THE COMPANY-BUILDING FIELD NOTEBOOKRESEARCH EDITION / SEPTEMBER 2026
Startup
Research.
Search
POST-MORTEM LIBRARY / AI-era failures (2023–2026)

Jasper (partial failure — valuation collapse, company survives)

  • Company — Jasper AI (formerly Jarvis, formerly Conversion.ai)
  • Sector — AI content generation / marketing SaaS
  • Founded / Died — 2020 / did not die; internal valuation cut and strategy reset from mid-2023
  • Lifespan — Ongoing
  • Capital raised$125M Series A (October 2022, Insight Partners, Coatue, Bessemer) at a $1.5B valuation; ~$131M raised in total
  • Peak scale — Reported ~$75M ARR and ~100,000 customers in 2022 — among the fastest early ramps ever recorded for a SaaS company. Both figures are company-stated.
  • What it built — A marketing copywriting tool built on OpenAI's models, with templates and workflows layered on GPT-3.
  • Stated cause of death — Jasper did not fail. In July 2023 it cut its internal 409A valuation, reduced its revenue forecast, and later replaced its CEO. The company positioned this as a shift from a point tool to a marketing platform.
  • Evidenced cause — ChatGPT launched on 30 November 2022 — six weeks after Jasper's $1.5B round — and gave away free the core of what Jasper charged for. Growth stalled almost immediately. The condition is thin-wrapper exposure: when the primary value is prompt engineering and UI over a third-party model, that model provider's consumer product is a direct substitute. Jasper survives because it moved fast into brand-voice controls, enterprise workflow and compliance — the parts a general chatbot does not provide.
  • Warning signs visible earlier — The gross margin structure (paying per-token to a supplier who also sells to your customers) was visible at the time of the round.
  • Money outcome — No liquidation. Late investors are very likely underwater on paper. Employees who joined on 2022 valuations hold options struck above the reset price — a common and rarely discussed form of employee loss in a down round.
  • Post-mortem qualityNot applicable, and that is the point. Partial failures produce no post-mortems at all, which means the most common outcome in the AI era — a company that survives at a fraction of its peak mark — is almost entirely undocumented.
  • Transferable lesson — Build where the model provider will not go.
  • SourcesMaginative on the valuation cut · AI Beat · The Information/secondary coverage of the CEO change

Read the wider evidence

This entry is reproduced from the supplied research, with its inline source links retained. It has not been independently re-reported for this website conversion.

Read the complete chapter, source list, and methodological notes →
← Back to post-mortems