GhostView

The problem is specific and timely. A distinct new cohort of builders has emerged in the last 18-24 months: non-engineers and light-engineers shipping apps built largely by AI code generators. They ship fast, get modest early traffic (50-2,000 users, per the founder — a plausible range), and then stall because they have no instrumentation discipline and no analytics muscle memory. Mixpanel/Amplitude are overkill and require event schema design these users won't do; Hotjar and Clarity give recordings but not synthesis. The founder's second observation — that nobody watches 47 recordings — is the real pain. Session replay tools have always had a consumption problem, and it gets worse for solo builders with no time. The third observation is the sharpest: when an AI agent wrote the code, returning friction evidence as video breaks the loop, because a human has to translate the video into a prompt before the fix happens. Structured JSON that can be pasted into Cursor closes the loop agent-to-agent. That is a real insight about how this cohort actually works, not a generic 'AI-powered analytics' pitch. Where the problem gets softer: the target user is a segment famous for churn and low willingness to pay. Many vibe-coded apps never reach revenue, and their builders abandon them within weeks. So the pain is real but the wallet behind it is thin and transient.

The product is a lightweight script tag that captures clicks, scrolls, form fills and full sessions, plus an LLM synthesis layer that reads across all sessions to surface drop-off points, rage-clicks, and behavioural anomalies with citations back to specific sessions. Ninety-second install and a no-signup live demo are strong go-to-market choices for this audience — they will not sit through a sales call or a schema-mapping exercise. The genuine differentiator is not the recording (commoditised) but two things: (1) synthesis across the whole user base rather than per-session replay, and (2) JSON output designed to be pasted back into an AI code editor. The first is a UX bet — that builders want an editorial summary, not a library of tapes. The second is a workflow bet that could become a real moat if GhostView builds tight integrations with Cursor, Lovable, v0 and Bolt (e.g. one-click 'send friction report to Cursor'). Without those integrations, the JSON angle is a nice demo but not a lock-in. Defensibility is the honest weakness. PostHog already has session replay, is open-source, and has been shipping AI features aggressively; Hotjar and Microsoft Clarity are free or near-free. Any of them could ship 'AI summary of your sessions' as a feature. GhostView's edge has to come from being obsessively better at the vibe-coder workflow — install time, integration with AI builders, and the quality of the synthesis prompt chain — not from the underlying recording tech.

Bottom-up: the addressable population is builders on AI app platforms. Lovable, v0, Bolt, Replit and Cursor collectively have, based on what I know through my training data and public statements, several hundred thousand to low millions of active builders, though the share who actually ship apps to real users is much smaller — plausibly 50-200k globally at any given moment. If GhostView captures 2-5% of shipped-app builders at, say, $20-40/mo blended ARPU (their published post-beta pricing will pin this down), that is a $10-50M ARR ceiling on the narrow wedge. That's a reasonable angel outcome zone — not a venture-scale prize, but a credible $5-30M ARR business that could be acquired by PostHog, Vercel, Replit, or an analytics incumbent. The expansion path — moving upmarket into seed-stage startups that have raised money and want lightweight analytics — is where the real dollars sit, and it's a natural adjacency. But that's also where the incumbents fight hardest.

PostHog (open-source product analytics + session replay, YC alum, well-funded as of my training data): the most direct competitor. GhostView must be meaningfully better on install friction and AI-builder workflow, because PostHog will add AI summarisation if it hasn't already. Hotjar (session replay for marketers, owned by Contentsquare): serves a different buyer (marketing teams on established sites), so less direct but sets pricing expectations. Microsoft Clarity (free session replay): the price floor. GhostView cannot compete on cost of recording — it competes on synthesis and workflow. June.so (product analytics for B2B SaaS, positioned as auto-generated reports): a good analogue for the 'we synthesise for you' positioning, though targeting a different segment. Vercel Analytics / Lovable's own built-in analytics: platform-native alternatives that could eat this wedge if the AI-builder platforms decide analytics is table stakes. As of my training data, PostHog had raised a Series C and Contentsquare had acquired Hotjar; the AI-native analytics category (June, Mixpanel's AI features, Amplitude's) was heating up but nobody was specifically targeting the vibe-coder cohort.

1) Willingness to pay from vibe-coded app builders is unproven and probably low — many are hobbyists. Mitigation: the founder should validate paid conversion during beta with 20-30 pilot users; if fewer than ~15% convert at published pricing, the wedge needs rethinking. 2) Platform risk — Lovable, v0, Bolt, or Cursor could ship native analytics and evaporate the wedge overnight. Mitigation: become the default recommendation via partnerships or integrations before platforms build their own; no guaranteed defense. 3) PostHog or Clarity ship 'AI session summary' as a checkbox feature. Mitigation: GhostView must own the AI-builder workflow (JSON to Cursor, direct integrations) so it's not just 'recording + summary' but 'closed loop back to the agent that wrote the app.' 4) Churn — this cohort abandons projects quickly, so even paying customers may churn in 60-90 days as they move to the next side project. Mitigation: none obvious; the founder should measure cohort retention early and honestly. 5) Solo/small team execution risk — no team info provided; unknown whether the founder can build both the recording infra (non-trivial: rrweb-style DOM capture, storage costs) and the AI synthesis layer.

1) Vibe-coded app builders will pay $20-50/mo for synthesised session insights — validated by 15+ paid conversions from the beta cohort within 60 days of turning on billing. 2) The JSON-to-Cursor loop is a real workflow, not a demo gimmick — validated by observing whether beta users actually paste GhostView output into their AI editor and ship fixes, measurable via a simple 'did this fix ship' survey or integration telemetry. 3) GhostView can hold a defensible position against PostHog adding AI summaries — validated by whether the founder ships 2-3 native integrations with AI builder platforms (Lovable, v0, Bolt, Cursor) in the next 90 days that PostHog cannot easily replicate.