I recently audited a website that cost around £15,000, part-funded through a business development grant. The owner didn’t come to me because anything looked broken. The site loaded. It took orders. To the human eye, it was fine.

I scored it 4 out of 10.

That number surprised the owner, and I understand why. Nothing on screen looked like a 4/10. But that’s exactly the problem, and it’s the reason I’m writing this article. The most expensive mistakes in this build were the ones nobody could see. Invisible to a person scrolling the site. Perfectly visible to a search engine crawling it.

This wasn’t a one-person failure

Here’s what makes this case worth writing about. This wasn’t a business owner who found a cheap freelancer on a marketplace. This was the proper route, done by the book: a development grant, a funding body, and a chain of approved providers drawn from that body’s own list. A business mentor, a marketing expert, and finally the web design agency itself. Every one of them funded. Every one of them recommended.

And every one of them signed off on a site that, underneath the surface, was actively working against the business it was supposed to grow.

That’s the uncomfortable part. The system worked exactly as designed. Money was allocated, experts were engaged, boxes were ticked, a website was delivered. What the system never included, at any point in the chain, was someone technical reading what had actually been built.

What I found underneath

I’ll keep the business anonymous, but the findings matter because I see versions of them constantly.

Links still pointing at the development domain. Parts of the site linked back to the agency’s staging URL. A visitor might never notice. A search engine absolutely does: it can crawl and index a duplicate copy of the site, splitting the authority the real domain should be earning. The business was, in effect, competing against a ghost version of itself.

Placeholder text live on a paid-for page. Published filler copy. Nobody in the entire chain of funded experts had read the site top to bottom.

A broken map embed on the contact page. One of the highest-intent pages a business has, and it didn’t work.

Product claims that created legal risk. The copy made therapeutic-style claims that can’t lawfully be made in a regulated category. The marketing expert didn’t catch it. The agency didn’t catch it. The owner had no idea it was even a category of risk.

Not one of these is visible from the owner’s chair. All of them are visible to the engines, the regulators, or both.

Script-kiddie web design

There’s a phrase in software for someone who runs powerful tools they don’t actually understand: a script kiddie. That’s the honest description of what happened here. Modern platforms and page builders make it easy to produce something that looks like a professional website. Templates handle the design. Plugins handle the features. What they don’t handle is the part that was always the actual job: launch discipline, technical hygiene, search fundamentals, and copy that won’t get a business in trouble.

At £15,000, funded and vetted, this business didn’t buy web development. It bought the script-kiddie version of it, assembly without understanding, and the difference doesn’t show up on screen. It shows up over the following years, as rankings that never build, enquiries that never grow, and a business owner who concludes that “websites don’t really work for us.”

That’s the damage. Not a broken site. A site that quietly holds a growing business back while everyone involved believes the job was done well.

What £15,000 should actually buy

If I’m reviewing a build at this budget, this is my minimum standard. These things should exist before the final invoice is paid:

A clean launch. No staging links, no placeholder text, no broken embeds, nothing indexed that shouldn’t be. Every page read, every link clicked, every form tested. This is baseline professionalism, not a premium extra.

Performance that holds up on a phone, evidenced against Google’s own tooling. Not “it feels fast on the office wifi.”

Search fundamentals done as work, not as a line item. Proper titles and descriptions, logical heading structure, the right pages indexable, redirects in place if it’s a redesign, and one canonical version of the site online, not two.

Copy that’s been reviewed for what it claims, not just how it reads. In any regulated space (skincare, food, health, finance, safety) someone has to read the copy with that hat on.

Ownership and access in the business’s name. Domain, hosting, analytics, admin. Not the agency’s.

Documentation and a handover. What was built, how to update it, and what happens after launch.

Notice what’s not on that list: features. Failed builds are rarely missing features. They’re missing finishing, and finishing is precisely the part that’s invisible to everyone except the engines.

If you’re spending grant money on a website

Scope the project independently of the funding amount. Decide what the business needs first. If it costs less than the grant, spend the difference on content and photography, where most sites are genuinely starved.

Build in a verification gate. Make the final payment conditional on an agreed technical checklist: performance evidenced, every page proofed, every link tested, one indexed domain, ownership transferred, handover delivered. A good agency won’t object. A bad one will reveal itself on the spot.

Get one independent technical review before sign-off. Not from the funding body’s list; that’s the chain that just failed. An hour or two of genuinely independent review costs a fraction of the build and catches exactly the category of invisible problem described above, while there’s still leverage to have it fixed.

The honest conclusion

Nobody in this story set out to do harm, and the business wasn’t scammed in any way a court would recognise. It got something harder to spot and harder to fix: a funded, vetted, expert-approved build that was doing quiet damage from the day it launched. Damage no one in the chain was equipped to see.

If you’ve had a site built recently, grant-aided or not, and it looks fine but performs like it’s dragging an anchor, the checklist above is where I’d start. And if you’d rather I looked at it, that’s exactly what my redesign audit is for.

Get the redesign audit → or book a discovery call.

On this page