Before/After: We Cut Our Own Homepage's Reading Level From Grade 12 to Grade 5

Our old homepage scored a Flesch-Kincaid reading grade of 12 to 17, depending on the section. That’s bar-exam territory for a page meant to tell a stranger what we do in five seconds. Sentences ran 17 to 27 words long. Two floating cards near the top claimed “+120% enquiries” and “4.9/5.0.” Neither number had a source. Three testimonials were credited to initials only.

Nobody set out to write it that way. A homepage gets built in pieces over a year — a stat added here, a feature announced there — and at some point it stops being a page and becomes a pile of everything the business wanted to say at once.

The rule we set

A 15-year-old should be able to read our homepage and know who we help, what we offer, and what problem it solves. Plain English, no agency jargon. We used a Lean Six Sigma frame to get there: define value from the customer’s problem, measure what actually matters, cut anything that doesn’t earn its place, and keep one source of truth for every fact on the page.

That meant translating our own vocabulary before we could translate anything for a reader. “AI automation” became “tasks that run by themselves.” “Client acquisition engine” became “a website that gets you customers.” “Lead qualification” became “finding out who’s ready to buy.”

What came out

The floating stat cards went first. Neither number had a source we could stand behind. The anonymous testimonials went too — a quote credited to initials looks made up, whether it is or not. So did the industries strip, the blog teaser, and a separate stats strip. What’s left states only what we can check: 17+ projects, clients in Nigeria and the UK, and 5.0 on Google from 5 reviews, verified the week we rewrote the page.

The first attempt was wrong

We’re not going to pretend the first version worked. The plain-language rewrite brought the homepage down to roughly 840 words at a grade 5 reading level. On paper, that was exactly the brief. In the browser, it looked thin: no visual anchor, no sense that real work happens here, nothing to hold a visitor’s eye before they’d read a single word. We’d fixed the reading level and broken the thing a homepage actually has to do — make someone want to stay on it.

The rebuild kept the plain language and added what the first draft was missing: real screen recordings of our own tools and client projects running in the browser, looping on the page instead of sitting there as a static screenshot, colour-blocked panels that walk through each problem we solve, and one clear way to start — a single input box, not a wall of choices.

What’s on the page now

Four problems, stated the way a customer would say them, not the way an agency would pitch them: people visit but don’t contact you, customers wait too long for a reply, nobody finds you on Google or in AI answers, the same office task eats an hour every day. Each one gets a plain-English fix, not a “contact us for a custom quote.”

Nothing on the new homepage claims a result we can’t point to. The rewrite is backed up in full — the old version and both drafts — so if a change doesn’t hold up, reverting it costs an afternoon, not a redesign.

Try it yourself

Read our homepage the way a stranger would — cold, no context, on your phone. If you can’t tell what we do, who it’s for, and roughly what it costs inside five seconds, tell us. That’s the test our old homepage failed. Get in touch if you want the same read run against yours.