We Tested AI Search Across 10 Industries — Here's What Actually Gets You Recommended

AI heavily weights TripAdvisor ratings and booking platform reviews (Booking.com, Expedia) when recommending hotels. Hotels that appeared in AI results almost always had 4.5+ ratings on at least two major travel platforms. Google Reviews alone were not sufficient — AI cross- referenced travel-specific sources.

Key signal: Hotels mentioned in editorial travel content ("best boutique hotels in..." articles from travel bloggers or publications) had a dramatically higher recommendation rate. A single feature in a curated "best of" list often correlated with consistent AI recommendations.

Restaurants were the most competitive category. AI platforms showed a strong preference for restaurants that had been written about by food bloggers, local media, or guides (Michelin, Time Out, local "best of" lists). Review volume on Google and TripAdvisor mattered, but editorial mentions were the clear differentiator.

Key signal: As we documented in our Lyon restaurant study, niche specialisation matters enormously. A restaurant known as "the best bouchon in Lyon" gets recommended more reliably than a generalist with higher overall ratings. AI loves clear positioning.

Google Reviews and local directory listings dominate for dental practices. AI recommendations leaned heavily on Google Business Profile completeness (services listed, photos, Q&A section filled out) and presence in healthcare-specific directories (NHS Choices, Dentistry.co.uk, local dental associations).

Key signal: Dentists whose websites had dedicated pages answering common patient questions ("How much do dental implants cost?", "Is Invisalign right for me?") were recommended far more often than those with generic brochure-style sites. AI matched patient questions to practices that had already answered them.

For law firms, AI showed a strong bias toward specialisation. General practice firms were almost never recommended. Instead, AI consistently surfaced firms that clearly positioned themselves in a specific area: employment law, family law, commercial property. Legal directory presence (Chambers, Legal 500, Avvo) served as a major trust signal.

Key signal: Published thought-leadership content (blog posts, articles explaining legal concepts in plain English) correlated strongly with AI recommendations. Firms that wrote about the topics people actually search for ("What are my rights if I'm made redundant?") were the ones AI pointed users to.

Real estate was dominated by agents with strong local content. AI recommended agents who had published neighbourhood guides, market reports, and area-specific advice. Presence on Rightmove or Zoopla alone was not enough — the AI needed signals beyond listing portals to understand who the trusted local experts were.

Key signal: Google Reviews were important, but what set recommended agents apart was community-level content. An agent with a blog post titled "Moving to Chorlton: Everything You Need to Know" consistently outperformed an agent with more reviews but no local content.

Beauty salons presented a unique pattern: Instagram and social media presence played a larger role than in most other industries. AI platforms referenced salons with active social profiles alongside strong Google and Treatwell/Fresha reviews. Visual platforms seemed to feed into AI's understanding of salon quality and relevance.

Key signal: Salons that were featured in local "best of" articles or beauty blogger roundups had a notable edge. Those with booking platform presence (Treatwell, Fresha) in addition to Google were recommended more consistently than those relying on a single review source.

Wedding venues showed the highest correlation with editorial mentions of any industry we tested. AI consistently recommended venues that appeared in wedding blogs, bridal magazines, and curated venue lists (Hitched, Bridebook, WeddingWire). Review platforms mattered, but the editorial layer was decisive.

Key signal: Venues with detailed, structured websites (clear pricing tiers, capacity information, photo galleries with descriptions, FAQ pages) were recommended over venues with beautiful but information-sparse sites. AI needs facts to work with, not just aesthetics.

Auto dealers were one of the most review-dependent categories. AI platforms placed enormous weight on Google Reviews volume and recency, as well as presence on AutoTrader and specialist review sites. Dealers with fewer than 50 Google reviews were rarely recommended, regardless of their rating.

Key signal: Dealers who specialised ("independent BMW specialist in Manchester") outperformed large generalist dealerships in AI recommendations for specific queries. AI matched niche queries to niche expertise. Transparent pricing and financing information on the website also correlated with higher recommendation rates.

For accountants, AI recommendations were driven by a combination of Google Reviews, professional directory listings (ICAEW, ACCA, local chamber of commerce), and website content quality. Firms with blog content covering practical tax and financial topics were favoured significantly.

Key signal: Accountancy firms that clearly stated who they serve ("small business accountants in Birmingham" or "tax advisors for freelancers and contractors") were recommended over firms with vague positioning. Specificity was the differentiator: AI matched user intent to clearly-defined services.

Home renovation was the industry where trust directories mattered most. AI heavily referenced Checkatrade, MyBuilder, Bark, and Houzz alongside Google Reviews. Contractors listed on multiple trade platforms with verified reviews were recommended far more often than those relying solely on their own website and Google presence.

Key signal: Contractors who published project case studies, cost guides ("How much does a loft conversion cost in London?"), and before/after galleries on their websites were consistently recommended. AI wants content that answers the exact questions homeowners are asking — and rewards businesses that provide it.

Every single industry showed the same trend: businesses recommended by AI had reviews on multiple platforms, not just one. Hotels on TripAdvisor and Booking.com. Dentists on Google and NHS directories. Contractors on Checkatrade and MyBuilder. AI treats multi-source validation as a trust signal — one platform is an opinion, three platforms is a consensus.

This aligns with what we've observed in our broader research on AI visibility signals: single-platform businesses are nearly invisible to AI search. Diversification is not optional.

Across all 10 industries, businesses that had been written about by a third party — a blogger, a journalist, a local publication, a "best of" list, an industry roundup — were recommended at a significantly higher rate than those with no editorial footprint. This was true regardless of the size of the publication.

In AI search, mentions are the new backlinks. A local food blogger writing "this is the best Thai restaurant in Camden" carries real weight. AI models absorb these mentions during training and use them as evidence that a business is worth recommending.

This was perhaps the strongest and most actionable pattern. In every industry, recommended businesses had websites that addressed real customer questions — not just marketing copy. Dentists explaining implant costs. Contractors publishing renovation timelines. Hotels detailing what's within walking distance. Accountants breaking down tax deadlines.

This makes perfect sense when you understand how AI search works. When someone asks "What's the best dentist for implants near me?", the AI looks for content that directly addresses that question. A business that has already published an answer to that question on their website is the obvious choice. As we've explored in our comparison of SEO and AI visibility, this is a fundamental difference from traditional search: AI recommends entities, not just web pages.