The problem with your English homepage is almost never your English.
It is that you wrote it in the register of a tender document, and a tender document is the one kind of English that is designed not to persuade anybody.
This is why proofreading does not fix it. A proofreader corrects grammar, and the grammar is usually fine. What is wrong is pitch: too formal, too abstract, too careful, written in the voice of an institution rather than a person who knows something.
The good news is that this is a much smaller problem than a language problem. There are four specific tells, they are all repairable in an afternoon, and the instinct producing them is one that the research says costs you credibility rather than buying it.
The research that should settle the instinct
Most people writing in a second language raise the formality on purpose. It feels safer, more professional, more likely to be taken seriously.
Daniel Oppenheimer tested that, in a paper with a title that makes the point by example: Consequences of erudite vernacular utilized irrespective of necessity, published in Applied Cognitive Psychology in 2006. He took texts and produced more complex versions by swapping words for longer synonyms, then asked readers to judge the author.
Needless complexity made authors look less intelligent, not more. The mechanism he identifies is processing fluency: text that is harder to read makes the reader think worse of whoever wrote it. He also noted that most of the students surveyed admitted deliberately inflating their vocabulary to seem clever.
One honest caveat. The participants were undergraduates assessing writing samples, not procurement managers assessing a vendor. So I would not claim the effect size transfers to your pricing page. The mechanism is the part I would bet on, because processing fluency is not a property of students.
Which means the formality you added to be taken seriously is, on the available evidence, working against you.
The four tells
Formality calibrated too high. Business English learned from contracts, reports and EU documents rather than from conversation. It produces sentences like “Our solution enables organisations to optimise their procurement processes.” Nobody has said that out loud.
Abstract nouns where a verb belongs. This is the most common and the most damaging. “The implementation of our platform results in a reduction of manual data entry” instead of “our platform cuts manual data entry”. The information is identical, the second version is half the length, and only the second one sounds like somebody who has actually done it.
A feature list standing in for a claim. Three bullet points describing what the product has, with no sentence saying what it does for whom. This is not a language failure at all, it is a positioning failure that hides inside careful English, and it is the one that costs the most.
Translated idiom. Phrases that are correct and land oddly, usually because they are a direct carry-over. These are hard to self-diagnose, which is the one genuine argument for a native speaker, and I will come back to that.
Notice that none of these is an error. All four would pass a grammar check, which is exactly why they survive.
A published list of the words to drop
You do not need to develop taste for this from scratch. The British government publishes its style guide, including a list of words it tells its own writers to avoid, with the replacement for each.
A selection, because these are the exact words that appear in Estonian B2B English:
utilise, use “use”. deliver, when applied to something abstract, use “make”, “create” or “provide”, since the guide points out that pizzas and post are delivered and improvements are not. facilitate, say specifically how you are helping. leverage, use “use” or “influence”. incentivise, use “encourage”. liaise, use “work with”. streamline, use “simplify”. transform, describe what you are actually changing. in order to, usually not needed at all.
It also tells writers to avoid metaphors, on the grounds that they slow comprehension, with “drive” singled out: you can drive a vehicle, not a scheme.
Two things make this list more useful than a style opinion. It is written for a readership that cannot opt out, so it has been tested on the least forgiving audience there is. And it is British English, which is the variety your European buyers expect anyway.
One important limit from the same guide, because the opposite error exists. Technical terms are not jargon. If your buyer’s job involves the word, use the word, and explain it the first time. Simplifying a technical term that your reader uses daily reads as condescension, and that is its own credibility problem.
Two tests that take ten minutes
Read it aloud. Not silently. Out loud, in English, to the wall. Every sentence you stumble over or run out of breath in is a sentence to rewrite, and the stumble is almost always a nominalisation or a clause that should have been a full stop. This test finds about eighty percent of the problem and requires no English skill beyond what you already have, because your ear in a second language is better than your eye.
The one-sentence claim. Cover your homepage and write, in one sentence, what you do and who for. If you cannot, the page cannot either, and no amount of editing will fix it. This is the test that catches the feature-list problem, and it is a thinking exercise rather than a language one.
A third, if you have an English-speaking customer: send them the page and ask what they think you sell. Their answer is your actual copy, whatever is on the screen.
When to hire a native speaker
Worth doing, and almost always for the wrong job.
Do not hire them to write volume. Writing volume requires knowing your buyer, your category and your product, which is knowledge you have and they do not, and a fluent page about the wrong thing is worse than an awkward page about the right thing.
Hire them for three narrow jobs. The headline and the one-sentence claim, where every word carries weight and idiom matters most. A read-through pass to catch the translated phrases you cannot hear. And the ten or so sentences a buyer reads before deciding, which is a far smaller surface than your whole site.
That is a few hours of work a quarter rather than a retainer, and it is the version that pays.
Before lunch
Forty minutes, on your homepage and nothing else.
Read it aloud and mark every sentence you stumble in. Rewrite each one as something you would say to a colleague who asked what your company does.
Then search the page for the words in that list above, starting with “solution”, “utilise”, “enable” and “deliver”, and replace each with the specific thing you actually mean.
Then write the one-sentence claim at the top. If it took you twenty minutes to write, that was not a copywriting problem, and it is the most valuable twenty minutes on this list.
If you’d like a second opinion on whether your English page is underselling what you do, drop me a line on email, WhatsApp, phone or LinkedIn and we can have a quick chat, and my notes on content marketing cover the rest. I’m taking on a small number of contracts at the moment, so I’ll tell you straight whether it’s something I could help with.