Every week, your website either wins you work or costs you work. In most engineering and manufacturing businesses it is quietly doing the second.

The losing is quiet, which is why it goes unnoticed for years. Nobody lands on an engineering website, reads two pages and thinks these people are terrible. The real failure is duller than that. A serious buyer with a live requirement reads three pages, learns nothing that helps them decide, and closes the tab. No complaint. No bounce anyone investigates. No record of the enquiry that did not happen. The business carries on believing its pipeline reflects the market, when it reflects the market minus everyone the website failed to convince.

This is the hardest kind of commercial loss to act on, because there is no evidence of it. You can count the enquiries that arrived. You cannot count the ones that were considered and abandoned on a Tuesday afternoon by a project manager you will never meet.

The root of it is a briefing problem rather than a design problem. Most technical websites were built to describe the company: here is who we are, here is what we make, here is our history and our accreditations. Buyers do not use them that way. They use them to decide whether you are worth a conversation. Those are two different jobs, and almost every site in the sector has been built for the first one.

A tidy website that does no commercial work is an expensive one

There is a comfortable assumption behind a lot of underperforming websites, and it goes like this: it may not be doing much, but it is not doing any harm. It looks tidy, it is up to date, it does not embarrass us in front of a client. That is the standard most technical businesses actually apply, and it sets the bar in the wrong place. The right comparison is what a competent site would have earned you over the same period.

Think about what you would tolerate from a salesperson. If you employed someone who attended every meeting, described the company accurately, answered nothing the buyer actually wanted to know, never asked what problem they were trying to solve, and never once suggested a next step, you would not describe that person as neutral. You would describe them as expensive. Most industrial websites behave exactly like that, at greater scale, in front of more buyers, every day of the year.

The cost turns up somewhere other than the build fee. It is the tender you were never invited to, the enquiry that went to the competitor whose site made the case more clearly, and the buyer who quietly concluded you work with people much larger or much smaller than them. None of that appears in a report. All of it appears in the pipeline as an absence.

Serious buyers pre-qualify you before you know they exist

The behaviour that makes this expensive is well documented. The 2026 State of Marketing to Engineers study, found that engineers spend 62% of the buying process on independent online research, and that vendor websites remain one of the two most important sources of information they use. Gartner’s finding that 67% of B2B buyers would prefer a rep-free buying experience points the same way. The evaluation is happening. You are simply not present for it.

This changes the status of the website. It has stopped being what a buyer checks after you have spoken to them and become the audition for the conversation. By the time somebody picks up the phone, they have already decided you are plausible, and they made that decision alone, against whatever you have published, in a few minutes, without asking you a single question.

The same applies to referrals, which is the objection we hear most often. Businesses with strong reputations tell us the website does not matter commercially because the work comes through recommendation. It is a reasonable inference and it is wrong. A recommendation ends with a name, and the person who receives that name looks the company up before they act on it. The website is the second step of the referral, and it is where a warm introduction either gains weight or quietly loses it.

There is a second buyer in this process now, and it does not get bored or give you the benefit of the doubt. A growing share of early research runs through AI assistants that read what you have published and compress it into a short summary for someone who may never visit the site at all. That summary is built from whatever is there. A site that names sectors, states outcomes and carries dated evidence produces a specific answer. A site of capability lists produces a sentence that could describe any firm in the country, or nothing at all. The thinness that used to cost you a visitor now costs you the summary the visitor reads instead.

A referral that arrives at a site full of stock photography and capability lists cools quietly. The buyer forms a vague impression of a competent, generic engineering firm, files it, and does nothing urgent about it. The person who made the recommendation never finds out. Neither do you.

Feature-led copy answers a question the buyer has not asked

The most common failure on a technical website is well-written content that lists what the business can do without ever addressing who it is for or what it fixes. Machine lists. Service pages. Accreditations. Tolerances. All accurate, all specific, all describing capability in the abstract, with nothing that connects any of it to a buyer’s situation.

In a mature sector the buyer already assumes you are broadly capable, a point we have argued at length elsewhere: technical excellence not a differentiator, it is the entry fee. Their question is narrower and more anxious than what can you do. Have these people solved a problem like mine, for a business like mine, and can I tell that from this page without having to ask?

An audit reported by SME Today in August 2026 looked at 100 UK engineering websites drawn from the Made in Britain directory. Fifty-eight of them used sector or industries language that signals who they serve. Forty-two did not. That is four in ten technical businesses giving a visiting buyer no way of knowing whether they are the right sort of customer. The methodology was a simple keyword check, so treat the exact figures as directional. The pattern will be familiar to anyone who has looked at fifty of these sites in a row. The firms explain what they make very well. They make it much harder to see who they help.

The purest example of this sits at the top of the homepage. The hero line is the single highest-leverage piece of content the business owns, because it is read by everybody and it sets the frame for everything underneath it. Most companies spend it on something that could sit above any competitor’s logo without looking out of place. Precision engineering excellence since 1978. Your trusted manufacturing partner. Quality, delivered. We have written before about the vocabulary the sector shares and how completely buyers have learned to ignore it – it’s just wall paper!. The hero line is where that costs the most, because it is where you have the reader’s full attention and roughly one sentence of it.

We run a version of the same exercise in almost every website review. We ask the MD to open their own homepage and read it aloud as though they were a buyer with a live problem and three competitor tabs already open. Most get four or five lines in before they stop, usually with some version of the same sentence: this does not actually say anything. The copy itself is rarely at fault. It went through several drafts and everyone signed it off. What nobody ever wrote down is what the page was supposed to make a stranger believe.

A hero line that does commercial work names the buyer, the problem, and the outcome, in that order, in plain language. Precision machined components for aerospace subcontractors who cannot afford a late delivery. Structural steel for main contractors working to fixed programme dates. Neither of those is elegant. Both of them qualify. A buyer reading either one knows within a second whether to continue, and a buyer who continues for the right reason is worth more than ten who scroll past out of politeness.

There is a simple test. Put a competitor’s name and logo above your homepage copy. If nothing about the page looks wrong, the page is not saying anything about you.

Your proof is the most powerful content you own and the most likely to be missing

Ask any buyer in engineering, manufacturing or construction what they most want to see from a supplier they do not know, and the answer is consistent. Evidence. A description of a comparable problem, what the supplier did about it, and what happened as a result. Case studies and testimonials are the closest thing a website has to a reference call, and they are available to a buyer at eleven o’clock at night without anyone having to ask you for permission.

They are also the content most consistently absent. In that same audit of 100 UK engineering sites, only 40 carried any visible reference to case studies. Six in ten of these businesses, all of them with decades of delivered work behind them, had put none of it where a buyer could find it.

The reasons are always the same and they are all understandable. Client confidentiality. Nobody has time to write them. The client would have to approve it. And underneath those, the quiet one that does most of the damage: inside the business, the work does not seem remarkable. It was a Tuesday. It was what you do. The commercial value of a delivered project is invisible from the inside precisely because it is routine, and it is the routine competence, described specifically, that a nervous buyer is looking for.

When case studies do exist, they are usually written as process descriptions. Three paragraphs about the approach, a photograph of the finished item, no statement of what the client was actually struggling with, no numbers, no named sector, no consequence. That is a record, not an argument. A case study earns its place when a buyer recognises their own situation in the first two lines. Everything else in it is supporting material.

Where the evidence sits matters almost as much as whether it exists. A lot of technical businesses file their strongest material under News, undated, alongside a Christmas closure notice and a photograph of a charity walk. Buyers looking for proof of a specific capability do not browse a news feed. Evidence needs to sit where the claim is made: the sector page carries the projects in that sector, the service page carries the work that used that service. A case study that a buyer has to hunt for is doing about as much commercial work as one that was never written.

The same principle applies to testimonials. Sixty words from a named person at a named business about a specific problem you solved will outperform twenty anonymous lines about professionalism and good communication. Buyers discount anything the supplier could have written themselves, and most testimonials on industrial websites read exactly as though they were.

Most of this is fixable in a few days of honest editing

None of this argues for a rebuild. Rebuilds are how technical businesses spend a lot of money and change very little, because the brief that produced the original site is usually the brief that produces the new one. The problem sits upstream of the platform and the design: nobody ever decided what the site was supposed to do commercially.

Take one page at a time and ask a single question of each: does this give a serious buyer confidence that we understand their problem and can solve it? The homepage first, because it is read the most. Then whichever service page attracts the most traffic. Then your evidence, or the fact that you do not have any published. Most technical websites can be made materially more effective in a few days of honest editing, without touching a line of code.

Your website is not a brochure. It is the first commercial conversation you have with every buyer who finds you, and it is the only one you are not present for.