Before a prospective client speaks to an architecture practice, they may already have formed a view of whether it is right for their project. They might have explored its website, looked up its directors on LinkedIn or asked an AI tool which practices do this kind of work – and even a personal recommendation usually prompts a check rather than replacing one.

What they find often tells them less than it should. Most practice websites show the building and leave out the decisions: strong photography, a project title, a location and perhaps a paragraph of description. The reasoning a client is weighing up – what the brief demanded, which constraints shaped the scheme and why one option won – is rarely visible.

That gap has a cost. A practice may have a strong reputation and still struggle to show it is ready for a more demanding brief, a larger commission or work beyond the projects it is already known for. In a tight market, with RIBA’s Workload Index barely positive, the shortlist does more of the filtering, and a practice that is hard to assess from the outside is easier to leave off.

This piece examines what prospective clients are trying to find out before they make contact, why the expertise behind the work so often stays hidden, and how practices can make their thinking visible without adding to their directors’ workload.

AI is changing how prospective clients research complex services, unfamiliar technologies and potential suppliers. A buyer may still begin with Google, a referral, a LinkedIn post or an industry recommendation – but they may now also use ChatGPT, Perplexity or Google's AI-enhanced search experiences to clarify a problem, compare approaches and decide which firms are worth investigating further.

For expert-led B2B businesses, this does not mean chasing a shortcut into every AI-generated answer. It means making their expertise, offer and evidence clear enough for the right people to find, understand and trust. A mention in an AI response is not a business outcome if the website leaves a prospective client uncertain about what the firm does or why they should make contact.

Measuring this is harder than it looks. Referral reports show only the visits that can be traced back to an AI tool, and the same question can produce a different shortlist from one run to the next. A fairer picture comes from several measures read together and judged against qualified enquiries.

This piece examines why an AI visibility scan cannot diagnose the real issue on its own, what a more useful review should investigate, and how to measure AI search visibility without relying on referral traffic alone.

Most B2B websites are built for the moment they launch: a homepage, a few service pages and a contact form. They may look credible and do the job initially. But when marketing needs to launch a podcast, build a sector page, publish a news hub or support a new campaign, the site can quickly become a constraint rather than an asset.

Each new idea can mean re-engaging a developer, briefing a new template, waiting for a quote and securing budget approval before anyone can act. The cost is not only financial. Campaigns lose momentum, competitors publish first, and teams stop proposing ideas because they have learned the website cannot easily support them.

This is not an argument against bespoke development. Some businesses have genuinely complex requirements that call for it. But many primarily need a website that works as an active marketing tool – helping the right people understand and trust the business, while giving the internal team the flexibility to develop the marketing that reaches them.

This piece examines what makes a good B2B website, why launch-focused builds so often fall short, and how to create a site that can keep supporting your marketing as the business evolves.