Website Content Audit: Does Your Copy Do Its Job?

A website content audit, for a small site, checks whether a first-time visitor can tell what your product is from the headline alone, and whether your other pages back it up. Read your copy aloud, mark what sounds vague, then swap in another product's name and see if the line still works. A line that survives the swap is generic.

Words are the cheapest thing on a site to fix, and the first thing a visitor has to work with. This guide gives you one method for checking them in a sitting, with a pen and no special software.

What is a website content audit for a small site?

In marketing, a content audit usually means an inventory. Every page goes into a spreadsheet with traffic numbers, and each one gets a verdict: keep, update, or delete. That is the right job for a blog with hundreds of posts, or for a site where search visits are the main goal.

This guide is the smaller job. You have a handful of public pages, and you want to know whether the words do their work. If your site is large, do the inventory first, then come back and run this method on the pages that matter most.

What does Clarity and Promise ask of your words?

Our audit scores every public site on the same six areas, and the first one is about words. Clarity and Promise asks whether a first-time visitor could say what this is after reading your headline, and whether your other pages tell the same story. We put it first because the other areas lean on it: a tidy mobile layout and a specific button do little if nobody can tell what the product is.

In our sample report, an audit of ismyappslop.com dated 28 September 2026, Clarity and Promise scored 98, and "What works" credits a headline and subheadline that say what the product does, who it is for, and the value. "Slop tells" for that area reports none observed.

For outside support, the US government's plain-language guide calls writing for your audience its first rule (Digital.gov, page published 11 September 2025, read 5 October 2026). Google's self-assessment questions for content ask whether a page leaves readers feeling they need to search again for better information (Google Search Central, page updated 5 October 2026, read 5 October 2026). That page is about search content, but the question travels: would a visitor leave to look for a plainer answer?

How does the read-aloud and mark-up pass work?

Your eyes skim past empty phrases. Your mouth does not. Reading aloud slows you down enough to hear the filler.

1. Strip the design. Copy the words of one page into a plain document. Start with the homepage, then pricing, signup, and about.
2. Read it aloud. Use a normal speaking pace, as if you were explaining the product to a friend across a table. Do not fix anything yet.
3. Mark as you go. Use 4 marks:

• Underline any sentence where you stumble, run out of breath, or reread.
• Circle any claim with nothing behind it, such as "trusted," "fast," or "powerful" with no number, example, or name.
• Box any word you would never say to a friend, such as "leverage," "seamless," or "next-generation."
• Draw a wavy line under any sentence you could drop on a different site, word for word.

1. Count the marks per page. The page with the most marks gets fixed first.
2. Finish the sentence. Say your headline, then say "This is for" and finish it in one breath. If you need "and also," the headline is carrying too many ideas.

What is the swap-the-name test?

The wavy line is a guess. This test checks it. Take your headline, your subheadline, and your first 3 feature lines. Swap your product name for another product's name, or swap your category word (receipts for invoices, recipes for workouts). If the line is still true and still reads naturally, it is generic.

Specific lines break when you swap. They name your input, your outcome, or your person, and those things do not transfer to another product. Our homepage FAQ lists the vague headline among its examples of slop, and this test is a quick way to find yours.

What does vague copy look like, and how do you fix it?

Hypothetical example: a small app turns photos of receipts into an expense report for freelancers. The left column is a first draft. The right column is a rewrite to test, not proof of better results.

The draft headline fails the swap test: "supercharge your workflow" fits a calendar, a notes app, or a chat tool. It also earns a box around "next-generation" and a wavy line under the whole sentence. The rewrite names the input (receipts), the outcome (an expense report), and the person (freelancers). Swap in another product and it stops being true.

Do your other pages agree with the headline?

Clarity and Promise has a second half: does the rest of the site agree? Read your pricing, signup, about, and footer text aloud with the same 4 marks, and hold each against the homepage headline. Watch for a page that speaks to a different audience, a page that describes a different promise, and leftover template text or placeholder names nobody replaced. Each one makes a visitor wonder whether they landed in the right place. Fix the page that is out of step, and keep the headline as the anchor.

What if an AI tool wrote the words?

Run the swap-the-name test first, before the read-aloud. Drafted copy often reads smoothly, so it can pass a read-aloud without a single stumble and still fit a thousand products. The swap test does not care who wrote the line.

For every line that survives the swap, add one detail only you hold: your real input, your real price, or your real outcome. Then read it aloud to make sure it still sounds like you.

Does the next step read clearly?

Read each button and link as if it were your only clue, and check that the label says what happens when you press it. Button wording, forms, and the full path to a purchase are covered in our landing page audit guide.

Can an audit do this for you?

Part of it. The report scores Clarity and Promise on its own line, lists what works and any slop tells, ranks the fixes, and puts a verdict at the top, so you get an outside read on your headline and on whether the rest of the site agrees with it. For friction in navigation, forms, and mobile use, see our website UX audit guide.

What no report can know is what you meant to say. That is why the read-aloud pass comes first: it gives you your own marks, and the report becomes a second opinion on them.

Which marked lines do you rewrite first?

For each mark that matters, write 3 short things: what a visitor reads, why it may confuse them, and what you will change. Then work in this order:

1. The headline and the one-sentence promise, since every other page leans on them.
2. Pages that disagree with the headline.
3. Button labels and next steps.
4. Everything else, such as tone and polish.

Words you wrote yesterday sound different today, so read the new version aloud again before you publish it.

What is a good first step today?

Give it 10 minutes. Take the headline, the subheadline, and the first 3 feature lines on your homepage. Read them aloud, make the 4 marks, and run the swap-the-name test. Rewrite only the headline, and read it aloud again.