All work

Capital Preferences · 2014–2024 · 2 min read

Building for Enterprise

Took a product from nothing to live deployments at J.P. Morgan and Vanguard.

The challenge

Every financial adviser has to profile a client's attitude to risk before giving advice. It's a mandatory part of know-your-client, and most firms did it with psychology-based questionnaires and conversations. Those left room for subjectivity that no enterprise could accept at scale, and they weren't reliable enough to run a digital-only service.

I joined Capital Preferences as employee number one, and there was no product yet. BNY Mellon had asked the founder's consulting firm, Capital Position, to find which clients to target with personalised products, and the market research turned up a gap. The answer came from Shachar Kariv's experimental economics research into how people decide when they play risk games.

I started as an analyst and grew into senior product manager. Later, the company needed me closer to its clients, so I became head of customer success and managed two customer success managers. It showed me that product is where I do my best work.

The approach

  1. Winning Vanguard on accuracy

    Vanguard wanted a digital-only advice service, but its advisers overrode so many profiles that a human process wouldn't carry over. I led the bid: I designed and ran a test of our profiles against the ones its advisers had settled on for existing clients, then shaped the product around the results, and we won.

  2. Built to be embedded

    Vanguard showed that each client needed the experience to fit its own journey, so I chose one configuration layer over custom builds and defined what clients could change and what had to stay fixed for the research to hold. It shipped as an HTML component, clients chose what data reached us, and I onboarded firms from Vanguard to ones as small as Lemmi, each going live fast.

  3. A bet on ESG

    As sustainable investing boomed, J.P. Morgan Asset Management wanted its advisers' tools to reflect what clients cared about. I helped shape the ESG activity and worked with our research team to define an optimisation engine that turned risk and ESG preferences into n=1 portfolios of real funds, built around J.P. Morgan's needs. It launched well before US states began barring ESG investing, and while it never matched risk's scale, it proved the research could carry into new domains and opened the way to the preference activities that followed.

  4. Into Merrill's adviser workflow

    At Merrill we again had to beat alternative tools, this time in a head-to-head pilot with its advisers. I ran the trial, trained the advisers and led the product changes their feedback called for, and the pilot won a broader rollout.

The impact

  • Grew from $0 to $10M ARR.
  • 100% uptime over 24 straight months.
  • 300K+ investor profiles created and plans recommended.
  • Advisers overrode far fewer risk profiles once they used our profiles.

Live at

NextPutting It on the Record