Directions Brain

Directions Saturday — acquisition wedge first, deck structure, and the licensing/partner workstream

6 min read

Directions Saturday — acquisition wedge first, deck structure, and the licensing/partner workstream

Source: raw/meetings/notetaker/2026-08-15-directiones-sautdday-11-10.json (Notetaker title: “Directiones Sautdday”)

Summary

Full-team Saturday session that settled the strategic sequencing question and produced the working plan for the next two weeks. The key call: the first thing to prove is an efficient SMB acquisition channel — traffic routed to a partner brokerage — not internal workflows or an own-license broker; joining a single partner to optimize their op-ex was explicitly rejected as an “FDE-for-brokers” path that proves nothing about being a unicorn (2026-08-15-acquisition-wedge-before-workflows). The investor deck was structured chapter by chapter (Problem → Vision → The Solution → Go-To-Market → Competition → Early Validation → Plan/Phases → Ask), ownership was divided, and two heavy research workstreams were opened: US licensing mechanics via founder-focused licensing-service companies, and finding a partner brokerage willing to take warm traffic — with tomer-shalom as a door-opener to strong American agencies.

Key takeaways

  • Acquisition Wedge > Workflows (the session’s core decision). Optimizing one partner’s op-ex makes the team field-deployed engineers for brokers and doesn’t prove a modern AI-native model; the unicorn path starts with a channel that reaches SMBs efficiently (later extensible to B2C), routes traffic to a wholesaler/brokerage partner, and builds licenses + internal broker + tools in parallel.
  • Nizan’s guardrail: acquisition can’t live in a void — end-user experience, pricing, and brand must be strong enough for NPS and a growth loop (the lemonade lesson), so brand and brokerage experience get real investment even while the “metal” behind them is a temporary partner. GTM model = cover-style: a home page that signals “something different in town,” integrations (canopy-connect-style), an MVP that could stand in ~1.5 months.
  • Vision formula: “We are building the AI Native Insurance Broker for SMBs” — reads the business (what it does, what could break it), composes coverage from the whole market, and runs the brokerage with AI so insurance stops being an annual document and becomes a system that grows with the business.
  • The Solution = three axes × jobs-to-be-done: customer-facing (verification, monitoring, documentation, intake, data gathering), broker↔carrier (market selection, quoting, comparative rating), internal operations (policy checking, day-to-day client service). From the existing ~20-job list, pick 5–6 core jobs and attach AI/tech to each. V1 narrative (Guy): a free insurance check-up — pull existing policies via canopy-connect/InsurGrid, build the full coverage picture, AI flags overpaying/underinsured/overlapping/missing protection, execute recommendations through own or network brokers.
  • Broken-incentives narrative: traditional brokers earn more when the client’s policy is worse (“we’re the angel, they’re the devil”); the fix is a neutral comparative rater + fixed-fee model where the broker’s take is independent of the plan chosen. (Tension with AX-INS-5’s named-coverage differentiation — see Open questions.)
  • Licensing workstream (Guy): investors will ask “how do you open a broker tomorrow morning”; map required licenses per state, the designated-responsible-person (DRLP) construct, and the licensing-service companies that get non-US-citizen insurtech founders licensed at known cost — meet them and price the package. Also: is there an anti-US path (Europe) if the US proves too complex?
  • Partner-brokerage workstream (Guy): find an existing brokerage that takes the warm traffic (for a share or fee) and could anchor a Letter of Intent in the deck; tomer-shalom meeting planned to open doors to strong American agencies/brokerages.
  • Validation workstream (Nizan): 4–5 qualitative calls with US brokerage owners/managers (red-state preference, e.g. Texas) on jobs, workflows, AMS, carrier relations, and appetite for platform-sourced traffic; an independent “state of the broker” survey; the prior 4-person qual sample must grow to ~10 with organized insights.
  • Traction-before-round debate: fake-door landing pages and PPC were weighed and deprioritized — Guy: two weeks of campaigns yields weak signal vs US companies with real customers, and maor-friedman hinted the deck alone might close him; Saar: a 10/10 deck still isn’t enough for US top-tier funds — deeper validation (brokers, experts, licensing) is the “something more.”
  • Round sizing (open): Nizan argues the super-broker vision needs ~4M drip; ties to the insurance-seed-fundraising debate.
  • Deck economics: chapters color-coded cheap (AI-completable overnight: problem data, competition tables) vs expensive (human-deep: Solution, GTM, Plan, regulation, partnership, interviews); chapters 2–4 (Vision/Solution/GTM) are “the heart” and get most of the two weeks. Ownership: Guy — Plan/Phases, regulation/licensing, partner; Nizan — Problem, Vision, Competition, GTM; Saar — technology and integrations layer (MS Teams/Zoom/Google Meet capture, Canopy/InsureGrid/comparative-rater integrations, aggregation, webhooks, AI explaining policy differences).
  • Brand/domain: early ideas (broker.insure, oneinsure, bestinsure, coverage; insure.com-class domains cost tens of thousands); final name+domain deferred, possibly with a brand consultant.

Decisions

  • MVP = acquisition play routing SMB traffic to a partner brokerage; internal broker and workflow tooling come after. Promoted to 2026-08-15-acquisition-wedge-before-workflows.
  • Validation for the next two weeks is qual research + survey + regulatory answers, not live acquisition campaigns (in-session working decision).

Action items

  • nizan-shifman + saar-arbel — write “The Solution”: sharp AI-native-broker definition, 5–6 core jobs-to-be-done across the three axes, connected to product value and tech. Due 2026-08-18.
  • saar-arbel — add the technical layer: integrations (MS Teams, Zoom, Google Meet, Canopy, InsureGrid, comparative raters), aggregation/webhooks, AI-driven policy comparison.
  • nizan-shifman — schedule 4–5 qual calls with US brokerage owners/managers (red-state preference). Due 2026-08-18.
  • nizan-shifman — build and run the “state of the broker” survey for Early Validation. Due 2026-08-25.
  • guy-barkat — regulatory research: per-state licenses, DRLP role, licensing-service companies + package costs. Due 2026-08-25.
  • guy-barkat — meet licensing/compliance service companies; summarize terms/costs for the Plan and Regulation chapters. Due 2026-08-25.
  • guy-barkat — set the tomer-shalom meeting to open doors to US agencies/brokerages (partner + LOI candidates). Due 2026-08-20.

Open questions

  • Does the acquisition-wedge + fixed-fee/neutral-rater narrative drift from AX-INS-5 (differentiation = named-coverage category creation)? The deck now differentiates on incentive alignment and continuity, not an owned category name — no decision has amended the axiom.
  • AX-INS-2 tension: phase 1 has the partner brokerage holding the licenses and executing; when exactly does BoR/customer ownership transfer to the company, and what stops the partner from keeping the book?
  • Fixed-fee vs commission economics: does fixed-fee survive contact with the ~10–20% commission annuity math in us-insurance-distribution-economics?
  • 20M raise — vision-scale runway vs stage-honest sizing.

Graph View

Backlinks