Directions Brain

Directions — Broker Jobs-to-Be-Done Map (Saar/Nizan)

4 min read

Directions — Broker Jobs-to-Be-Done Map (Saar/Nizan)

Source: raw/meetings/notetaker/2026-08-16-saar-nizan-11-05.json

Summary

Working session (Notetaker participant metadata empty — core-team gate overridden per user request; transcript confirms only saar-arbel and nizan-shifman) that decomposes the end-to-end commercial-brokerage workflow into nine sequential jobs-to-be-done, from first-meeting qualification/exposure mapping through commission reconciliation, for the product now referred to throughout as “Actually” (previously the working name “Ainsure” — see smb-insurance-portfolio-brokerage). The session produced the full broker-jobs-to-be-done-map synthesis and a plan for the investor jobs-map slide (heat map + old-world/new-world comparison + value summary). It closes with a same-day research push: booking US broker interviews via userinterviews.com for tomorrow and researching brokerage infra/API vendors (e.g. CoverForce).

Key takeaways

  • Nine jobs defined end to end: Qualification & Exposure Map → Intake & Data Gathering → Market Selection & Submission → Quoting → Binding, Onboarding, Billing & Certificates → Endorsements & Mid-term Changes → Claims Advocacy → Renewal Prep & Retention → Commission Reconciliation. Full detail in broker-jobs-to-be-done-map.
  • Market selection vs. quoting, sharpened as distinct jobs: market selection/submission is a judgment call (which of ~100 carriers are eligible for this risk, e.g. skip Lemonade, submit to Travelers) narrowed to a shortlist; quoting is mechanical execution (retrieving actual price numbers via self-service portals) once submissions are in.
  • Certificates of Insurance (COI) flagged as a high-volume, low-effort win: ~400M COIs issued per year in the US; Actually’s plan is to auto-generate a COI for every client from day one instead of the broker producing it manually on request.
  • Commission reconciliation named “the silent revenue leak”: agencies frequently under-collect earned commission because matching carrier statements against books is manual and painful — the pain is “invisible until quantified,” making automated statement-matching a candidate operational alpha source.
  • Claims decoupled from binding-day scope: first version may just route the client to the right carrier claims channel/index rather than handling the claim inside the platform, since some clients file directly with the carrier (Nizan’s example: Menorah) and others go through the broker.
  • Investor narrative structure locked for tonight’s prep: opening statement → jobs-to-be-done heat map (grouped, e.g. customer/carrier/office in a Venn framing) → old-world vs. new-world zoom-in on select jobs → value summary (growth, margins, efficiency, speed, better coverage — team said “six” value points but only five were named in-session, sixth unspecified). Saar’s explicit guardrail: stay at the jobs/value level, do not drop into microservices detail for investors.
  • Legal-adjacent intake handling: for intake/data-gathering’s litigation-like evidence-collection cases, the current lean (per a Huri suggestion) is to gather the information and refer out to lawyers for revenue share rather than represent clients in court.
  • Research cost reality check: userinterviews.com/Listenna-style recruiting runs ~$98/session (platform fee + participant incentive) — high enough to trigger a look at cheaper alternatives (Dscout, UserTesting) before scaling the campaign.
  • Known terminology gaps flagged for follow-up with Guy or an outside expert: “binder,” agency-bill vs. direct-bill mechanics, and portal-raking — specifically to avoid duplicating or mis-scoping jobs 4 (market selection) and 5 (quoting).

Decisions

  • No standalone team decision locked this session — this was drafting/definitional work refining the jobs map and investor narrative structure. See broker-jobs-to-be-done-map for the resulting synthesis.

Action items

  • saar-arbel — Schedule interviews with US brokers for tomorrow via userinterviews.com, 11:00–20:00 Jerusalem time, against open calendar slots.
  • nizan-shifman — Prepare, by tonight, the full jobs-to-be-done heat map (grouped, old-world/new-world zoom-in comparison, value summary) for the investor deck.
  • saar-arbel — Research brokerage APIs/infra providers (e.g. coverforce) — sign up where relevant and evaluate what can integrate into Actually.
  • nizan-shifman — Stand up and configure the userinterviews.com research campaign for US-based independent commercial-lines insurance brokers (role/seniority filters, Google Meet format); also price-check cheaper alternatives (Dscout, UserTesting/“LisnA”) given the ~$98/session cost.

Open questions

  • What exactly counts as “binder,” and how agency-bill vs. direct-bill billing routes differ — needs an expert pass (Guy or an external broker) before finalizing jobs 4–5 scope.
  • Where the line sits between claims handling Actually owns directly vs. routes/decouples to the carrier or a partner — deferred decision.
  • Whether the sixth named “value” point (beyond growth, margins, efficiency, speed, better coverage) gets specified before tonight’s deck pass. (unknown — needs source)
  • Cheaper qualitative-research vendor vs. userinterviews.com/Listenna, given per-session cost pressure on the discovery budget.

Graph View

Backlinks