Product differentiation, hiring founders, and going deep on hard problems

Executive overview

Most products fail not because they lack differentiation or quality in isolation, but because they lack both simultaneously in a way that matters to users. Ayo Omojola distils lessons from scaling Cash App from sub-50K to 50M+ monthly actives, then applying them at Carbon Health.

The throughline: pick a domain that matters, be different and better in a way users value, keep teams small and senior, and don't stop asking questions until you reach the actual end of a problem.

The compound effect of doing 10 things right — not just one — is what creates lasting differentiation.

The differentiation framework

• Different alone is cheap — you just have to look at what exists and build something else.
• Better alone is cheap — you can charge more for a marginal upgrade.
• The bar: different and better in a way that matters to the end user.
• Cash App's cut-through differentiator was instant — instant cashout, instant card issuance, instant stock/Bitcoin proceeds.
• "Instant" worked because asynchronous money movement had no technical reason to persist — just inertia.
• Must also be in a domain with real economics; art can clear the first three bars but won't scale.

Building a startup within a company

• Small, tightly knit, senior teams outperform larger ones — fewer communication paths, more trust.
• Cash App launched and scaled with ~11–12 people; headcount discipline was enforced by CFO Sarah Friar.
• The "startup within a startup" illusion breaks because internal teams don't face existential fear — easy to accumulate headcount that outpaces results.
• Protect the team from organisational friction; Brian at Square gave Cash App a genuine shot.
• The hero customer must be unambiguous — every trade-off went to the consumer, which made people angry internally.

Hiring founders and building the team

• Roughly half of Ayo's Carbon Health hires have been founders; the thesis has proved correct.
• Standard recruiting pipelines filter out non-traditional candidates before a hiring manager sees them — intentionally break that by listing founders welcome in job postings.
• Cost 1: founders immediately call out waste and dysfunction — you have to sit and hear everything that's wrong.
• Cost 2: high attrition at the two-to-two-and-a-half year mark as ambition pulls them toward their next thing.
• Net benefit: higher output and a levelling effect on the whole team.

The relationship-first hiring philosophy

• "You pick the people, but they pick when." Time between first meeting and hire can be 18 months.
• Operationalise by meeting people continuously and asking: how can I add value to your life now?
• If you know someone wants something, have the ability to provide it, and it costs you nothing — not doing it is irrational.
• Ayo made ~600 introductions in one year: jobs taken, rounds closed, advisors placed.
• Being a good matchmaker compounds over time; give away cards instead of holding them.

Going deep — don't stop until you reach the end

• The Cash Card required visiting card manufacturing factories, testing ~1,000 combinations of laser-engraving settings, plastic overlays, paper, and finishes before shipping v1.
• Experts give answers that reflect their prior experience, not necessarily your specific constraint — keep pushing.
• The execution owner must become the domain expert; it doesn't have to be you, but someone has to own the depth.
• "I can't do it because X said so" is not sufficient — require: here is the contract clause, here are the exact consequences.
• Ask what actually breaks if we don't do it this way. Many regulatory constraints from prior jobs are applied inappropriately, creating worse user experiences for no reason.
• In complex, constrained environments (regulated industries, physics-bound products), avoiding the details doesn't save you — it just makes success fortune rather than skill.

Crossing into healthcare from fintech

• The connecting thread: expertise in regulated businesses, not expertise in money specifically.
• Going deep on regulatory wallpaper (product, engineering, legal, compliance in the same room, text on the projector) is the transferable muscle.
• Healthcare network effects matter more than merit alone — many deals depend on who you know, not the quality of what you're building.
• Direct-to-consumer healthcare (Carbon's model) reduces reliance on payer/employer networks and plays to consumer product skills.
• For B2B health tech: a crisp, specific use case surfaces the right decision-maker faster than broad networking.

Carbon Health

• Fully vertically integrated: builds and runs clinics, employs providers, and owns all software end-to-end.
• ~130 clinics across ~17 US states; urgent care and primary care; virtual care available.
• Software covers the full stack — patient booking through insurance claim submission.
• Example innovation: continuous glucose monitor streams blood glucose directly into the EMR; providers can intervene in real time, no $200/month subscription required.