Positioning
Not Ahrefs / Semrush
No claim of full SERP database. Core is AI-answer share + crawl + loops.
Not a ChatGPT screenshot folder
Same prompt set, multi-model checks, full answers stored, re-check after ship.
Not a generic chatbot
Strategist Chat is scoped to project passport, gaps, loops, and evidence.
Not “AI will replace your team”
Operators own the ship. Software owns measurement and backlog discipline.
Access control
Public registration is off by default. Invite codes are reviewed manually. Google OAuth is available for sign-in when configured.
Honest current limits
API session still uses workspace headers in parts of the product — hardening (cookie/JWT session, non-spoofable admin) is on the 90-day trust roadmap. We disclose this instead of pretending enterprise SSO exists.
Admin surface
Operator admin is path-gated and not linked from public marketing. Invite review and Arena publish require admin controls.
Logging
Backend uses structured request logging with request IDs. AI routes have rate limits and daily budget guards.
What we store
Project profile, prompts, model answers from checks you run, crawl findings, loops, knowledge entries, invite applications, and usage for billing/ops. Public Arena never exposes real brand names without consent + pseudonyms.
What we don’t invent
We do not fabricate SERP positions, customer logos, or Arena scores. Empty samples show “No data yet”. Dollar impact stays Modeled until you connect analytics.
Models
AI checks run through OpenRouter on models you select. Provider terms apply to model inference. Evidence stays in your workspace records.
Export & delete
Workspace admins control project data. Legal requests and DPA terms cover processing. Contact via legal pages for formal data subject requests.
Commercial
Proof policy
We do not buy logo walls. A design partner blurb publishes only when there is a fixed prompt set, real checks, and (optionally) a consented Arena match. Until then, this section stays a frame — not fiction.
Slot reserved
Marketing Lead · B2B
Prompt set · multi-model baseline · ≥1 closed loop with re-check
Publishes only with measured evidence
Slot reserved
SEO / Growth
Crawl + AI gap map · share vs named rival on same prompts
Publishes only with measured evidence
Slot reserved
CEO / Founder
Board pack with Measured/Derived/Modeled tags · loop throughput
Publishes only with measured evidence
Ship-or-die list for diligence. Progress appears in Build Log — not vapor decks.
Now → 30 days
30 → 60 days
60 → 90 days
Category wedge
AI answer share is becoming a discovery channel. Most tools stop at screenshots. We stop at closed loops.
Workflow moat
Passport → checks → battle → loops → operator → board pack. Acquirers buy the operating system, not a single chart.
Evidence graph
Full answers, citations, findings, and re-checks create defensible data density per workspace.
Honest trust tags
Measured / Derived / Modeled is a diligence asset. Fake precision dies in acquisition review; labeled uncertainty survives.
Serious diligence gets a private data room (cohorts, retention, unit economics) — not a public vanity dashboard. Start with invite + Build Log + live traction above.
Access control
GetLoopLoop is invite-only. If you want access, submit a short request. We review applications manually and send an invite when it is a fit for Marketing, SEO, or leadership teams who will actually run the loops.