Openridge AI · Prepared for Joybody Physio & Pilates · July 2026

Can AI see Joybody yet?

thejoybody.com · Apr 9 – Jul 14, 2026 · every trend measured against event-marked windows; unlogged campaign days verified not to distort the reads

For most of this engagement, the honest answer was no — we discovered in late June that the May schema deployment had silently failed, so AI systems reading your site found almost nothing. That's also why nothing moved for eight weeks, exactly as our model predicts. On June 30 the foundation went genuinely live (and we verify it's still live, every cycle). This report covers what happened in the first two weeks since — and what has to happen next.

Brand appearances on Google
+43%
▲ 14.8 → 21.1 per day
Corrective era vs the 12 days after the foundation went live
Visits sent by AI assistants
6
▲ 4 of 6 in the last four weeks
Real ChatGPT click-throughs — the only client in our portfolio with them
Brand-search clicks
+29%
▲ 1.29 → 1.67 per day
Before the engagement vs post-foundation
The scorecard

The honest one-page answer

The foundation went live Jun 30 — most rows below are "two weeks in," and say so.
The questionHow we measure itWhat happenedAnswer
Can AI systems read your practice's identity?Fetching the site exactly as AI crawlers doInvisible until Jun 30 (failed deploy, caught + fixed) — now live & re-verified every cycle✓ live since Jun 30
Do AI assistants recommend Joybody?Mentions in 200 real physio/pilates questions, 4 platforms8.5% of answers — unchanged so far; market leader sits at just 16%▬ not yet — expected this early
Did AI send actual visitors?Site visits arriving from chatgpt.com6 visits, accelerating (4 in the last month)▲ first click-throughs
Are more people finding the brand on Google?Brand-name searches shown · clicked / dayAppearances +43% · clicks +29% since the foundation▲ early lift
Do you appear for treatment searches?Non-brand Google appearances / day35 → 37 — flat▬ waits on service pages + citations
Can we measure bookings?Clicks from the site into Jane0 recorded in 60 days — tracking gap found & root-caused, fix queued✗ instrument broken
What else is driving traffic?Everything in analyticsGoogle Ads running all period + a Jul 4 Facebook post + newsletters — none logged with us? need the full picture
Question one

Do AI assistants recommend Joybody?

200 identical questions across ChatGPT, Gemini, Claude and Perplexity, every run. Markers: the failed May deploy · the real Jun 30 foundation.

Not yet — and two weeks in, that's the expected answer. AI assistants recommend what trusted third-party sources say, and until June 30 they couldn't even read your own site properly. The citation work now underway (PainHero, Yelp, Instagram live; Bing, RateMDs, BBB in flight) is what moves this line — and the market leader sits at only 16%, so the ceiling is low.

AI answers naming Joybody (live search) % From AI memory (no search) %

The honest detail: the purple line was flat before the foundation too — because there was nothing for AI to read. That's actually the strongest evidence in this report: across our whole portfolio, clients with a live foundation moved; Joybody without one didn't. The pre-agreed test is the week-12 measurement in late July against the clean June 29 baseline.

The numbers
RunLive search %From memory %Failures
May 53.66
May 202.50
Jun 157.02.00
Jun 29 (pre-foundation anchor)8.01.00
Jul 148.52.50
Question two

Did AI send anyone? Yes — six visitors, and counting.

Visits arriving on thejoybody.com directly from chatgpt.com.

Joybody is the only client in our portfolio receiving real AI click-throughs. Product brands get recommended and the shopper never clicks; healthcare works differently — people click through to vet a provider before trusting them. Four of the six visits came in the last four weeks.

Six visits won't fill a schedule — but the trajectory matters, and so does what it implies: when AI recommendations grow (question one), the clicks follow here in a way they don't for product brands. The missing piece is knowing which of these visitors booked — see the tracking fix below.

Question three

What moved on Google in the first two weeks?

Three periods: before the engagement · the corrective era (May 8 – Jun 29) · after the foundation (Jul 1–12).

Brand searches — early but real. Appearances +43% and clicks +29% in the twelve days since the foundation and on-page fixes went live. Treatment searches ("physio surrey", "clinical pilates near me") are flat — as expected, because the content and citations that move them haven't shipped yet.

Brand appearances / day Brand clicks / day

Two honesty notes. First, twelve days is a short window — we call this "early and consistent with the fix," not proof. Second, we verified the lift is not your July 4 Facebook post or the newsletter (excluding those days, the numbers hold). One thing we need from you: your Google Ads have run all period — if they bid on the name "Joybody," the organic numbers above are actually understated.

The numbers
PeriodAppearances/dayClicks/dayGoogle position
Before (Apr 9 – May 6)15.91.292.4
Corrective era (May 8 – Jun 29)14.81.432.1
Post-foundation (Jul 1–12)21.11.672.6*

*Position moved from 2.1 to 2.6 because you now appear for more searches, and new appearances start lower — dilution from growth, not a drop.

What happens next

The plan, and what we need from you

We need from you

  • The "how did you hear about us?" question in Jane intake (Casey is setting this up — please green-light it). Your booking clicks are currently invisible to analytics (we found and diagnosed the gap); this question is the durable fix, and at another client it identified ChatGPT-sourced customers within two weeks.
  • Google Ads details — what campaigns and keywords are running? If the ads bid on "Joybody," it changes how we read the brand numbers.
  • GA4 access + one setting — we have read-only; we need edit access to extend data retention from 2 to 14 months (every week of delay permanently loses a week of future history).
  • The running list — the July 4 Facebook post and your newsletters showed up in the data unannounced. A quick heads-up when you post/send/advertise keeps every report clean.
  • Uninstall the Wix Bookings app — it still advertises a broken booking path to AI agents.

We're doing

  1. Finishing the citation sweep (with Ava) — PainHero, Yelp, Facebook, Instagram live; Bing, RateMDs, BBB next. This is the lever that moves AI recommendations.
  2. Service-area pages — the content that gets you found for treatment searches, not just your name.
  3. Schema v4 — corrected map pin + linking your profiles, once the citation URLs land.
  4. Re-verifying the foundation every cycle — Wix can silently undo it; we check the raw code each review.
  5. The week-12 measurement (late July) — the pre-agreed test of AI recommendations against the June 29 baseline.

What we can say today — and what we won't say yet

We can say: your practice's identity is now genuinely readable by AI systems, verified in the raw code — for the first time. Brand appearances +43% and clicks +29% in the twelve days since. Six real visitors have arrived from ChatGPT, four in the last month. And the eight flat weeks before June 30 are themselves evidence: without the foundation, nothing moves.

We won't say yet: that AI recommends Joybody more (8.0 → 8.5% is statistical noise — the pre-agreed test is the week-12 measurement); that the brand lift is proven (twelve days is directional); or anything about bookings — the tracking gap means we honestly can't see them until the Jane intake question is live.