Cases · seven engagements · one method
Not a lucky case. A repeatable playbook.
Three growth proofs (our dogfood, Birdview PSA, GAC auto retail) show what the loop does in practice. Four baseline audits (Whitewill Dubai real estate, Gorbilet tourism, Nonton retail, LS ELECTRIC manufacturing) show why a brand is invisible before the work starts — and the diagnostic pattern repeats across categories.
Engines measured
- ChatGPT
- Claude
- Perplexity
- Gemini
- Grok
- DeepSeek
- Kimi
- Google AIO
- Copilot
Seven case studies · sorted by case number
One method. Seven categories.
Each case is sourced from the live Hermes Visibility dashboard plus public AI-engine queries we ran during the engagement. Original case write-ups are linked from each case page.
Case 01 · AI marketing platform (our own brand)
Humanswith.AI
1,000+ · Worldwide · 12 weeks
We tested the playbook on the hardest target first — ourselves. No incumbent advantage, no friendly journalist on speed dial, no warm-list of clients vouching for us. Just the loop, on our own brand, for three months.…
Case 02 · B2B SaaS · Project Management
Birdview PSA
21.5% · Canada and USA · 8 weeks
Birdview had strong AI visibility inside the narrow PSA (Professional Services Automation) cluster — Perplexity 40%, Claude 41%. But in the broader PMS (Project Management Software) cluster where the larger audience l…
Case 03 · Automotive retail
GAC
9,042 · St. Petersburg, Russia · 6 weeks
GAC was AI-blind in St. Petersburg auto retail — one mention across nine AI engines, while competitors got recommended daily. Six weeks of format-driven content (calculations, competitor comparisons, 'should I buy?' g…
Case 04 · Premium real estate
Whitewill
Surface-first · Dubai, UAE · 12 weeks
Whitewill is a strong Dubai-real-estate brand offline — but for 121 high-intent investment queries on AI engines, the brand had zero citations at baseline. Meanwhile Medium carried about 97 mentions, Engelvoelkers 92,…
Case 05 · Tourism · river tours
Gorbilet
94.6% · Russia · 10 weeks
Gorbilet's tourism niche behaves differently from B2B SaaS or real estate. The total AI citation surface for the topic produced 8,861 mentions — and Gorbilet already held 289 of them with 100 branded mentions and a 35…
Case 06 · Retail
Nonton
200 · Russia · 12 weeks
Nonton had strong brand-search results — 127 mentions when AI engines were asked about the brand by name. But across 160 non-branded category queries, the brand showed up in exactly one answer. Brand recognition was n…
Case 07 · Manufacturing · industrial automation
LS ELECTRIC
Alternatives, comparisons, distributors · Global · Russia regional site · 12 weeks
LS ELECTRIC's regional site had 66 AI citations on the topic — the entity was recognised. But 64 of the 66 citations landed on the homepage and only two on the support section. The global LS ELECTRIC site held 170 cit…
7
case studies · 7 categories
9
AI engines measured per case
6–12
typical first-lift window · weeks
Browse cases by industry
Same playbook. Vertical-tuned.
Each industry page surfaces the agency-era proof in that vertical plus the modern AI-search playbook we ship today for it. Useful if you want to see vertical-specific examples before the strategy call.
Healthcare
Clinics, telemedicine, medical SaaS — YMYL-safe AEO.
12 cases →Logistics
Freight, fulfillment, supply-chain SaaS — trade-lane content.
6 cases →Real estate
Brokerages, developers, property tech — per-agent authority.
2 cases →E-commerce
DTC, marketplaces, retail tech — comparison-content engine.
3 cases →Why this works
One loop. Seven different categories.
A buyer might wonder why the seven cases share the same shape. It is not coincidence; it is the method. Every engagement runs the same loop: Hermes measures the per-engine citation gap, ContentOS ships the trust-graph content that closes it, Website Agentic Optimization gets the schema right, then Hermes re-measures. The numbers move because the loop moves them, not because the category is special. A B2B SaaS, a real-estate brokerage, and a regional industrial site converge on the same shape of result. That repeatability is the proof you cannot fake.
How the loop runs
What 90 days looks like.
The same sequence ran for every case below. Industry-specific deliverables differ; the loop shape does not.
- Week 1. Hermes baseline scan: per-engine citation map across 9 AI engines, ranked against the 5 closest competitors.
- Week 2–3. Trust graph: Organization, Person, Service schema; sameAs to verified profiles; third-party authority anchors (Clutch, Google Maps, vertical-specific).
- Week 4–7. ContentOS ships 6 chunk-ready pieces (research-driven, factcheck-verified, citation-target). Each piece a candidate for AI extraction.
- Week 8–10. Off-site authority: third-party publication placements, podcast guesting, vertical-association listings.
- Week 11–12. Hermes re-scan + report. Per-engine delta vs baseline, gaps that closed, gaps that need a second round.
Agency archive · 2019–2025
Looking for the agency-era work?
Performance, SEO, CRM, content, growth — thirty-four preserved cases from before the 2025 platform pivot.
The seven cases above use a modern proof model: per-engine AI mention rate, citation tracking across nine engines, weekly delta reports. The agency-era work used a different proof model entirely (Google Analytics, CRM dashboards, ad-platform metrics) and lives in its own archive so the two don't get read with the same lens.
Categories include healthcare (clinics, dental, medical), logistics (Infinity Logistix), e-commerce, real estate, auto, tooling (CRM, outreach, Notion), and agency positioning.
Case questions buyers ask
Frequently asked about the cases
How long until results show up?
Growth proofs in this set ran 6 to 12 weeks. GAC reached the default-recommendation position in six weeks. Birdview PSA hit a 23-times ChatGPT lift in eight weeks. Humanswith.AI dogfood ran for three months. Baseline audits (Whitewill, Nonton, LS ELECTRIC) ship results in execution quarters that follow the audit.
Are these numbers verified?
Yes. Each case is sourced from the live Hermes Visibility dashboard plus the public AI-engine queries we ran during the engagement. The original case write-ups are linked from each case page on this site and on gregshevchenko.com/research.
Why a mix of growth proofs and baseline audits?
Growth proofs show what the operating method does in practice. Baseline audits show why a brand is invisible before the work starts — and the diagnostic pattern repeats across categories. Both are useful to a prospect deciding whether the same loop applies to their situation.
Will my industry work?
We have shipped the playbook in B2B SaaS, auto retail, premium real estate, tourism, retail, and industrial manufacturing. The Hermes audit shows the closeable gap in your category before you sign — if the third-party source set or commercial-intent surface is thin, we will tell you so.
Can I talk to one of the clients?
Yes. After the strategy call we connect you with a current client in a comparable category. We do not run references blind to either side.
Why is the playbook the same across categories?
The mechanics of getting cited by AI repeat: map the query landscape, ship chunk-ready content on the surfaces AI engines trust for your category, measure weekly. The channel mix changes per niche (Medium for B2B SaaS, operator sites for tourism, Engelvoelkers-class publications for Dubai real estate) but the loop is the same.
Want a case like one of these?
The strategy call shows your closeable gap. The audit shows your AI mention rate.
An engineer from our AI search team runs your brand through Hermes before the call. You see exactly where each engine names you today. We tell you which of the seven cases looks most like your situation — and whether the same loop applies.