Massage equipment is a textbook "legacy-brand-dominated" category. When consumers ask ChatGPT or Perplexity for "a professional massage at home" or "alternatives to a massage table," the default answers surface brands that have been selling to physical therapy clinics for two decades. According to Statista's 2026 Wellness E-Commerce Report, 63% of at-home wellness buyers under 45 begin their product research inside a conversational AI tool.
Magarri's flagship product — a mattress-based massage kit that installs in 30 seconds without tools and fits 99% of mattresses — has an unusually clear value proposition. But in AI search, those differentiators rarely made it through. The team chose Agentic Page to deliver a semantic payload to AI crawlers and deploy a Competitor Hedging Graph against the legacy players.
"We're not trying to beat Earthlite or Master Massage. We just want AI to know that for anyone who doesn't want to install a full massage table, Magarri is an answer."
— Sarah Chen, Head of Brand, Magarri
Challenge: a differentiated product invisible in a legacy-brand category
Head of Brand Sarah Chen ran a cross-team audit in early 2026. Three things were clear:
- Target buyers were asking AI question-shaped queries: "how do I do a professional massage at home," "massage solutions for small apartments," "massage gifts for a partner." These prompts lean toward scenarios, not brand names.
- Magarri's Shopify site delivered its differentiators through video, client-rendered comparison tables, and collapsed FAQ sections. Humans saw the value; AI crawlers saw empty divs.
- Earthlite, Master Massage, and Saloniture already appeared consistently in AI comparison answers. Magarri barely registered.
The goal became specific: for three prompt categories — "portable massage equipment," "at-home massage solutions," and "alternatives to a massage table" — Magarri needed to be in the AI shortlist.
Solution: why Magarri chose Agentic Page
In the first week, Agentic Page delivered:
- ACCC Diagnosis — first scan: 54 (Fair band). robots.txt not allowing major AI crawlers, hero video without a transcript, competitor comparison table client-rendered, "30-second setup" differentiator buried behind a third-fold tab, URL semantics weak.
- AI Mirror Site — 34 core pages. 6 PDPs, 8 use-case articles, 5 comparison reviews, 9 FAQs, 6 blog tutorials. Cutting tokens 59% while raising the density of critical attributes.
- Traffic Monitoring: AI-source traffic broken down bot by bot (GPTBot, PerplexityBot, ClaudeBot, Google-Extended) with prompt-level attribution.
Implementation: deploying a Competitor Hedging Graph against the incumbents
Phase 1 — Baseline + Phase 2 — Mirror generation: Two weeks baseline across ChatGPT, Perplexity, Claude, and Gemini. Then the team reviewed every PDP to confirm differentiators — "30-second setup," "fits 99% of mattresses," wood grade, adjustable angle range — were preserved as structured tables.
Phase 3 — Competitor Hedging Graph
A structured, LLM-parseable comparison table plus decision tree pre-encoded the buyer's trade-off variables against Earthlite, Master Massage, and Saloniture:
- Setup time and tools required (30 seconds, no tools vs. 20–40 minutes, multiple tools)
- Space footprint (zero permanent footprint vs. dedicated room)
- Mattress / surface compatibility (99% of mattresses vs. dedicated table)
- Price range and total cost of ownership
- Primary use case fit (couples, postpartum, small apartments, recovery)
Phase 4 — Content-side EEAT lift: Converted competitor comparison table from client-rendered to server-rendered; added structured transcripts to hero video and product demos; pulled "30-second setup" up from third-fold tab to top-line summary; added "who this is for / who this isn't for" block covering remote workers, athletic recovery, couples, and postpartum recovery.
Results: seven weeks, a challenger inside the AI shortlist
- AI Citations: +268%. Mirrored pages cited in AI responses nearly 3.7x as often.
- Prompt Coverage: +319%. Prompts where Magarri appeared grew from 41 to 172.
- Non-Branded Presence: +420%. Magarri mentioned more than 5x as often in prompts not including "Magarri."
- ChatGPT Head-to-Head vs. Legacy Brands: 0% → 35%. Magarri went from barely appearing to showing up in over a third of direct comparison answers.
- "Massage Table Alternative" List Appearances: 6% → 58%. In AI-generated lists of "alternatives to massage tables," Magarri moved from occasional to regular.
- ACCC Score: 54 → 91. All four dimensions moved into the Excellent band.
- AI-Sourced Conversion & AOV: Conversion +48%, AOV +22%. Buyers from AI platforms converted higher and spent more.
"Challenger brands used to feel outgunned by the incumbents. Agentic Page showed us something different: in the AI era, a clearly differentiated product is the loudest voice in the room — if AI can actually read it."
— Sarah Chen, Head of Brand, Magarri
Key Takeaways for small DTC brands in legacy-brand categories
- Don't wait to be discovered. Deploy a Competitor Hedging Graph encoding every trade-off against incumbents.
- Promote your differentiator to the top fold. Buried selling points are invisible selling points in AI search.
- Transcribe every demo video with timestamps. Media-only content is uncitable.
- Add a "who this is for / who this isn't for" block. LLMs reward explicit use-case fit over generic marketing copy.
- Track post-click signals. Citations are the entry; engagement is what keeps you in the RAG pool.
Industry playbook: AI Visibility for Clean Beauty & Skincare DTC →