Guide
Help AI search find and understand your practice
By Mark Gossman, Founder and Chief Strategist, Modern Harvester · Published September 30, 2026 · Updated October 31, 2026
Try this first. Open ChatGPT and ask the question your ideal patient would ask, with your town in it. Which practices does it name? Is yours one of them?
If not, it may not be anything you did. This guide covers the four areas to check first: access, patient-focused content, directory details and credible public information. Usually you don't need a new website.
No checklist guarantees an AI recommendation. Tools, questions and results vary.
Why this is worth reviewing
In BrightLocal's 2026 survey of 1,002 U.S. adults, 45% reported using AI tools for local business recommendations in the previous 12 months. That describes local-business discovery generally, not healthcare bookings or the proportion of searches using AI. (Source: BrightLocal, Local Consumer Review Survey 2026)
For your practice, the practical question is whether someone researching your services can find accurate information about who you help, where you work and how to take the next step.
How to think about AI search
The four steps below are a maintenance framework for your online information, not a claim about a universal recommendation algorithm.
1. Check that relevant search tools can reach your public pages
Ask your website team to check crawl access, indexing controls, firewall rules and the content returned by important pages. Keep patient portals, private systems and confidential information protected.
Your technical team should use the current first-party documentation rather than assume every AI bot has the same role.
If you use Cloudflare, inspect the actual settings and logs for relevant search requests.
Make essential facts available as accessible page text. Do not rely only on an image, a booking widget or content loaded after browser interaction.
Use structured data that accurately describes the organization, services and eligible entities on the page. No schema type or llms.txt file guarantees a citation.
2. Answer the questions people ask about your practice
Review whether your pages clearly explain:
- the services and conditions within your actual scope;
- each provider's accurate credentials and role;
- your location, hours and any telehealth availability;
- what a first visit involves and how payment works.
Recommended writing approach: answer the page's main question early, use descriptive headings, keep the language plain, and explain the next step.
For medical claims, use appropriate evidence and review by a qualified clinician. Add the real reviewer and review date only after that review happens.
Compare an abstract sentence such as "We take a whole-person approach" with an informative pattern such as: "[Provider name and verified credential] offers [service within scope] in [location], with [verified visit options]." The second version tells a reader exactly who you help, where and how.
3. Keep important listings accurate
Review your practice name, address, phone, providers, services and hours across the listings relevant to your practice. These may include Google Business Profile, Bing Places, applicable healthcare directories and professional listings.
Fix outdated information where you have control, and document requests where you do not.
Directory visibility can matter, but a statistic from a particular citation study is not a universal ranking weight. Focus on accurate information rather than chasing a promised algorithm shortcut. (Source: Yext study of 6.8 million AI citations)
4. Build credible public information without manipulating reviews
Keep provider credentials, relevant publications, association profiles and other legitimate public information accurate. Seek genuine recognition, not invented endorsements.
For Google reviews, do not offer incentives or selectively solicit positive reviews; its platform policy prohibits those practices. (Source: Google Maps user-contributed content policy) The FTC rule separately prohibits fake reviews and incentives conditioned on a positive or negative sentiment. (Source: DLA Piper: the FTC's approach to fake reviews)
Public review replies should not confirm patient status or disclose patient-specific information. Have the practice's privacy/compliance lead approve its actual process.
Track what you observe
Choose a consistent set of realistic questions based on your practice's city, services and audience. Record the tool, date, question, search mode where visible, response and sources. Repeat checks consistently and save the outputs.
Treat this as a sample. An increase in mentions is not the same as an increase in inquiries or booked appointments. Track those outcomes separately through the practice's approved process.
How long does it take?
The timing and extent of changes vary. Fixing a page does not guarantee when a tool will revisit it or whether it will include the practice in an answer. Modern Harvester's AI Visibility Program starts with 90 days of work, with regular comparisons against a baseline, and can extend if more work is needed. It does not promise a recommendation by a set date.