Reviews have acquired a second job. They still persuade patients reading them, and now they supply the material an assistant uses to justify naming you. Understanding that changes what you pay attention to.
The justifying sentence
When an assistant names two or three practices, it typically adds a short reason for each. That sentence is usually assembled from review content and profile data.
So your reviews are not just a rating. They are the raw material for how you get described in a channel where a patient may never see the reviews themselves. If several of your reviews mention that staff explained the process clearly and the monitoring period felt safe, that becomes your characterisation. If the recurring theme is difficulty reaching the phone, that becomes your characterisation instead.
What matters, and in what order
The weighting differs from what patients respond to:
- Recurring themes. The most influential factor. A summary is built from patterns, not from your best review.
- Recency. Current reviews carry more weight than old ones, so a strong profile from three years ago decays.
- Volume. Enough reviews to establish a pattern. A handful cannot produce a confident summary.
- Breadth across platforms. Consistent sentiment on several sources is more robust than depth on one.
- Substantive content. Reviews with detail give a summary something to work with. Star ratings with no text give it nothing.
- Rating. Matters, but a five-star average with six reviews and no text is weaker input than a solid average across sixty detailed ones.
That last point is worth absorbing, because it inverts the usual instinct to protect a perfect average.
Influencing it honestly
The line here matters. You cannot tell patients what to write, and you should not try. What you can do is make the experience you want described actually happen, and make the ask routine so that the pattern reflects reality:
- Get the operational basics right. Reviews describe what patients experienced. Phone responsiveness, wait times, billing clarity, and how well the process was explained are the recurring themes in this specialty, and all four are fixable.
- Explain the process well in clinic. Patients who understood what would happen say so, unprompted.
- Ask consistently so recency stays healthy, per the previous lesson.
- Ask across a small number of platforms, not just one, for breadth.
- Never suggest wording, never solicit clinical detail, never incentivise. Beyond being prohibited, engineered reviews read as generic and contribute little to a summary.
Why a perfect rating is not the goal
A flawless average with low volume is weak input and reads as implausible to patients. A genuine average with substantial volume, recent activity, and thoughtful responses to criticism produces a better summary and more trust.
This is also the practical argument against gating, from earlier in this chapter: suppression produces a thin, uniform, low-credibility body of reviews, which is exactly what an assistant cannot build a confident description from.
Watching what gets said
Add this to the AI monitoring from chapter five:
- Ask an assistant about your practice and read the descriptive sentence carefully.
- Ask what patients say about your practice, and see which themes surface.
- Note anything unfavourable and check whether it reflects a genuine recurring complaint. Usually it does.
- Compare the description with a competitor description to see which themes differentiate you.
- Record it on the same monthly schedule, so you can see change.
The honest conclusion
You cannot optimise your way to a favourable AI summary. It is downstream of what patients actually experience and choose to write about.
Which means the most effective reputation work in this channel is not marketing at all: answer the phone, explain the process, be clear about cost, and keep asking for reviews. Chapter eight turns to referrals and payer directories.