Directions Brain

Directions — Value-Led GTM, Price Levers, and Book Purchase

7 min read

Directions — Value-Led GTM, Price Levers, and Book Purchase

Source: raw/meetings/notetaker/2026-08-20-directions-19-00.json (Hebrew, 22:00 Jerusalem, immediately after the jasmyne-mcdonald interview)

Summary

The session that converted a week of discovery into a plan. B2C is confirmed, GTM is to be derived from the values the technology actually delivers rather than from a slogan, three concrete price-reduction levers are named, and buying books of business emerges as the team’s favoured acquisition channel — hours after an operator gave them real valuation numbers. A VC deck for ~30 funds and a Huri meeting are scheduled, along with a demo intended to prove capability rather than describe it.

Key takeaways

  • Decision: stay B2C, and derive GTM from values. The named value set is efficiency, transparency, price reduction, holistic coverage, speed, and service quality. The GTM must be built on one or a combination of these, must be a sharp message, and must be backed by a real ability to deliver the promise — explicitly not a slogan.
  • “We’re efficient and cheap” is not a GTM. Saar’s challenge was the sharpest moment: even if the system is 30% more efficient than an existing broker and can source a better, cheaper policy, that still does not say how you spend the first $5M on acquisition. Nizan’s answer: start segment research from the value — find who is most price-sensitive and most poorly served today, then pick an initial segment where a real delta can be promised (his example: ~30% price reduction for construction).
  • Three price-reduction levers, now explicit:
    1. Smart initial shopping — appetiting and quoting across the whole market rather than the shop across the street.
    2. OpEx pass-through — return part of the operational saving to the customer via cashback, discounts, or added services.
    3. The renewal lever — at renewal, deliver a price decrease instead of the customary increase, which also drives retention.
  • The renewal lever rests on a claim our own evidence contradicts. Saar argued brokers barely remarket after year one because it isn’t worth it to them. But greg-ehly described a systematic 60-day pre-renewal review with proactive remarketing whenever a claim-free account jumps double digits — deliberately absorbing lower commission for retention. Flagged rather than resolved: the lever may still work, but “incumbents don’t remarket” is not a safe premise to put in front of investors.
  • Guy doubts price is the selling point at all. His candidate differentiator is an AI-led digital UX with human backup for a small slice — a Robinhood-style brokerage experience, which he notes essentially no broker offers today. Note this sits alongside, not behind, the price levers; the session did not choose between them.
  • Book of business purchase became the favoured acquisition channel. Guy proposed the lean version: buy policies only, without the org structure, secretary, and overhead — net book, loaded straight into the digital system. His arithmetic is that this can work even at 60% retention, because the purchase price sits near estimated LTV, which would make book purchase a genuine acquisition channel rather than a roll-up. Saar called it his favourite of everything discussed and “very Monday” — workflow-oriented. Nizan’s addition: buy books, improve customers’ prices at renewal, retain them, and generate word-of-mouth referrals from the better service.
  • This is AX-INS-3’s revisit trigger. The axiom says “no agency roll-ups — we build the book, we don’t buy it,” with a revisit trigger for a pure book purchase in the exact segment at distressed pricing. The team is now distinguishing exactly that — buying an agency versus buying a net book — and Huri is noted as disliking classic roll-ups. No axiom has been amended; this needs a logged team decision.
  • The cash-flow problem is named and unsolved. In brokerage the money arrives only after a year, so there is no immediate cash flow, churn before renewal is a live risk, and the business “lives at current plus one year.” Requires accurate forecasting in the model.
  • The VC objection they expect: buying a book generating 70K looks like value destruction. Nizan’s answer is that this is only true absent unit economics showing how price deltas, retention, and efficiency still leave the business profitable — which is now a deliverable, not an argument.
  • Backoffice and AMS read: the market is ruled by 20-year-old AMSs with no new market leader, and new entrants build workflows around the AMS rather than replacing it — which the team reads as space for an integrated player. A related idea: a free app performing repetitive customer tasks (cancelling a policy, connecting to carrier APIs, payment reminders, chasing missing documents) as organic growth. Guy sourced this from analysing Reddit threads where agents complain about customer requests they can’t service efficiently, estimating the system could handle three quarters of them.
  • Huri’s enthusiasm is specific and worth building to: he responded strongly to AI explaining coverage choices to the customer — why Coverage A and not B, why one carrier over another — presenting options and exclusions transparently so nobody is surprised at claim time. Guy sees showing 20 options with reasoning as both good service and an acquisition hook for customers previously burned by brokers who never explained.
  • Referrals are earned, not featured. Nizan: word-of-mouth comes from something genuinely good happening in the product, not from a referral feature — so GTM should be inside-out, grounded in product quality.
  • Licensing is the easy part. Guy will book a US broker-licensing service early next week; an individual agent/brokerage license is “a piece of cake” (about two weeks of exams) compared with the difficulty of buying a book. The license can be registered to an American employee as DRLP. (The session also floated “buying a person” to put their name on the license — recorded verbatim as said; this would need real legal review before it becomes a plan.)

Decisions

  • Continue with B2C, with GTM derived from the delivered values rather than from a slogan.
  • Price alone is insufficient as a GTM; it must be translated into segment-specific messages with a quantified promise (e.g. ~30% reduction for construction).
  • Build a staged product plan (V1/V2), with V1 as automatic appetiting plus quoting on existing integrations.
  • Build an appetiting/quoting demo on integrations and APIs, to show real capability to Huri and VCs rather than describing it.
  • The deck targets ~30 VCs and must be exceptional in the opening and main body.
  • Meet alon-huri at the end of next week, with a short prep call on Sunday, arriving 70–80% resolved.

Action items

  • nizan-shifman — roll 80–90% of existing knowledge into the VC deck: how brokers make money, where the market is broken, jobs to be done, and the values the system delivers. Due: 2026-08-30.
  • guy-barkat — set the alon-huri meeting for end of next week (Wednesday, or after Saar’s wedding) and make contact Sunday for a short prep call. Due: 2026-08-23.
  • nizan-shifman — coordinate the team’s next working session in the group chat (Saturday afternoon or after the game) to continue on the deck and the model. Due: 2026-08-22.
  • guy-barkat — book a call early next week with a US broker-licensing service, and examine registering the license to an American employee/DRLP. Due: 2026-08-24.
  • saar-arbel — begin researching and sketching the quoting and appetiting flow, including initial use of integrations and APIs, and the V1 architecture. Due: 2026-08-30.

Open questions

  • Price versus experience was not decided. Nizan’s levers and Guy’s Robinhood-UX thesis are both live, and the GTM message can only carry one headline. jasmyne-mcdonald’s churn ranking — mishandling first, price second, location third — is a data point against price as the lead.
  • Is the 60%-retention book-purchase arithmetic sound? It assumes purchase price ≈ LTV and that a bought book survives a transition. jasmyne-mcdonald’s own transition failed for lack of seller support. See bought-books-convert-cheaper-than-paid-acquisition.
  • How do the price levers survive the aggregator comparison? Saar raised it directly: US aggregators are already connected to every carrier and already return competitive quotes.
  • The renewal lever’s premise that incumbents don’t remarket is contradicted by greg-ehly. Needs checking before it reaches the deck.
  • “Buying a person” to hold a license needs legal review before anyone repeats it externally.

Graph View

Backlinks