Hire the provider that can show a dated baseline, audit the full local-business stack in order, keep Invisalign questions separate from general dentistry, and report AI visibility apart from booked consultations. Reject unsupported percentages and any promise that ChatGPT placement can be guaranteed.
This is a hiring guide, not an agency ranking. The useful order is the one a provider must be able to inspect and document: Google Business Profile, name-address-phone consistency, structured data, reviews, third-party citations, then repeated measurement. Citevio is one option and should be tested against the same checklist as any other provider.
What should an agency verify in Google Business Profile first?
Verify ownership, the real-world practice name, primary category, address, phone, website, hours and location-specific services before discussing AI tactics. Citevio’s published datasets do not contain a Google Business Profile completeness score, so this step stays numberless.
Google’s Business Profile representation guidelines require accurate real-world identity, a precise address or service area, and a phone and website that represent the individual location. Ask the agency who owns the profile, what access it needs, what it will change, and how every edit will be logged. A complete profile is verifiable; a claim that one edit will trigger a ChatGPT recommendation is not.
How should an agency test NAP consistency?
NAP means name, address and phone. The agency should compare those fields across the website, Google Business Profile, Bing, major directories and structured data, then document every mismatch. Citevio has not measured a percentage effect of NAP inconsistency on AI mentions.
A search answer attributed the claim that inconsistent NAP can reduce AI mention probability by 40-60% to Kompozy’s local AI guide. The current cited guide discusses consistent business data but does not publish a study supporting that percentage. Citevio did not measure that effect, so there is no replacement percentage here. The defensible action is a field-by-field consistency audit, not a probability claim.
What structured data should an Invisalign agency inspect?
Inspect LocalBusiness or Dentist identity, address, phone, URL, clinician and visible FAQ content, then validate that the markup matches the page. Structured data is readable identity infrastructure; it is not a recommendation switch.
| Schema state | Raw count | Hiring implication |
|---|---|---|
| No LocalBusiness or Dentist schema | 284/527 (53.9%) | The provider should first establish a readable business entity. |
| LocalBusiness present, no FAQPage | 225/527 (42.7%) | Identity exists; visible question coverage remains separate. |
| Both LocalBusiness and FAQPage present | 18/527 (3.4%) | Both are uncommon in the sample; the study did not test recommendation outcomes. |
Google’s LocalBusiness documentation says structured data provides a standardized description of page content and recommends validating it before release. The published counts show a large implementation gap, not that schema causes AI visibility.
What do review volume and rating actually tell you?
They show reputation context, not a universal eligibility cutoff. In Citevio’s 523-practice Google review reading, the median was 250 reviews and 4.9 stars; 506/523 practices, 96.7%, were already at or above the industry-claimed 4.3 line. That line cannot distinguish most practices in this sample.
| Review measure | Raw result | What it means |
|---|---|---|
| Review count | Median 250 | Sample context, not a target. |
| Rating | Median 4.9; mean 4.83 | Most practices were tightly clustered at high ratings. |
| At or above 4.3 | 506/523 (96.7%) | The industry-claimed line was already crossed by nearly the whole sample. |
| At or above 4.5 | 484/523 (92.5%) | Still not a useful separator in this sample. |
| At or above 4.1 | 516/523 (98.7%) | Also not a useful separator in this sample. |
An Aether Agency article reports that locations recommended by ChatGPT averaged 4.3 stars, while a search answer strengthened that into an “effectively ineligible” claim below roughly 4.0-4.3. An average is not a cutoff. Citevio’s test shows that 4.3 is not discriminating in this dental sample; it does not show that reviews have no influence.
Review acquisition also has a compliance boundary. The FTC’s review guidance for marketers warns against selective positive-review requests, undisclosed incentives and fake reviews. Ask who writes the review request, whether every eligible patient is treated consistently, and how the practice retains approval.
What third-party citations should an agency be able to verify?
Ask for the exact external pages and profiles that corroborate the practice’s identity, clinicians, services and location. A citation list should distinguish the practice’s own site from independent sources and show whether an engine actually returned those sources.
Citevio’s agency visibility study v2 contains 82 manually collected answer blocks dated 20-21 August 2026. Seventy-eight blocks recorded a search call: 72 returned at least one source and six returned zero; four blocks recorded search_calls=none. That dataset concerns agency-selection answers, not patient-facing Invisalign practice recommendations, so it demonstrates a source-audit method rather than a dental outcome. The raw answer and agency CSVs are published at Citevio Open Data.
A provider should show where each citation lives, who controls it, whether the practice is named accurately, and whether the page is accessible without a login. Empty profiles and an undifferentiated list of links are not proof that an assistant used them.
How should an agency measure Invisalign AI visibility?
Agree the real Invisalign buying questions, record a dated baseline, track each engine separately, preserve the question set, and report named, cited and booked-consultation outcomes as different fields. A single screenshot is not a trend.
Citevio measures visibility on ChatGPT, Perplexity, Gemini and Google AI Overviews. The question set is fixed and agreed with the client at the start, and logged. It does not change mid-month, because a set that changes cannot be compared month to month.
For the exact hiring question behind this page, the August 2026 v2 CSV contains four answer blocks. Perplexity named and cited Citevio in neither recorded run; ChatGPT named and cited it in one of two. That 1/4 result is a baseline for this agency-selection question, not a result for any dental practice. The existing Invisalign and cosmetic query study remains the patient-side page; this guide answers the owner’s hiring question.
OpenAI’s crawler documentation says OAI-SearchBot is used for ChatGPT search results. Include it in the technical check, but keep the inference narrow: access can be tested; a recommendation cannot be promised.
Which unsupported claims should you challenge before signing?
Challenge any percentage without a sample, denominator, date and method; any rating average presented as an eligibility cutoff; and any promise that a technical edit will cause ChatGPT to recommend the practice.
- NAP probability: the 40-60% line lacks a supporting study in the cited page, and Citevio did not measure it.
- Rating eligibility: 4.3 is an outside claim being tested, not a Citevio threshold; 506/523 practices already cleared it.
- Causation: schema, reviews and source coverage can be audited, but these counts do not prove which change causes an engine to name a practice.
What should the final hiring scorecard contain?
The scorecard should make the provider’s work reproducible: baseline prompts, engine list, procedure-level reporting, booked-consultation definition, access checks, NAP audit, schema validation, review compliance, citation evidence, contract terms and an explicit no-placement-guarantee statement.
| Ask for | Pass condition |
|---|---|
| Baseline report | Named prompts, dates, engines, runs, named/cited outcomes and source URLs. |
| Visibility versus booked consultation | Two separate fields with a written definition for each. |
| Engine coverage | Each engine reported separately, with failures marked as not measured rather than zero. |
| Technical access | OAI-SearchBot, robots.txt, server response and raw-HTML checks. |
| NAP consistency | Field-by-field discrepancy list; no unsupported probability percentage. |
| Guarantee boundary | No agency can guarantee that an AI engine will name a practice, and Citevio does not offer that guarantee. AI placement cannot be bought; there is no paid slot in an AI recommendation. |
Compare providers against the same evidence. Citevio publishes its current scope and pricing, measurement method and raw market research. The adjacent veneers and smile-makeover guide shows how the same scorecard changes for another elective procedure.
Before requesting proposals, check whether your own site is readable:
Prefer to question the engines by hand? See how to check what ChatGPT says about your practice.
How did we measure this?
Schema and review figures were re-counted from the published June 2026 dental practice web anatomy CSVs. The denominator is 527 US dental practice websites for schema and 523 practices with a Google review reading for rating and review count. These are general dental-practice denominators, not an Invisalign-only sample.
The agency-selection baseline comes from the v2 answer CSV: 82 manually collected ChatGPT and Perplexity blocks dated 20-21 August 2026, including four blocks for the exact hiring question. Collection used chat interfaces, personalization was not excluded, and four blocks with search_calls=none are not counted as “not cited.” Files and limits are published at Citevio Open Data.
Common questions
What should a baseline AI visibility report include?
It should name the agreed patient questions, engines, location, date, run count, whether the practice was named, which URL was cited, and who appeared instead. Keep the question set stable. A baseline is useful only when a later reading can be compared with the same method.
Should an agency report AI visibility or booked consultations?
Both, but as different measures. AI visibility records whether an engine named or cited the practice for an agreed question. A booked consultation is a downstream practice event. Reporting can connect the stages, but it should not present visibility as proof that it caused the booking.
Which AI engines should an Invisalign practice track?
Citevio measures visibility on ChatGPT, Perplexity, Gemini and Google AI Overviews.
Can an agency guarantee ChatGPT recommendations?
No agency can guarantee that an AI engine will name a practice, and Citevio does not offer that guarantee. AI placement cannot be bought; there is no paid slot in an AI recommendation.
Why should an agency check OAI-SearchBot?
OpenAI identifies OAI-SearchBot as the crawler used to surface websites in ChatGPT search. An agency should check robots.txt, server responses and raw HTML for that crawler. Passing the check is an access condition; it is not evidence that ChatGPT will recommend the practice.