You are deciding whether “we have proprietary data” is a real moat or a pitch line. Many Seed and Series A AI teams confuse a growing log pile with defensibility. A data moat is a closed loop: scarce or hard-to-copy inputs, outcome-linked feedback, and a measurable product lift that brings more of the same data. Without the loop, you have a corpus, and competitors can catch up to a corpus.
What you are deciding
| Claim | Real moat signal | Weak signal |
|---|---|---|
| “We have lots of data” | Scarce, legally usable, outcome-linked; improves evals with usage | Public scrapes, customer-owned dumps you cannot retrain on, stale snapshots |
| “Data network effects” | Product value rises because nodes interact or shared operational memory compounds | More rows in a warehouse with no product lift |
| “Our model is the moat” | Domain evals and harness beat frontier APIs on the jobs you sell | Calling GPT/Claude with a thin UI |
| “Flywheel” | Each session creates corrections, exceptions, and ground truth you reuse | Accept/reject buttons nobody clicks; no trajectory store |
| “Customers stay for the data” | Switching loses their history of decisions and edge cases | Switching loses a chat transcript they can export |
fn-content has no verified atom yet for Seed–Series B time-to-minimum-viable-corpus, retention lift from proprietary loops, or quality-vs-quantity labeling spend. fn-content tracks it as benchmark request: AI data moat signals. Until then, use the named public sources above. Do not invent a “typical” dataset size.
Scale effects are not network effects
Casado and Lauten separate network effects (value rises because participants interact over a shared interface) from data scale effects (more training or retrieval data improves predictions even when users never interact). Most AI application pitches describe the second and call it the first.
Their enterprise observation still maps to application AI in 2026:
- Minimum viable corpus is cheap relative to later data. You can bootstrap with crawl, customer trade, transfer learning, or synthetic data. That gets you into the market without giving you a moat.
- Acquisition cost rises. Unique long-tail examples get harder to find, secure, and label.
- Incremental value falls. New batches overlap existing coverage. Past a domain-specific asymptote, more of the same does little.
- Freshness decays. Streets, policies, buyer language, and edge cases go stale. Keeping the corpus current is ongoing work, not a one-time scrape.
The Eloquent Labs support-chatbot curve they cite is domain-specific, not a universal law. Use it as a caution: know your coverage curve before you tell investors the moat widens forever.
When data defends
| Condition | Why it holds | Who frames it |
|---|---|---|
| Proprietary or exclusive sources | Competitors cannot buy the same feed; vendor scrutiny itself filters rivals | a16z: secure proprietary sources; compliance as a gate |
| Outcome-linked labels | Corrections and results train the next eval, not vanity metrics | Sequoia: trajectories → evals → harness fixes |
| Quality before volume | Narrow, high-fidelity loops beat broad mediocre dumps | Bessemer principle 10; EvenUp human review example |
| Workflow + multimodality | End-to-end job with integrations and mixed inputs beats a wrapper feature | Bessemer principles 2 and 8; RAG on industry data as a floor |
| Operational memory | History of decisions, exceptions, approvals, and failures stays in-product | Sequoia online-learning loop; operator judgment on compounding benefit |
| Instant user reward | Users label when feedback improves their work now | FounderNexus session: feedback that pays instantly, not altruism |
Bessemer’s Vertical AI Part IV warns that models will not stay a moat as infrastructure costs fall. Ask why your product beats what a buyer can assemble from public models and public data. Industry-specific retrieval, compliance, and end-to-end workflows are the practical answers they emphasize.
Build the loop in four layers
| Layer | Operator move | Public anchor |
|---|---|---|
| 1. Eval | Turn real work into graded tasks (prompt, context, grader). Stop vibe-checking alone. | Sequoia / Harvey: benchmark before you own more of the stack |
| 2. Capture | Log trajectories: context in, tools called, output, edits, undos, retries. | Sequoia online learning; failed task → new eval |
| 3. Improve | Pick the lightest fix: RAG/context for missing facts; SFT for format; preference for taste; RL for specialized skill; distill for cost/latency. | Sequoia / Lin Qiao framing in Huang’s piece |
| 4. Contract | Trade early discounts for usage minimums and structured feedback obligations when you need the first turns of the flywheel. | Operator judgment from founder rooms (FounderNexus session) |
Two sources of differentiated data that repeatedly show up in operator rooms (without closed-session numbers): collect what nobody publishes, and apply decades-style domain fluency about what is signal versus noise in a niche. Commodity enrichment feeds are table stakes.
Bessemer’s EvenUp lesson fits layer 1–3: EvenUp chose early human review as a quality investment, not a failure to automate, and scaled once the feedback was trustworthy.
Decision table: invest in the loop or not
| Situation | Invest in a data loop now | Wait / do something else |
|---|---|---|
| Vertical workflow with repeated edge cases and human corrections | Yes. Capture trajectories and grade them. | — |
| Thin chat UI on a frontier model with no system of record | — | Prove retention and a painful job first (Bessemer: high-ROI product before data theater) |
| Buyer will not let you use their data for training | Build per-tenant memory and evals they own; or redesign the value so you do not need cross-tenant training | Do not claim a cross-customer moat you cannot legally create |
| Base model releases erase your fine-tune every quarter | Shift investment to harness, evals, and proprietary context (Sequoia “why now” on open weights) | Stop treating last quarter’s weights as the company |
| You can buy the same dataset as three competitors | Compete on GTM, workflow depth, and brand (a16z holistic defensibility) | Do not pitch “our data” as the story |
Worked situations
Seed, vertical workflow AI, ten design partners. Stand up a private eval of 50–100 real tasks before you brag about a moat. Log every accept, edit, and undo. Trade a discount for weekly structured feedback. Cite Bessemer quality-over-quantity and Sequoia’s eval-first path. Do not tell Series A investors you have network effects because the Postgres table is growing.
Series A, usage up, win rate flat vs a wrapper competitor. Audit whether new data hits the long tail or only duplicates the head (a16z distribution warning). If lift is flat, invest in scarcer labels and harness fixes, not another scrape. Revisit pricing so outcomes you improve are the unit you charge for (seat vs usage vs outcome).
Series B, “our model is the moat” in the board deck. Replace the slide. Show domain eval delta vs frontier APIs, trajectory coverage, and switching costs from operational memory. Bessemer: models commoditize; multimodality and workflow integration do not as fast. Market the job, not the model (market AI without saying AI).
Sources
- a16z, The Empty Promise of Data Moats. Martin Casado and Peter Lauten, 9 May 2019. Scale effects vs network effects; rising acquisition cost / falling incremental value; Eloquent Labs chatbot coverage curve (~20% effort → ~20% coverage; ~40% intent asymptote in that study); minimum viable corpus; proprietary sources and holistic defensibility.
- Bessemer Atlas, Part IV: Ten principles for building strong vertical AI businesses. 28 Jan 2025. Principle 10 quality over quantity (EvenUp human review); principle 8 multimodality / models not a reliable moat; principle 2 end-to-end workflows vs commoditized features; industry-specific RAG as a foundational layer.
- Sequoia, Own Your Intelligence: A How-To Guide. Sonya Huang, 19 Aug 2026. When proprietary data argues for owning stack slices; evals before post-training; harness and trajectories; Harvey Legal Agent Benchmark (1,200+ tasks, 24 areas, 75,000+ rubric criteria); research team of seven.
- FounderNexus. Founder sessions. Data as the ingredient a startup can own; unpublished collection and domain fluency; feedback that rewards the user instantly; small data still signals. Not a survey. No closed-session numbers.
- fn-content #22. Benchmark request: AI data moat signals (MVC size, loop latency, retention lift from proprietary loops).
Related
- How to market an AI product without leading with AI
- Seat vs usage vs outcome pricing for AI SaaS
- AEO vs SEO for B2B SaaS startups
Founders who have tested whether their loop compounds or only piles up logs will pressure-test yours in a FounderNexus session.