- B2B buyers do most of their research before they speak to anyone, on Google, on LinkedIn and increasingly in ChatGPT and Perplexity.
- For B2B and SaaS, SEO and AI visibility are about being the credible answer at that stage, not about traffic volume.
- Our six steps: map the buyer's questions, make the site pass the "are they legit?" test, build pages per use case, make your brand unambiguous to AI, write for the sponsor, and measure demos rather than visits.
- Proof from our UK work: Unmind (an SEO-safe replatform), Sova (an enterprise-grade, fast site) and Wirehouse (a site we look after day to day).
For a UK B2B or SaaS company, SEO and AI visibility have one job: to make sure that when a buyer researches the problem you solve, on Google or in ChatGPT, Perplexity or Google's AI Overviews, they find you, trust you and book a demo. Traffic is a means to that, not the goal.
I co-founded Saigon Digital, and a large share of our work is with UK companies that sell to other businesses: software platforms, HR and employment services, assessment technology and specialist consultancies. This post explains how we approach search for that kind of company, and what we have learned from three of them: Unmind, Sova Assessment and Wirehouse Employer Services.
Why do B2B buyers decide before they book a demo?
Because most of the buying happens without you. 6sense's 2024 Buyer Experience Report found that B2B buyers are nearly 70% of the way through their purchasing process before they engage a seller, and that buyers, not sellers, make first contact more than 80% of the time.
In practice, that research looks like this:
- A problem search. Someone searches the problem in their own words, not your product category.
- An AI shortlist. They ask an AI assistant to compare options and name a few vendors.
- The legitimacy check. They check your website and LinkedIn to answer one question: "are they legit?"
- The internal pitch. They forward what they found to colleagues before anyone fills in a form.
If you are not visible and credible at each of those steps, you are not on the shortlist. That is true however good your demo is.
What makes search different for B2B and SaaS?
The usual SEO playbook was built for high-volume consumer searches. B2B is different in five ways, and each one changes how we work.
Search volumes are small, and each lead is worth a lot. A query that 50 people a month search can be worth more than one with 50,000 if those 50 are heads of procurement. We judge keywords by buyer intent and deal value, not by volume.
Sales cycles are long. Three months is quick in B2B, and enterprise deals can take over a year. Content has to serve every stage, from the first problem search to the security questionnaire.
The sponsor is not always the user. The person who signs is often a revenue owner, a head of operations or a head of AI strategy, not the team that will use the product. Pages need to answer the sponsor's questions about risk, payback and how quickly it goes live.
Precision matters more than personality. Buyers in financial services, legal and HR notice a vague or wrong claim immediately, so we write with subject-matter review and the client approves every piece.
AI engines can confuse you with someone else. If another company has a similar name, or your product name is a common word, ChatGPT may describe the wrong business. Fixing that is entity work, not keyword work (more on that below).
What did we do for Unmind?
Unmind is a UK workplace mental health platform that sells to large employers. When we met them, their marketing site ran on a Webflow setup that had become a bottleneck: the backend was slow, it was hard to build new page types, and the translation plugin was unreliable just as they were growing internationally.
We rebuilt the site on Next.js with Prismic, using Slice Machine so the content team could assemble pages from ready-made sections. Two things mattered most for search and demand:
- No search equity was lost in the move. We treated technical SEO and redirect mapping as part of the migration, not a clean-up job after launch, so Unmind's rankings came across intact. (Our replatforming guide covers this in detail.)
- The content team can publish without developers. That sounds like an operations win, and it is. It is also a demand-generation win: when marketing can launch a landing page for a new buyer persona, sector or campaign in an afternoon, demo campaigns stop waiting for a development sprint.
For a SaaS company, that second point is where demos come from. Every audience (HR directors, wellbeing leads, CFOs) and every use case deserves its own page that speaks their language and ends in a clear next step. A CMS your team actually uses is what makes that possible.
"Their team brought our vision to life with precision and creativity." George Stokes, Unmind
What did we do for Sova Assessment?
Sova Assessment is a UK HR technology company whose assessment platform helps employers hire and develop people. Their site sat on an ageing, monolithic Craft CMS build that looked dated and performed poorly.
For an enterprise software company, a slow site is a credibility problem. A buyer comparing three vendors quietly downgrades the one whose site feels sluggish, before anyone reaches a demo. Faster pages also rank better and lose fewer visitors.
We moved Sova to a Jamstack architecture (Next.js with Strapi), chose the CMS around what the content editors needed, and built a library of more than 30 reusable components. New pages are assembled from blocks that already look right and load fast, so the site keeps pace with the product.
"Their reliability, efficient turnaround time, and open communication allowed for a seamless and productive partnership." Shahzad Ahmed, Sova Assessment
What do we do for Wirehouse?
Wirehouse Employer Services is a UK firm that gives employers HR, employment law and health and safety support. We look after their website day to day: hosting and security patching, plus the compliance updates a regulated advice business needs, such as privacy notices, their AI notice and cookie consent.
Working closely with a specialist B2B firm like Wirehouse taught us something we now apply everywhere. The demand is usually there, but the business cannot see it. Employers search in their own words ("how do I handle gross misconduct", "do I need an HR consultant") rather than in the firm's service names. Mapping those searches, and seeing which pages appear for them and which do not, is the first step in our process for every B2B client.
How do we turn search into booked demos?
These are the six steps we follow with UK B2B and SaaS clients.
1. Map the buyer's questions, and size the market
We start with how buyers describe the problem, not how you describe the product. We pull the searches and AI prompts around that problem, group them by buyer and stage, and add up the monthly demand. That gives you a real number for the market you are competing for, and shows where competitors are winning it.
2. Make the site pass the "are they legit?" test
Before anyone books a demo, they check whether you are real, established and safe to buy from. That means case studies with named clients where you can, clear pricing or engagement models, security and compliance information, a visible team and leadership, and fast, error-free pages. This is the cheapest conversion improvement most B2B sites can make.
3. Build pages for each use case, sector and buyer
One "solutions" page cannot speak to a bank, an insurer and a pharmaceutical licensing team at once. We build focused pages per sector and per use case, each answering that buyer's questions and ending in one clear next step, usually a demo.
4. Make your brand unambiguous to AI
AI assistants build their answers from what they can verify about an entity: your name, what you do, who you serve and who else talks about you. We tighten this with consistent naming everywhere, Organization and Product schema (structured data), links between your site and your verified profiles, and content that states plainly what you are and are not. It is also how we fix the "ChatGPT thinks we are a different company" problem. Our entity SEO guide explains the method.
5. Write for the sponsor, with no filler
We publish articles that answer the questions sponsors and evaluators actually ask: comparisons, implementation timelines, risk, ROI and how you differ from the incumbents. They are written by our team, reviewed for accuracy, and approved by you before anything goes live. No generic AI-written padding.
6. Measure demos, not visits
We set up tracking so every demo request is recorded with its source, including visitors referred by AI assistants, which most analytics setups lump together with ordinary referrals. The monthly report leads with demo requests, branded and non-branded search, and AI visibility for your priority questions. Traffic comes after that.
What should a UK B2B or SaaS team measure?
If you only track four things, track these:
- Demo or enquiry requests by source. Organic search, AI referrals, LinkedIn and direct, recorded as key events in GA4 and matched in your CRM.
- Non-branded search visibility. Growth in people finding you without already knowing your name is the clearest sign SEO is working.
- AI visibility for your priority questions. Whether ChatGPT, Perplexity, Gemini and Google AI Overviews mention you, and describe you correctly, when asked the questions your buyers ask. See our guide to AI visibility metrics.
- Pipeline influenced. Deals where the buyer touched your content before talking to sales, even if they arrived by referral in the end.
For more on the SaaS side of keyword strategy, see our guides to SaaS SEO and B2B SEO.
Why work with a Vietnam-based team on UK search?
Because you get a senior, Western-led team that builds websites as well as optimising them, at a cost that makes sense for a growing B2B company. We are based in Ho Chi Minh City and work in UK hours for UK clients: our afternoon is your morning, so calls, reviews and fixes happen during your working day.
The other advantage is that one team owns the whole journey. The same people who write your content can fix the page it lands on, the form behind the demo button and the tracking that tells you whether it worked.
If you want to see where you stand, our AI visibility audit shows how Google and the main AI assistants describe your business today, or you can talk to us directly.
Frequently asked questions
How long does SEO take to produce demo requests for a B2B company?
Expect compounding results rather than instant ones. Technical and conversion fixes can lift enquiries within weeks, because they help the visitors you already have. New content usually takes several months to rank and be picked up by AI assistants. Our SEO results timeline explains what typically moves when.
Do we need a new website first?
Usually not. We start with the site you have, fix what holds it back and add the pages buyers need. A rebuild only makes sense when the platform itself is the bottleneck, as it was for Unmind and Sova.
Can you write for regulated or technical sectors such as financial services?
Yes. We research each topic properly, write precisely, and build in review by your subject-matter experts. You approve every piece before it is published, so nothing goes live that your compliance or product team would not stand behind.
What if ChatGPT confuses us with a company that has a similar name?
It is a common problem and a fixable one. We make your brand's identity explicit and consistent across your site, structured data and third-party profiles, then test the AI assistants regularly to confirm they describe you correctly.
Want to be the brand AI engines recommend?
We make businesses visible on Google and inside AI answers, measured in leads, not vanity metrics.
![Hotel Local SEO: How to Rank a Hotel for "Hotels Near [Location]" Searches](https://d1yy38tw7zsl9z.cloudfront.net/blog/nov-01.png?w=640&h=400&fit=crop&auto=format&q=75)
