How AI Search Engines Are Changing Purchase Decisions: What the Research Says

Purchase decisions are changing. We don't say that as hyperbole — it's a documented shift playing out in data, research studies, and the behavior of real customers.

Here's what the research says about how AI is changing the customer's path to purchase.

The research phase is moving into AI

The traditional customer journey started with identifying a need and moved on to a Google search. Today, an ever-larger share of the research phase happens in AI chatbots.

6sense (B2B Buyer Experience Report 2025, 4,000+ respondents): 94% of B2B buyers actively use AI tools when researching vendors. The adoption pace is dramatically higher than two years ago, when AI involvement in B2B purchasing was estimated at ~35%.

G2 (The Answer Economy, April 2026, 1,076 B2B decision makers): 51% of software buyers start their research in AI chatbots — up from 29% in April 2025. AI isn't a single touchpoint; it's a layer across the entire research process.

McKinsey Digital (2025): The number of professional AI queries has grown 7× over the past two years. The adoption curve is steeper than the rise of mobile search in 2010–2014.

How AI answers influence vendor selection

Bright Edge research (2025) examined how users work with AI recommendations when choosing products and services:

73% of users who received a specific recommendation from AI verified it further — but only with one or two sources, not across the whole market. AI recommendations drastically shorten the comparison phase.

68% of B2B buyers said they would consider a vendor recommended by AI even if they had never heard of it before — provided the AI described it positively and specifically.

Average shortlist length generated by an AI recommendation: 2.8 companies. Versus the average shortlist from a Google search: 5.4 companies. AI shortlists are shorter — but more deliberate.

The trustworthiness of AI recommendations

Edelman AI Trust Research (2025): 71% of professional users trust AI recommendations for work decisions more than they trust paid advertising. 54% trust them as much as or more than recommendations from colleagues — traditionally one of the most trusted B2B sources.

This is a key shift for GEO: AI recommendations carry a high perception of trustworthiness. A company that AI recommends as an authority acquires that trust implicitly.

Customers coming from AI convert better

The conversion data is consistent across platforms and industries:

Semrush (June 2025): AI-referred visitors convert 4.4× better than organic Google traffic. The specific numbers: 14.2% vs. 2.8% (Opollo, 2026 AI Search Benchmark Report, analysis of 312 B2B technology companies).

Klaviyo (e-commerce data, 2025): The average order value of AI-referred customers is 23% higher than average. The hypothesis: customers arriving from an AI recommendation are further along in their research and have higher intent.

B2B SaaS specifically: The average sales cycle length for AI-acquired leads is 18% shorter (HubSpot Sales Report, 2025). By handling the research phase, AI accelerates the customer's decision-making.

Demographic adoption patterns

Pew Research (2025): The average age of an AI user for professional research is 34. The highest adoption of AI for work purposes is in the 25–44 age group (78% use it regularly).

This is exactly the group that will make up the decision-making layer of most companies over the next 5–10 years. Their buying habits are setting the standard for the entire decade.

Adoption by industry: The highest AI adoption in the purchase process — technology (81%), financial services (74%), professional services (69%), healthcare (61%) (Gartner, 2025).

What this means for companies in practice

A summary of the key implications of the research for business strategy:

1. Shortlists form earlier. If you're not in AI answers, you're not on the shortlist for customers who started their research in AI. And in the tech segment, that's more than 50% of them.

2. AI trust is a real advantage. A company cited by AI as an authority gains a credibility advantage that converts into shorter sales cycles and higher order values.

3. Don't wait for mainstream adoption. The research shows that AI adoption in the purchase process is already mainstream in tech and professional services. Other industries follow 12–24 months later.

4. Have GEO data, not just intuition. Test AI answers to your category queries regularly. Track AI-referred traffic in your analytics. Have a baseline so you can see the trend.

Frequently asked questions

How many customers use AI in purchase decisions today? 94% of B2B buyers use LLM tools during the purchase process (6sense, 2025), and 51% of software buyers start their research in an AI chatbot more often than on Google (G2, 2026).

Do visitors coming from AI convert better? Yes — according to Semrush, they convert 4.4× better than visitors from organic Google search. They arrive later in the decision-making process, often already carrying a specific recommendation from the AI.

In which stage of the purchase does AI play the biggest role? In the research phase and when building the shortlist. Customers ask AI for 2–3 recommended options and only then visit the websites — whoever is missing from the recommendation never makes it into the consideration set at all.

What should a company do first? Establish the baseline: test your customers' queries in ChatGPT, Gemini, and Perplexity and find out who the AI recommends and why. A GEO audit addresses this systematically.

---

Want to know where your company stands in AI answers for your industry? A free GEO preview will give you concrete data — no strings attached.