Found by AI is a three-part Velo series on how B2B buyers now use AI to choose suppliers. Each sector is found in a different way. Manufacturers have to unlock what they already know. Technology firms have to prove what they claim. Professional services firms have to earn the right to be cited.
Technology buyers have adopted AI faster than anyone. Forrester found that 94% of business buyers use AI during their buying process. Gartner reports that 60% of B2B buyers use GenAI in some form to shape their purchase decisions.
They don’t trust it blindly, though. 51% of buyers say they’re more likely to encounter misleading information from GenAI, and 49% say the same of sales reps.
This combination defines tech marketing today. Buyers use AI to build the shortlist, then check everything it tells them. The vendors that win are the ones that pass those checks.
Check one: “Is this real?”
Every vendor now calls itself AI-powered. AI engines see thousands of identical claims and treat them as noise, so they look for specific, checkable facts. A buyer asking “which field service tools integrate with Xero?” gets names drawn from integration pages, help pages, review sites and community threads. Your brand campaign plays no part.
Check two: “Will it work for us?”
Buyers expect to see the product working. More than 60% of business buyers use a trial, rising to 78% for purchases of $10 million or more. A trial doesn’t close the deal, though: Forrester found just over a third of trial users go on to buy the full paid version from the same provider.
Buyers compare before they trial and keep comparing during it. Content that answers “does it handle our setup?” earns you the trial, then helps turn it into a sale.
Check three: “Who else says so?”
Buyers look for other opinions. They use an average of seven information sources during a purchase. McKinsey counts ten channels across the journey, with 42% of buyers using more than 11 touchpoints.
Forrester found buyers check AI-generated information with peers, product experts and analysts. 69% of buyers also prefer to validate AI-generated insights with a sales rep. A customer’s review or a named case study carries more weight than anything you say about yourself.
All of this happens early. 68% of buyers have a front-runner before they begin, and that vendor wins 80% of the time. With around 95% of the market out of cycle at any one time, the real contest is to be the front-runner when a trigger fires. Typical triggers are:
- An outage at the current supplier
- A new use case
- A cost review at renewal
- A new security or data regulation
- An innovation that changes the business case
Steps to influence your appearance in LLMs: readable, answerable, verifiable
Make it readable (architecture)
- Move your help centre and documentation out from behind logins where you can. They’re some of the most-cited sources in software answers.
- Check how your pages load. Many AI crawlers don’t run JavaScript, so content that appears only after scripts load may be invisible to them.
- Add structured data for FAQs and software products. Fix crawl errors, speed and mobile issues.
Make it answerable (on-site content)
- Build your FAQ from sales calls and support tickets, the most accurate record of what buyers ask.
- Publish one page per integration, covering what connects, how and any limits.
- Write fair comparison and alternatives pages before a competitor defines you.
- Make pricing as clear as your model allows.
- Keep changelogs and release notes current, since freshness signals relevance.
- The company website is still one of the three most-used B2B touchpoints, alongside in-person sales and video calls, so it has to answer the checking questions directly.
Make it verifiable (off-site signals)
- Structure every customer story the same way: the problem, the setup and the result in numbers. Encourage customers to tell it in their own channels too.
- Run an active review programme on G2, Capterra and the platforms your buyers use.
- Take part honestly in the communities where buyers ask each other for advice.
- Keep partner marketplace listings complete, and invest in analyst and customer advocacy relationships.
Your first 90 days
- Days 1 to 30: Get access to Google and Bing Search Console, then review the errors it reports. Run a prompt analysis: ask ChatGPT, Perplexity and Google’s AI Mode the comparison and checking questions your sales team hears, and record who gets named and which sources are cited. Fix titles, headings, redirects and duplicate content. Open up your documentation and publish an FAQ from answers you already have. Expect errors to fall.
- Days 31 to 60: Build customer stories with numbers, integration pages and comparison pages. Fix JavaScript rendering problems and pages with high bounce rates. Expect rankings and traffic on key terms to rise.
- Days 61 to 90: Launch the review and advocacy programme and form a digital PR and citation plan. Rerun the prompt analysis and compare. Expect your AI appearance to grow.
Prove what you claim
AI visibility doesn’t replace your sales team; it decides which vendors reach a sales conversation in the first place. The tech firms that get recommended are those whose documentation, customers and reviewers all back up the same specific story.
Also in this series: In Part 1, manufacturers learn how to unlock specification data AI can’t currently read. In Part 3, professional services firms learn why only genuine thought leadership gets cited.
Want to know whether AI recommends you? Velo can run the prompt analysis on your market and show you who gets named, and why. Click here to request a prompt analysis report.