What Are Online Reviews?

Online reviews are public ratings and written comments that patients leave about a practice on sites such as Google Maps, Yelp and health directories. For a health practice, they are often the first thing a new patient reads and one of the few off-site assets you can influence directly. They also supply the patient language that search engines and AI tools see when they describe you.

How online reviews work

A review has three parts that matter. The star rating is a quick signal. The count tells a reader how many people have shared an opinion. The text is where the useful detail lives, such as what brought the patient in, what stood out and how the visit went.

Google uses reviews as part of local ranking. Its business help page says more reviews and positive ratings can help a business's local ranking. It lists review count as one input to prominence, which is Google's word for how well known a business is. Google does not say how much weight reviews carry compared with distance or relevance, so nobody outside Google can give you a formula. What you can say is that reviews are part of the picture on the map. Your Google Business Profile is where most of them appear.

AI answers are a different surface, and the evidence is less tidy. An MGMA-published study of 33,340 ChatGPT citation events across 200 practices found that review aggregators made up 3.8% of the cited sources. The same study found zero direct citations of Google Business Profile, which its author attributes to Google not exposing reviews as crawlable web addresses. The author runs a commercial service, so treat the study as one data point. It suggests that reviews help most on Maps and less as a direct source in a ChatGPT answer.

That does not make reviews irrelevant to AI. Reviews are one place where outside people describe you, and tools look for facts that several independent places agree on, which is the idea behind corroboration. If your site says you treat thyroid conditions and your reviews mention thyroid care, those two sources agree.

Detail matters more than volume. Compare these two kinds of review.

• A bare review. "Great place, five stars." It confirms nothing a stranger can use.
• A detailed review. It names the problem that brought the patient in, what the clinician did differently, and what changed afterward in plain words.

The second kind helps a new patient decide, and it gives a tool a sentence it can reuse. You cannot script it. You can invite specifics by asking a simple open question, such as what brought the patient in and what stood out.

Reviews also work as free customer research. Repeated praise tells you which words to use on your pages. Repeated complaints about wait times, billing surprises or hard-to-reach staff point to operations problems that no amount of marketing will fix.

Replies are part of the system too. Google says replying shows you value feedback. In a health setting, replies carry a privacy risk, covered under review response privacy.

Example: online reviews at a health practice

Take a chiropractic clinic with two providers in a mid-size city. This example is illustrative, not a real clinic. It has 140 Google reviews, but most are one-line ratings from three years ago, and the newest is five months old. A competing clinic nearby has 60 reviews, with the newest from last week, and many describe the problem and the result in a few sentences.

A patient with low back pain searches "chiropractor near me" on a phone. Both clinics appear in the map results. The larger clinic shows a higher count. The competing clinic shows recent, specific reviews. The patient opens both profiles and reads a few comments. The recent, detailed ones answer the questions the patient has, such as whether the clinician listens and whether the visit was worth the time.

Now the clinic fixes its approach. It does not buy reviews or push every channel at once. The front desk manager owns one task. After a visit that went well, the patient gets a text with one link to the Google profile and a short line thanking them for coming in. The manager follows up once, politely, and tracks the count each month. Over a few months, newer and more detailed reviews appear. The clinic also reads its reviews every month, spots repeated comments about parking confusion, and adds a map and instructions to its confirmation text.

None of this promises a ranking change, because Google does not publish how it weighs reviews. It does change what a patient sees at the moment of choice, and it gives the clinic accurate patient language to reuse on its pages.

For the AI side, the MGMA-published data is the sourced anchor. In that study, hospital rosters and association credentialing pages accounted for about half of the cited sources, and practice-owned websites for about 23%. Reviews were a small slice. So the sensible plan puts reviews to work for Maps and for patient trust, and puts equal effort into the sources AI tools cite most. See AI source map for how to find them.

How to check yours in 60 seconds

1. Search your practice name plus your city on a phone.
2. Open your Google profile and the top two review sites that appear.
3. On each, write down the review count, the date of the newest review and whether recent reviews describe a condition or treatment.
4. Check that the practice name, specialty and address match your website.
5. Look at whether you have replied to recent reviews.

If the newest review is months old, or most reviews are one line, you have found where to start.

What to do about it

1. Read your last twenty reviews and list the repeated praise and the repeated complaints.
2. Fix the operations problems that reviews point to before you ask for more.
3. Pick one primary platform, usually Google, and one or two secondary ones that patients in your specialty use.
4. Assign one person to own the process, and give that person a simple tracking sheet.
5. Invite patients who had a good experience, soon after the visit, with one direct link. Offer a single follow-up, then stop.
6. Ask an open question such as what brought the patient in. Do not script the words.
7. Reply to reviews in general terms, without confirming a visit, a service or a condition.
8. Keep your name, specialty and address identical on every profile, so tools connect the reviews to one practice.
9. Review the numbers monthly, and judge progress over months, not days.