A few weeks ago, we published an honest account of what we found when we audited bravand.com against the same framework we use for client work. The short version: a website that worked fine as a brochure, but wasn't working nearly hard enough as a business development tool. Good scores in some areas. Real problems in others. No softening.
That article got a decent response - and a recurring question: so what are you actually doing about it?
This is that article.
First, a thing worth saying out loud
We're a digital agency. Which might make you assume that fixing our own website is just a matter of someone turning round from their desk and getting on with it.
It isn't quite like that.
Bravand is a micro business. We have clients, projects, deadlines, and cash flow to manage - the same as any small organisation. Paid work comes first. Internal improvement work gets scheduled, phased, and budgeted like anything else. The cobbler's children are not automatically well-shod.
We're telling you this because it's relevant to you. If *we* have to plan and prioritise this work rather than just clicking our fingers, so will you. That's not a reason not to do it. It's just a reason to approach it sensibly - which is exactly what we've been doing.
What we've done so far
The audit gave us a prioritised list of recommendations. We've worked through the high-priority items over the past few weeks, fitting it around client work. Here's what's been done.
The big one: that team photo.
The largest single asset on our site was a PNG of the Bravand team - 757KB. To put that in context, it accounted for roughly 28% of our entire homepage's weight. We compressed it and converted it to WebP. It came down to 115KB. One file, one tool (Squoosh - free, browser-based), roughly an hour of work including re-uploading and testing.
We did the same with two background images while we were at it.
The content and SEO basics.
We rewrote the homepage meta title and description - both had been essentially unchanged for years and weren't doing anything useful for search. We fixed the page titles on three pages that were missing the Bravand brand name entirely. We updated the footer copyright date, which had been frozen at 2020 (it is, in case you hadn't noticed, 2026). We fixed the wrong company names in the alt text on our own 'Trusted by' section - an embarrassing one that had been sitting there unnoticed.
The case study content.
This one grew. While we were in the site fixing images and metadata, we took the opportunity to properly overhaul the written content on our case study pages - adding proper heading structure, rewriting some that were essentially stubs, adding calls to action. That took considerably longer than the technical fixes, but it needed doing and the timing made sense.
The navigation.
The audit flagged that our services dropdown had over 30 items in a flat list with no grouping or logic. We've now clustered these into five clear categories. The full services landing page is still to come - that's part of the sprint - but the navigation itself is already cleaner and more useful.
What hit a wall
One item on the list turned out to be genuinely unfixable, at least for now.
The audit identified approximately 2.5 seconds of savings available from addressing render-blocking requests on mobile. We investigated. It's a Webflow architectural issue - not something we can change from within the platform. We've noted it, accepted it, and moved on. Sometimes the answer is "not with the current setup."
Where we are now
When we started this process, our mobile PageSpeed score was 42 out of 100 - in Google's "poor" band. By the time we published the audit article in May, work done over the previous months had already moved it to 71.
After the fixes above, it's now sitting at 91.
Desktop has gone from 92 to 99.
Those aren't vanity numbers. Mobile page speed is a ranking signal. A slow mobile site loses visitors before they've read a word - research consistently shows more than half of mobile users abandon a page that takes longer than three seconds to load. Moving from 42 to 91 on mobile is a meaningful commercial improvement, not just a technical one.
One caveat worth knowing: PageSpeed scores fluctuate between runs. We re-ran the test a couple of hours after the 91 and got 78. Nothing had changed on the site - it's just how the tool works. Network conditions, server load, and test environment all affect the result. The scores that don't fluctuate (Total Blocking Time, Accessibility, Best Practices, SEO) are all consistently strong. The performance score is best read as a direction of travel, not a fixed number - and the direction, from 42 to consistently in the 70s and 90s, is clearly upward.
What's still to do
A chunk of the remaining work requires design, development, and SEO expertise. We've been honest with ourselves about that, and we've planned a dedicated sprint for July.
That sprint will cover:
- Alt text on all 25 team portrait photos (needs dev)
- Image SEO on the Our Work thumbnails
- A proper Services landing page
- AEO setup - FAQ content, schema markup, structured data - so we start showing up properly in AI-generated search results
- Accessibility fixes flagged by PageSpeed (colour contrast, link labelling)
- Internal linking from blog posts to relevant service pages
- A soft conversion option beyond the contact form - likely a discovery call booking
There's also a /refer page to build, which we're doing ourselves. If you send someone our way and they go on to become a client, you get 10% of the first project value back - as cash, credit, or a charity donation. It deserves more than a mention buried mid-article, which is currently all it has.
So, should you bother doing this?
Here's what three weeks of work - fitting around client projects, in phases, with a further sprint planned - has produced:
- Mobile performance score: 42 → 91
- Homepage significantly lighter and faster to load
- SEO basics corrected across multiple pages
- Case study content properly structured and worth reading
- Navigation that makes more sense
- An active plan for the more complex work
A fair chunk of this was done without any specialist help. Free tools, a few focused hours, and a willingness to actually look at what was there.
The rest needs expertise. And knowing which things need expertise, and roughly what's involved, is itself valuable - because it means you're not going into a conversation with an agency (us or anyone else) blind.
That's what an audit gives you. Clarity on where you are, what matters, and what to do next.
What this looks like as a service
We're now offering the Website & Digital Audit as a standalone piece of work, from £1,500 + VAT. Fixed fee. No tie-in. The report is yours to act on however works for you - yourselves, another agency, or with our help.
The example report - the one we produced using our own site - is available to download if you want to see the format and depth before committing.
If you're wondering whether an audit makes sense for where you are right now, get in touch and we'll give you an honest answer either way.
Read the full series
Part 1: "Your website might be "good". But that might be the problem..."
Part 2: "We audited our own website. Here's what we found"
