Directions Brain

connect-businesses-multi-vertical-api

5 min read

Connect Businesses — Multi-Vertical Unified API

Definition

Exploratory direction surfaced by guy-barkat in 2026-05-31-directions-connect-businesses-and-brain-scaling: a multi-vertical, “Plaid-style” platform that lets any app with distribution plug in any third-party service via a unified API + standardized contracts + bundled legal/compliance. Guy frames it as Taboola-for-services: matchmaking between demand (apps with eyeballs) and supply (companies with capability), reducing the per-deal N×N integration cost to N. The Brain is not central to this direction — this is a candidate alternative direction, not a Brain wedge.

Key points

  • Trigger anecdote. Guy is leading etoro’s API partnerships (Public APIs / Identity Trading). Two strategic deals are live: etoro CEO Yoni Assia talking with papaya-global (global payroll) about a wallet that routes a slice of salary into investments via etoro’s API, and deutsche-bank approaching etoro for a similar play. Same morning, two parallel data points landed: Israeli supermarket victory selling apartments and rami-levy / osher-ad selling cars (Sportage via the Rami Levy app). Pattern Guy noticed: every low-tech / medium-tech platform with distribution is suddenly trying to consolidate categories.
  • Why now. Cost of integration dropped sharply (AI). Distribution is the only durable moat for many incumbents — their core product margin is shrinking. Adding a category is suddenly cheap, and “everyone is on everyone” (super-app behavior bleeding from Asia/X into Israeli retail). A platform that turns N×N into N + handles legal/compliance/SSO/T&Cs/customer-support could ride the wave.
  • Reference templates.
    • Plaid — unified API for US banking. Single-vertical (banking), worth >$15B, took years per-bank. Their “Acquired” podcast episode is on guy-barkat’s reading list.
    • SnapTrade — Plaid-equivalent for trading platforms. Currently powers etoro partnerships (e.g. Seeking Alpha connects to eToro through SnapTrade without ever talking to eToro). Single-vertical fintech.
    • The thesis: stay multi-vertical instead of vertical-specialist. Same shape, broader surface, ride the cross-category demand.
  • Matchmaking analogy (Nizan). Like Taboola: it doesn’t care what the publisher’s business is — anyone with traffic gets monetized. Same here, but for full services rather than ad inventory: any app with distribution gets a catalog of plug-in services.
  • Possible product wedge. Bundle “do the integration once” with “we already have the legal/compliance contract + SSO + ToS standardization in place.” Customer-support handoff between supplier and consumer is the load-bearing operational problem (already solved by Plaid/SnapTrade in their respective verticals — proof it’s tractable).
  • ICP candidates. Not Silicon Valley high-tech (they will build internal). Sweet spot: low-tech / medium-tech platforms with millions of users and weak product-engineering depth — supermarkets, retail super-apps, regional fintech apps, CRM platforms looking to bolt on capabilities without a year of integration work.
  • Adjacent failure mode (Saar’s pushback). For a platform like wolt doing a partnership with a giant like El Al, the bottleneck is legal/commercial (closing the deal), not technical (building the connection). The unified-API thesis only saves time on the technical leg, which may be the easier leg. Counter-evidence (Nizan): even Monday CRM × Facebook never got prioritized internally — multiple engineering teams will not sign up for one-off integrations even when they want them.

Why-now (Saar’s challenge: “this would have been a great idea five years ago too”)

Three changes:

  1. Cost of building a multi-vertical platform dropped — what Plaid took years to do, this could ship in weeks/months.
  2. Larger appetite specifically among low-tech / medium-tech players to do integrations they previously skipped on cost grounds.
  3. Distribution is increasingly recognized as the new moat — incumbents are pressured to monetize it before they get eaten by AI-native challengers (this matches nizan-shifman’s “long-tail of startups will disappear; only big-distribution incumbents survive” thesis from the meeting).

Evidence

Open questions

  • What’s the patent / defensibility? Saar pressed multiple times: “How are you not just a services consultancy?” Guy’s cut: speed-to-market is the moat (low-tech ICP can’t run a 6-vendor compliance + integration process; you can stand them up in days). Unverified.
  • Cost of customer acquisition. No PLG motion; pure B2B sales on both sides. Each new supplier and each new consumer needs onboarding + legal. Chicken-and-egg marketplace dynamics.
  • TAM vs take-rate. Plaid is one vertical worth $15B+. If multi-vertical is real, the ceiling is much higher — but the required take-rate to get there is unknown.
  • Vs the vertical-use-case-led-brain direction. This is a non-Brain direction. Treat as a parallel exploration, like prediction-markets-regulatory-play. Team holds three open directions now.
  • Suppliers bypassing the platform. Once a supplier and consumer know each other through the platform, what stops them from cutting the platform out at renewal? Plaid/SnapTrade have presumably solved this; need to study how.
  • ICP shape — supplier side or consumer side? Guy’s instinct is to start with apps that already expose APIs (suppliers); Saar pushed back that low-tech consumer apps are exactly the ones that don’t have APIs yet, so the platform may need to build those too — which collapses back into a services company.
  • First mover already in market? Make / Zapier are nearby (Nizan raised both). The differentiation Guy claims: those are technical glue; this is bundled commercial + legal + compliance + UI sponsorship.

Graph View

Backlinks