EVIDENCE · BOX-GRADED · RAW DATA PRESERVED
Cases — evidence from the field.
Real engagements · real measurements · BOX-graded evidence · raw AI-engine output preserved.
ANSWER · AEO-EXTRACTABLE
Two vAEO cases at launch · both with full methodology disclosure + BOX-graded evidence.
Barcelona Boat Rental is the first published client case. D2 Full Snapshot delivered 2026-04-23 · cross-engine probe across 5 AI engines on 20 commercial queries. Pre-audit measurement showed Barcelona Boat Rental at 19% Share of Model versus dominant competitor Click&Boat at 44% across the probed query set · BBR cited zero times in Perplexity for the top 4 commercial queries · Gemini consistently misclassified BBR's category (charter vs rental). Citation Stack L1 + L2 fix cycle is delivered. L3 Semantics cycle measurement pass scheduled Q3 2026 per Proof Ladder P3 probabilistic timing. Real numbers · real client (named with permission) · raw AI-engine output preserved as audit artifact.
vAEO applied to itself is the dogfood case · mandatory at launch per P1 self-AEO compliance gate. Before vaeo.ai went live, we ran the same 5-engine probe on our own site. We documented baseline measurements · captured every Citation Stack 4-layer fix we made to our own pages · published the case live as audit artifact. Cobbler's son has shoes. Methodology proven against the methodology owner. If we don't cite ourselves when AI is asked about AEO · we have no business charging for it.
Every case on this page:
- Has client-verified permission OR is self-disclosure (dogfood)
- Reports metrics with BOX grade methodology disclosure
- Links к public Answer Capsule summary
- Preserves raw AI-engine output as audit artifact
- Discloses Proof Ladder cycle status (which layer assessed · fixed · measured · pending)
Cases we don't publish: unverified testimonials · screenshot-only «results» · cases without methodology disclosed · cases without client permission · cases without raw data preservation.
What cases prove · what they don't.
Two cases at launch. Both with full methodology disclosure. Both with raw measurement data preserved as audit artifact. Both honest about which Proof Ladder rung the case sits on.
What cases prove
- Citation Stack 4-layer methodology is real · reproducible · applied to actual brands
- Cross-engine probing (5 engines · ChatGPT · Perplexity · Gemini · AIO · Claude) surfaces gaps single-engine measurement misses
- L1 + L2 mechanical fixes deliver Citation Stack score uplift deterministically (Proof Ladder P2)
- L3 + L4 measurement happens on probabilistic cycle (Proof Ladder P3 · 30/60/90)
What cases don't prove
- Causation between AI Visibility uplift and downstream business outcome (Proof Ladder P4 · not promised by any honest vendor)
- Guaranteed outcomes for similar engagements («similar results await you» = anti-pattern · we don't say this)
Two cases isn't a portfolio. It's two honest data points. We publish more as engagements close + clients give sign-off.
«Two cases · honest about what they prove · honest about what they don't. We don't fake portfolio breadth.»
Published cases.
Two cases live at launch. Sorted by recency — dogfood newest · BBR L1+L2 cycle complete · L3 in flight.
vAEO applied к itself.
Methodology applied к its own surface before launch · seven first-pass misses disclosed honestly (three initial · four caught while writing the page) · 35% aggregated Share of Model at Gate C pre-launch probe. MANDATORY at launch per P1.
Barcelona Boat Rental.
Pre-audit Share of Model 19% vs Click&Boat 44% across 20 commercial queries · 5 engines. L1+L2 fix cycle complete · L3 measurement pass scheduled Q3 2026 per Proof Ladder P3 probabilistic timing.
«One real client · one self-disclosure. Both with raw data preserved. Both naming the Proof Ladder rung honestly.»
What we publish · what we don't.
Honest publication discipline. Trust-building. Differentiates from vendors who publish unverified testimonials.
Every case on this page
- ✅ Has client-verified permission OR is self-disclosure (BBR = named with permission · dogfood = self-disclosure)
- ✅ Reports metrics with BOX-grade methodology disclosure (Bronze · Silver · Gold · Platinum per claim · per Methodology M7)
- ✅ Links к public Answer Capsule summary
- ✅ Preserves raw AI-engine output as audit artifact (you can audit our audit)
- ✅ Discloses Proof Ladder rung honestly (which layer assessed · fixed · measured · pending)
- ✅ Names the boundary — what was deterministic · what was probabilistic · what wasn't promised
Cases we don't publish
- ❌ Unverified testimonials («great results!» without numbers · without methodology)
- ❌ Screenshot-only «results» (cherry-picked outputs without reproducible methodology)
- ❌ Cases without methodology disclosed (you couldn't reproduce what we did · so it's not science)
- ❌ Cases without client permission (anonymization defaults к Cases-BRB pattern · NOT to publication without consent)
- ❌ Cases without raw data preservation (verifiable claim requires audit-able artifact)
- ❌ Cases promising specific outcomes для future clients («similar results await you» = anti-pattern)
Where future cases will come from
As D2 / O / Dm engagements complete + clients give sign-off · we publish. Anonymization via Cases-BRB pattern (Brand Reference Blinded · industry-context preserved · client identity withheld) used when client wants the case but not the brand attribution. Frequency: realistically 1-2 new cases per quarter at current engagement pace.
«Two cases is honest. Twenty-eight cases unverified would be theater. We publish what we can verify.»
Pipeline.
What's queued for publication — no fake «coming soon» grid · concrete pipeline:
Q3 2026
- BBR L3 Semantics cycle re-measurement (Proof Ladder P3 · 30/60/90 cadence post L1+L2 fix)
- Updated BBR case with L3 cycle numbers + Citation Stack health score progression
Q4 2026 (pending engagement timing + client sign-off)
- First Optimize O3 Citation Engine case (full-stack engagement · target name disclosure)
- First Dominate Dm2 Grow retainer case (3-month retainer outcome data · target name disclosure)
2027 (anticipated)
- First Cases-BRB anonymized pattern case (YMYL-vertical engagement where client opts for anonymization · industry-context preserved)
- Additional D2 Snapshot cases as engagements close
What you won't see in this pipeline: fake «coming soon» case placeholders. Generic «testimonial» cards. Vendor-supplied logos without verifiable case work behind them. Empty grid filling.
«Two cases now. More when engagements close and clients sign off. Not before · not faked.»
Where you go next.
If a case resonates with your situation:
- Primary path — start with the canonical audit. $500 D2 Full Snapshot. Cross-engine probe. Decision-ready artifact in 5 days. Same delivery format as the BBR case.
- Alternative path — book a 30-minute Cal.com discovery call. We'll scope which tier fits your situation.
- Agency path — if you're bringing this to your own clients · the Partnership Track has wholesale pricing + 18-month firewall + WL Identity Protocol.
«Real evidence · BOX-graded · raw data preserved. Two cases · two honest data points. Then you decide.»