The shift is already measurable

66% of UK senior decision-makers now use ChatGPT, Copilot, or Perplexity when researching suppliers. That number comes from survey data published in 2024. It is not a prediction. It is current behaviour.

The generational split makes the direction clear. 85% of buyers aged 25 to 34 use AI for supplier research. Among those over 55, it is 23% (Source: Sopro). As younger professionals move into purchasing roles, AI-assisted procurement becomes the default.

How buyers actually use AI in procurement

Buyers are not typing brand names into ChatGPT. They are describing problems and asking for solutions.

A typical query looks like this: "Which construction consultancies in the Midlands specialise in pre-construction planning?" Or: "What are the best AI training providers for professional services firms?"

AI platforms respond with a shortlist. Usually three to five companies. Sometimes with brief explanations of why each was recommended. The buyer reads the list, clicks through to a website or two, and makes contact.

The entire top-of-funnel process that used to take days of searching, reading reviews, and asking contacts now happens in a single conversation.

Where AI gets its recommendations

AI platforms do not simply repackage Google results. Research from Profound found that 90% of ChatGPT citations come from pages outside Google's top 20. The content AI chooses to reference is often completely different from what ranks highest on Google.

Each platform also has its own criteria. Only 11% of websites are cited by both ChatGPT and Perplexity (Source: Profound). Being visible on one does not guarantee visibility on another.

AI platforms pull from multiple signals to decide who to recommend:

• 01.
  Content clarityDoes your website clearly explain what you do, who you serve, and what makes you different? AI needs to understand your business well enough to recommend it confidently.
• 02.
  Third-party validationAre you mentioned on other websites, in industry directories, in news coverage, or in customer reviews? AI weighs external mentions heavily when deciding who is credible.
• 03.
  Structured informationSchema markup, clean site architecture, and consistent business information across the web help AI platforms parse and trust your content.
• 04.
  RecencyAI platforms favour up-to-date content. A website that has not been updated in two years signals that the business may not be active.

Why your Google ranking does not predict your AI visibility

Most B2B companies assume that good SEO means they will show up in AI search too. The data says otherwise.

Google ranks pages based on backlinks, keyword relevance, and technical signals. AI platforms recommend businesses based on how well they understand what that business does and whether they trust it enough to cite.

These are different systems with different logic. A company ranking first on Google for "construction project management software" might not appear in any AI recommendation for the same query.

47% of brands have no strategy for how AI represents them. They are optimising for Google but ignoring the platforms that an increasing share of buyers use first.

What this means for different industries

The impact varies by sector, but the pattern is consistent.

Construction and engineering. Procurement teams are using AI to shortlist subcontractors, consultants, and material suppliers. Companies with clear service descriptions and project case studies get cited more often.

Professional services. Accountancies, law firms, and consultancies are seeing AI become a first touchpoint for clients comparing providers. Firms with specific expertise pages outperform generalist websites.

Manufacturing. Buyers researching components, equipment, or contract manufacturers increasingly ask AI before contacting sales teams. Technical specifications and certifications help AI platforms make confident recommendations.

Education and training. Organisations researching training providers and edtech solutions use AI to compare options. Detailed programme descriptions and outcome data improve citation rates.

How to check your current AI visibility

You can test this yourself in about ten minutes.

Open ChatGPT, Perplexity, and Gemini. Ask each one a question that a buyer in your market would ask. Something like: "What are the best [your service] providers in [your region]?"

Note whether your company appears. Check if it appears accurately. Check whether competitors show up instead.

Do this across several different query variations. The results will tell you where you stand.

Most companies that run this test for the first time find they are either absent entirely or described inaccurately. Both are fixable. But you need to know the starting point.

What you can do about it

Improving AI visibility is not a single tactic. It is a set of practices that help AI platforms understand and trust your business.

• →
  Audit your current state.Test your visibility across multiple AI platforms with real buyer queries. Document where you appear, where you are absent, and where competitors are being recommended instead.
• →
  Make your content AI-readable.Structure your website so each page answers a clear question. Use descriptive headings. Write in plain language. Add schema markup so AI crawlers can parse your information.
• →
  Build third-party signals.Get mentioned in industry publications, directories, and partner websites. Consistent information across multiple sources reinforces your credibility with AI platforms.
• →
  Keep your content current.Regular updates signal that your business is active and your information is reliable. AI platforms weigh recency when deciding who to recommend.
• →
  Allow AI crawlers access.Some websites block AI crawlers without realising it. Check your robots.txt to ensure GPTBot, ClaudeBot, and other AI crawlers can reach your content.

The companies acting now have an advantage

AI search visibility is still a new discipline. Most B2B companies have not started working on it. The companies that do start now face less competition and build authority faster.

The parallel to early SEO is useful. The companies that understood Google in 2005 built positions that took competitors years to match. The same dynamic is playing out with AI search.

The difference is speed. AI adoption among buyers is moving faster than Google adoption did. The window to build an early advantage is shorter.