Field Notes - Nov 26, '25

Executive Signals

Proof beats primer: open with traction and slope, not market explainers
Polish in quarantine: isolate demo UI changes; cherry-pick safe presentation tweaks
Partition the risk: name de-risked facts; milestone the remaining unknowns
Partners are the new pipeline: free data management pulls labeling revenue when work starts
Forecasts need walls: committed revenue separate from speculative opportunities

Open by proving you’re real. First slide should pass the drunk test: who you are, one‑line what you do, rounded traction, and a single line of founder credibility. Show the product, not paragraphs. Keep the revenue chart full-bleed and emphasize slope. Close with a selective posture: you’re building regardless; capital accelerates R&D and GTM.

Use a 6‑slide flow: what/traction, problem, product screenshots, traction chart, GTM/why now, ask
Make the revenue chart the whole slide; label axes; highlight the slope
End with “building anyway” posture and a crisp use of funds

Craft the De‑Risking Narrative

Investors buy partitioned risk. State what’s de‑risked—customers, revenue, founder/market fit—then list the open risks with milestones to close them. Replace “can be big” with “will make it big.” Wear two hats with clarity: CEO grows the pie; shareholder decides when to sell.

List 3 de‑risked facts; name 2 live risks with dated milestones
Tie open risks to measurable outcomes, not vibes
Keep the “CEO vs shareholder” tradeoffs explicit in board materials

Define Venture‑Scale Paths Beyond the Core Tool

If core 3D tooling caps at tens of millions ARR, show the stack that compounds. Sequence adjacencies you can credibly win: AV 3D dominance, then robotics/industrial, then data and LLM services to the same buyers. The “why us” is distribution, data exhaust, and workflow lock‑in.

Gate each step with thresholds for ARR, gross margin, and net retention
Pre‑write the qualification tests that greenlight each expansion
Map the upsell path from workflow ownership to data/LLM services

Open‑Source Integration: Outcome First, Architecture Later

If today’s product leans on OSS, don’t front‑load that in a short pitch. Position it in diligence as a learning engine that funds development. The roadmap is proprietary automation that compresses label‑hours per scene and expands ACV.

Sell outcomes and traction first; discuss OSS provenance in diligence
Narrate “integration now → automation moat next” with concrete milestones
Tie milestones to unit savings: hours per scene down, unit cost down

Marketing

Frame Labeling as Durable Infrastructure

For AV and robotics, the world is non‑stationary—roads, signs, weather, behaviors keep changing—so data is never “done.” Combine that with scaling laws and you have durable, picks‑and‑shovels demand independent of any single OEM. Show software, not services: curation, labeling, and validation workflows.

Emphasize “environment evolves → ongoing relabeling” in positioning
Use UI screenshots to prove it’s software‑first, not a body shop
Avoid vendor name‑drops in five‑minute formats; stick to principles and proof

Charts That Land Without Lying

People skim. Make charts legible and persuasive. Start axes at a sensible floor, focus on the last 6–9 months, annotate two inflection points, and put current MRR in the top‑right. Keep raw numbers available for diligence.

Use full‑bleed charts with large labels and minimal gridlines
Annotate causal moments like channel launches or feature releases
Park detailed tables in the appendix for reviewers

Sales

Channel BPOs; Monetize Labeling; Give Data Management Free

Let labeling providers pull you into accounts. Offer data curation and management free to seed usage and stickiness; meter when labeling begins. Build orchestration that lets enterprises coordinate multiple vendors in one place. Smooth workforce volatility with contracts that convert to FTE after sustained demand.

Ship free data management; meter on label‑hours or task units
Build workforce and task orchestration for enterprises and BPO partners
Contract until three consecutive months above a hiring threshold, then convert

Separate Pipeline From Revenue You Can Count

Do not contaminate near‑term forecasts with logo soup. Show committed or contracted revenue for next quarter separately from late‑stage and future opportunities, each with time horizons and conservative ranges. This reads stronger and is diligence‑ready.

Split dashboards: “Committed next quarter” vs “Future opportunities”
Assign conservative dollar ranges to future opportunities; no speculative stacking
Review weekly to keep stages and timing honest

Product

Holiday‑Week Demo Rhythm That Lands

Thin‑attendance weeks reward observable progress—dry runs, dashboard reviews, and UI polish—not fragile integrations. Lock the golden path the day before the demo and apply AI‑generated UI improvements where they help comprehension. Gate the demo on reliability signals, not optimism.

Publish a golden‑path script with steps and screenshots by end of day prior
Run a 10‑minute, no‑commentary dry run; fix “first 60 seconds” friction
Gate on two checks: zero auth errors in last 24h and sub‑2s primary view load

Engineering

AI Pair‑Design for Demo Polish, Safely Isolated

Stabilize the core and quarantine “make it pretty.” Sit with the dashboard, generate UI variants with AI, and cherry‑pick only safe presentation changes into the demo branch. Keep rollback trivial so polish never risks the narrative.

Use 1–2 throwaway branches limited to copy, CSS, and layout
Freeze the demo branch 24–48 hours pre‑demo; allow only cherry‑picks
Target under 15 minutes to revert any UI change; keep rollback steps in the runbook