You have four completed visibility audits on Boulder-area practices, run in June 2026. None of them has been sent to anyone. Two of the four never even reached report form. This page pulls all four into one place with the real numbers attached, flags the figures that two runs disagree about, and then gives you a fresh list of med spas worth auditing next: the ones sitting on organic page two while the map pack at their own front door belongs to somebody else.

What was audited, at a glance

Only the first two are med spa work. Yamashiro and Boulder Dental Arts are orthodontics and general dentistry, audited with the same tooling in the same week, and they are here because they are part of the same unsent stack.

A note on what the citation index measures: it runs the same fifty questions across five AI chat tools (ChatGPT, Claude, Gemini, Grok, Perplexity) and counts who gets named. Google AI Overviews was skipped in every run, on purpose, because it has no chat API and needs a geo-located SERP call instead. So every "one of five" below means one of five chat surfaces, not one of six places people search.

Evolution Dermatology + Wellness

1455 Yarmouth Ave, Boulder. PA-led combined medical dermatology and med spa, no MD dermatologist on staff, no Mohs program. This is the deep one: four separate runs between June 9 and June 16, covering AI citation, site health, a 25-point map rank grid, organic SEO, and a sourced competitor study.

They are close to invisible in AI. Citation index 2.5 out of 100, last of the four practices measured. One tool out of five named them, and only Perplexity, in four of its ten answers. ChatGPT, Claude, Gemini, and Grok named them zero times each. Boulder Valley Center for Dermatology scored 19.8 and was named by four tools.

The detail worth putting in front of them: the single most-volunteered name across every answer was a generic "Boulder Dermatology", eighteen times, ahead of every real practice in the set. The AIs are reaching for a category label because no practice has made itself the obvious answer.

Organic search is carrying national reading traffic, not Boulder patients. They rank for 871 keywords with 196 in the top ten, which sounds healthy until you look at what those keywords are. Almost all the traffic comes from one blog topic (self-tanner safety, five variants at 480 to 590 searches a month, ranked one or two) plus their own brand name. They hold zero top-ten positions on dermatologist boulder, botox boulder, or dermatology boulder. The only local commercial terms they own are two spellings of microneedling boulder, about 70 searches a month each.

The map is their one genuine strength, and it is uneven. Across 275 cell-queries they appeared 160 times and took a top-three spot in 107 cells, average rank 3.5. They win the #1 spot most often on acne treatment boulder, eczema treatment boulder, skin cancer screening boulder, dermatologist near me, and cosmetic dermatology boulder. They lose every high-margin cosmetic term to dedicated med spas: Rinnova Skin and Body takes botox boulder, mole removal boulder, and laser skin resurfacing boulder, and Vasu Skin Solutions takes ipl photofacial boulder.

That is the strategic read the June 16 report lands on, and it still holds: they are defending the insurance-based medical terms they already win, while the cash-pay cosmetic terms that actually pay are contested but not locked.

Site problems, in order of what a fix is worth. Their schema is Organization only, with no MedicalBusiness or LocalBusiness type at all, which is the single highest-value technical fix on the list. The H1 "Better Skin Health is Our Mission" appears twice on the page and names no city. There is no embedded Google map. The footer copyright still reads 2022. The title tag is 32 characters. Page speed was never measured: both PageSpeed and Lighthouse returned HTTP 429 with no API key configured, and the report refuses to estimate rather than guess.

Skin Deep MD

Boulder med spa, Dr. Melody Denham. Picked as the strongest candidate out of a pool of fourteen Boulder med spas. Audited June 8: AI citation across five tools plus a site read. The live map grid was never run, because it needed paid API spend that was not authorized at the time.

Best of a nearly invisible group. Citation index 6.5, first of four, but the report is blunt that the whole cohort barely registers. Perplexity named them in nine of ten answers. Gemini, Grok, ChatGPT, and Claude named them zero times each.

The sharpest thing in this audit is not their own score. Gemini and Grok both confidently recommend the same rival, Boulder Plastic Surgery and Medical Spa, and Grok delivers it as a single top recommendation. Meanwhile Claude declines to rank Boulder med spas at all and sends patients to RealSelf and Google Maps instead, saying it lacks reliable enough information. That is a live, quotable demonstration that the question "who should I see in Boulder" is being answered by somebody else.

Their site is in better shape than Evolution's. Ugliness score 24 of 100. MedicalBusiness schema is already present, six schema types, copyright year current, 118 KB page weight, no DIY-builder template. Two real problems: the homepage takes 6.3 seconds to load, and there is no embedded map and no maps link anywhere on the site. "Boulder" is in the title tag but not in the H1. They surface Healthgrades and Vitals badges rather than their own Google reviews.

There is also a full redesign mockup already built for them, a real Astro site at 12.7 KB against their 118 KB, with verified desktop and mobile screenshots. Before any of that goes out, the handoff lists five things to fix, including that the provider photo in the mockup is a stock image and not Dr. Denham.

Their review count needs settling before anyone quotes it: verified Google says 4.9 on 34 reviews, their own site schema claims 4.98 on 40, and the June 14 sweep recorded 5.0 on 29.

Yamashiro Orthodontics and Boulder Dental Arts

Both audited June 16, both citation plus organic only, and neither has a written report. Three JSON files each.

Yamashiro scored 6.7, second of four, named only by Perplexity in seven of ten answers. The number that matters is the competitor's: Boulder Orthodontics scored 43.7, was named by four of five tools, and appeared in 52% of answers, roughly six and a half times Yamashiro's share. Organically Yamashiro ranks for ten keywords total, one of which is their own name, and their best non-brand position is rank 4 on "orthodontist in boulder co" (1,000 a month), the exact term Boulder Orthodontics holds at #1.

One caution before that number gets used: the SEO run measured orthodontistinboulder.com, while the June 11 shortlist lists their site as yamashiroortho.com. If the script pointed at the wrong domain, the ten-keyword finding is not about their real site. Check that first.

Boulder Dental Arts is the interesting inversion. It has the best AI visibility of all four practices (index 19.0, and the only one named by two tools, with Gemini naming it in nine of ten answers) and by a distance the worst organic footprint: 22 keywords, zero of them in the top ten, best position rank 20 on "boulder dental" (1,900 a month). Boulder Dental Center, by comparison, has 187 keywords and thirteen at position one.

Three numbers not to quote without checking

These came out of the audit files themselves and they are worth knowing before any of this reaches a business owner.

Evolution's citation index exists in two versions. The June 9 run says 2.7 but dropped six of fifty captures when an OpenRouter billing limit hit mid-run. The June 11 rerun says 2.5 on a clean fifty. Use 2.5. The June 9 markdown and HTML are superseded.

Competitor citation scores moved a lot in two days. Same prompts, same models, June 9 to June 11: Dermatology Specialists fell from 15.9 to 5.5 and from being named by all five tools to two. Mandel fell from 9.1 to 8.2. Boulder Valley held steady. Say "snapshot" out loud when presenting any of these, because a swing that size in a headline number will not survive a client rerunning it.

Two map pack runs disagree about who is winning Boulder, including about Evolution. The June 11 grid puts Evolution top-three in 107 of 275 cells. The June 14 cohort sweep scores the same practice at 2.7% top-three and calls it MIXED. The runs used different radii and different keywords, so some gap is expected, but there is a known name-matching bug in that tooling: the June 11 raw output initially reported Evolution appearing zero times in all 275 cells, because the matcher keyed on "Evolution Dermatology and Wellness" and the Maps title is "Evolution Dermatology". Nothing records whether the June 14 sweep was corrected for the same fault.

The clearest tell is Azeal Dermatology Institute: June 11 gives it 29 top-three appearances and the best average rank on the board, June 14 scores it 0.0% found and labels it invisible. One of those readings is wrong. Neither set should be published as settled until a reconciliation run is done.

The June 14 cohort sweep, for context

Sixty-five Boulder med spa and dermatology businesses, filtered to 4.5 stars and 25 reviews minimum, each measured on a 25-point grid across three high-intent keywords. Total cost $2.86.

One business dominates the Boulder map: RESTŌR Medical Spa, top-three in 68% of cells, found in 80%, median position 1. Sixteen businesses scored 0.0% found across all 75 cells, several of them large and well reviewed: Dermatology of the Rockies (4.8 on 1,147 reviews), Colorado Center for Dermatology and Skin Surgery (4.9 on 729), Youthbar (5.0 on 461), Azeal Dermatology Institute (4.9 on 121).

The pattern worth selling against is the middle: businesses found nearly everywhere and chosen nowhere. Aura appears in 74.7% of cells and takes top-three in 0.0% of them. Same for Alchemy Face Bar (70.7% found, 0.0% top-three), Lindsay Boyers MD (77.3% and 0.0%), Jenna Peart MD (76.0% and 0.0%). They are in the index and out of the running.

The med spas to audit next

Run on 2026-08-08. For every med spa in Boulder and Longmont whose street address we could place on a map, we ran five searches localized to that business's own coordinates: "med spa near me", "botox near me", plus the same three city-qualified terms in whichever city it sits in ("med spa boulder", "botox boulder", "laser hair removal boulder", and the Longmont equivalents). So this is what somebody standing in their parking lot sees. Desktop, top twenty organic results, 280 searches, $0.59 of API spend.

Of 56 med spas checked, 4 sit on organic page two with no map pack placement at their own address. That is the list you asked for, and it is the best-shaped group to approach: they rank well enough to prove the site is indexed and the content is real, and they are still losing the only result a local searcher actually clicks.

On page one organically, still shut out of the pack

2 med spas rank on organic page one for at least one of these searches and still do not appear in the map pack at their own address. Their site is doing its job and their Google Business Profile is not, which usually means categories, service areas, review velocity, or a listing detail rather than anything on the website.

Absent from the pack and outside the top twenty as well

36 med spas returned nothing on any of these searches from their own address: no pack placement, no organic result in the top twenty. These are a harder conversation, because there is no near-miss to point at, but the gap is larger and so is the upside.

Holding the pack today

14 of the med spas we checked do appear in the pack at their own address. They are not prospects for this pitch, and they are the names that keep coming up in everyone else's results.

How to read this, and what it does not say

The map pack is three slots and it is heavily distance-weighted, so a business can be absent from the pack at its own address and still show up half a mile away. This measures one point per business, not a grid, which is the cheap version of the 25-point rank grid run on Evolution in June. Treat it as a shortlist for where to spend a real grid, not as a finished diagnosis.

Five searches is still a narrow read. A med spa invisible on all of these may own "lip filler longmont" outright. Nothing here checks their own brand name either, which almost everyone ranks for, so nobody in the invisible group is invisible to their existing patients.

One matching detail worth knowing, because it changes the numbers: organic results are matched to a business by domain only. An earlier pass matched on the result title too, and an organic listing whose title was the single word "Boulder" then matched every practice with Boulder in its name, inventing a dozen rank-one placements that were not real. Pack entries are still matched by name, since pack titles are business names, so a practice that has changed its trading name may be under-counted there.

Businesses whose listed address sits more than 8km from the Boulder or Longmont city center were dropped even where our directory files them under those cities, and eleven more had no usable street address to place. So this is a floor, not a full count.

Where the files live

Everything above is in FindWebSiteClients/outputs/. Evolution has the fullest set: a June 9 audit in markdown and HTML, a June 16 three-dimension report, and separate findings files for SEO, map pack, and competitors. Skin Deep MD has a citation snapshot plus the Astro mockup. Yamashiro and Boulder Dental Arts are raw JSON only.

The evolution_triangle/ directory is empty. There is no Citation Triangle artifact anywhere, despite the name appearing in the run scripts.