the copy, don't redesign the layout. (If you have a friend who's too polite, pay them in coffee and watch their face on question two.)

3. Count your CTAs like they're expensive (because attention is)

The principle: each page should push toward one primary goal, and every competing ask makes each one weaker. Simple. The famous example: Notion, whose page has been praised for pointing at a single "Get Notion free" button. That works because the product's self-serve model makes "try for free" a genuinely obvious next step.

But here's where the "best practice" turns into cargo cult: the rule was always one decision per page, not one button. For a service business, a single CTA that says "Contact Us" is not a decision—it's a homework assignment. The visitor has to figure out what the email should say, whether they're ready, and whether you're the kind of person they want to reach. The decision at that stage might be "Read the case study that proves we can do this," or "Download the pricing sheet," or "Reply to this one question on a form." One of those is a smaller ask, and it's the one worth making.

That's also where the usually-quoted single-CTA advice and the reality of a small site diverge. You don't have a funnel team or a marketing site with ten interconnected pages. Clarity is your substitute. So the question is never "is there only one CTA on this page?" but "is there one obvious next step for the visitor standing in front of me?"

What people copy	What actually works
One CTA per page	One clear decision per visitor stage
Minimal hero with a lot of whitespace	A headline a stranger can repeat back
"Schedule a demo" in the nav	A next step that matches how ready the visitor is
Feature list under the hero	The outcome the customer actually gets

Use that table when you're auditing your own page. Check the box on the left and you've matched the visual pattern; check the box on the right and you've matched the mechanism. If the two don't line up, the design is decoration. (For more on the mechanics of landing-page structure, the proven best practices for SaaS landing pages are the longer version of this argument.)

4. Read it the way your customer speaks, not the way your CR file speaks

The principle: jargon is how a company's vocabulary takes over its homepage. You've said a phrase like "the leading collaborative intelligence platform" so many times it sounds normal to you. To a first-time visitor it sounds like a ransom note. The fix isn't to sound dumb; it's to sound specific.

Take the Calendly standard we met in step one: "schedule meetings without the back-and-forth emails." No glossary entry required. The page's headings should be the same way—sentences someone would say in an email to a colleague, not phrases from your pitch deck.

Running this audit pass is brutal but simple: go through every heading on the page and ask whether you'd understand it without having already drunk the company Kool-Aid. For a product you launched yourself, you will not be able to do this. Print the page, or screenshot it, and put it away for a day. Then read it with the eyes of someone who just googled the problem you solve. If a heading could be swapped with a competitor's and nobody would blink, the problem isn't differentiation; it's that the page never said what you do.

5. Make a trust ledger

The principle: conversion is the moment a stranger decides to trust you with something—their email, their time, a conversation. They don't know you, and your logo has no brand halo to lean on. So the page is doing trust work whether you planned it or not.

Walk through your page and list every claim that asks for trust: "We're the best," "Trusted by teams," "Enterprise-grade," and so on. Next to each one, put the concrete thing that backs it up. A testimonial with a first name and a specific outcome beats a logo grid of companies nobody's heard of (and beats a logo grid of companies they have heard of, if those logos are decorative and unsourced). An agency's case study that shows a real before-and-after does more work than a whole page of "we craft meaningful digital experiences." A portfolio with one strong, narratively explained project does more for a designer than a gallery of twelve half-finished pieces.

The practical output of this pass is a list. Eliminate every trust claim you can't back up. Then move the strongest proof as close to the CTA as possible. On a landing page for a tool, that usually means a screenshot of the actual product, clearly labeled, not a stock image of people staring at a chart. If your page is fundamentally a portfolio, the same logic applies—one proof point, positioned like evidence. It's the exact approach spelled out in our guide to building an industry-specific portfolio site.

6. Cut one section this week—no, really

The principle: for a site run by one person, focus is the only scalable advantage. Every additional section asks for a little more of a visitor's attention, and they have almost none. The default state of most sites is addition: a comparison table, a resources section, a "why choose us" list, a blog excerpt, a testimonial carousel, a client wall. A small site can't maintain content for all of those, and the visitor doesn't read most of them anyway.

So the audit ends with a cut. Pick the section you're most attached to—usually it's the one you spent the most time building, or the one your friend said was "really cool"—and hide it for a week. Not forever. Just long enough to see whether anyone asks for it. If no one asks, delete it. If they ask, you have your answer and a great story about being data-driven (we all need one).

This applies even to the risky stuff. Cut the paragraph that explains your company's origin story. Cut the FAQ question that exists because someone asked you once at a networking event. Cut the feature list that you wrote because the engineering side of you likes lists. What remains—a page that names the problem, offers a concrete next step, and shows one piece of proof—is not a masterpiece. It's better than a masterpiece: it's a page that respects a stranger's time.

If you did all six passes, you now have a list. Not a redesign plan, not a mood board—a list of edits with reasons attached. A headline that names the outcome. One decision per page, stage-appropriate. A handful of sections you're prepared to delete. That's the whole deliverable.

And one last thing, because the internet is full of people promising "50% more signups" if you only do their 47-point checklist: nobody can promise you a number. What the audit promises is more modest and more reliable—the fair chance that a stranger who has never met you will understand what you do, in the five seconds they give you, and make the one decision you actually want them to make. That's a perfectly good job for a homepage. It's also, annoyingly, the entire job.