Design Leadership · Product Process
Tri-track
A product process that put design evidence in front of Chatham's executives, rebuilt so every squad runs it.
Key outcomes
Film
How dual track became tri-track
Two and a half minutes on how the process works: what feeds each track, and how work moves between them. It's silent, so the captions carry it.
The film draws the process as tracks running left to right: vision, shaping, and delivery, with clients at the far end. Labels in the scene name what feeds each track and who owns it; they are listed after the captions. One piece of work is followed across each handoff, through the betting table, into clients' hands, and back around as feedback. It opens on dual track with one squad in 2021 and ends with every squad on tri-track. Two Onyx screens appear as mockups with invented data.
Overview
The Work
Tri-track is how Chatham decides what to build. It runs three tracks: vision, shaping, and delivery.
It grew out of a dual-track process I built with our engineering lead. On dual track a small team grew into three squads, and our GM, COO, Chief Strategy Officer, and head of product joined the shaping room. When we rebuilt it in 2026, I owned the shaping track. Every squad runs it now.
Challenge
The Problem
Dual track worked. A small team grew into three squads that had honest, evidence-based conversations with our executives about real trade-offs, and about the right time to chase an opportunity. Leaders came to trust UX research to de-risk a direction, so they decided with the most complete information we could give them.
It never had buy-in across the org. My engineering counterpart and I evangelized it to the people open to trying it. When shaping moved to product, most PMs struggled with the ambiguity and fell back on project-management habits. Then came reorgs, layoffs, and a rigid SDLC with layers of delivery managers and project managers. A process that ran on two people's evangelism couldn't survive that.
Where it failed was buy-in across the org.
Approach
How I Worked
Start with one squad and a weekly rhythm, Dual track
Our engineering lead and I gave the team a weekly rhythm: a shape-up meeting, one place to track requirements for the next release, and a fast way to feed insights from current work into future work. It grew from one squad to two, then three. Two more squads adopted it on their own. As it grew, I handed shaping for upcoming releases to emerging leaders.
Put executives in the shaping room, Dual track
Our squads ran some of the company's highest-profile work: the Portfolio Dashboard, and stronger audit and compliance workflows that moved advisors out of Excel and Jupyter notebooks for complex client performance reporting. Our executives wanted in. Our GM, COO, Chief Strategy Officer, and head of product were in every shaping meeting for our work streams. We brought them real trade-offs and decided together when to pursue an opportunity. Shaping with them taught the team to present design work, collaborate on the fly with sharp people, and hit deadlines on ambitious problems.
Use research to de-risk the bet, Dual track
We built trust in UX research as the way to test a product and design direction before committing to it. Leaders stopped deciding on gut alone.
Find the limit of evangelism, Dual track
We presented the process to other leads in our business area and helped them apply it. Our COO and Chief Strategy Officer loved it and pushed their own projects to follow it. But my engineering partner and I were supporting two or three squads at a time, and we didn't have the capacity to bring other squads up to speed. Some teams leaned in to running ahead and de-risking. Others weren't ready for change, or couldn't take it on without full support. It never became the formal approach. Through the SDLC change, the reorgs, and the layoffs, we kept what worked for our teams: regular research shares, planning well ahead, and a betting table for trade-off conversations with commercial and product partners.
Start informally, in front of our clients, Tri-track
Another reorg, and the company's AI transformation, gave us runway to rethink the process. Tri-track started informally with the demo for Chatham's Global Client Summit in NYC in June. Design leaders, product leads, the COO, GMs, and key business leaders combined live Onyx features with working prototypes of what comes next: agentic financial reporting, proactive risk alerts, end-to-end capital planning and trade execution, and lender intelligence. Together they showed Onyx as an operating system for the capital markets. Hundreds of clients saw it, and their feedback went straight back into shaping.
Formalize it through retros, Tri-track
After the Summit, we ran multiple retros and built tri-track out of them, with more structure. This time we had buy-in from leadership, and product management had matured as a practice. We set the level of detail to fit the work and cut the bad friction out of planning: documentation, recaps, and communication up and down the org chart. Small quads now shape the work, and most people own two or more tracks, so no single role carries the ambiguity alone.
Solution
What We Built
Three tracks
Vision sets where the platform is going. Shaping moves a team from an ambiguous ask to a committed bet. Delivery builds it.
Quads
A PM, a UX lead, an engineer, and an architect shape the work together. Smaller teams mean more people have a hand in the process.
Shared track ownership
Engineers own shaping and delivery. UX leads own all three. PMs own part of every track.
A documented shaping track
I documented the shaping track, presented it to the product and tech org, and built AI skills to support it. An AI working group keeps AI workflows consistent across all three tracks.
Results
What Changed
On dual track our squads became an example for the company. We made commitments, aligned with clients and leadership, presented what we built and planned at company-wide meetings, and hit our deadlines. It peaked at five squads: three we ran, and two that adopted it.
Every squad runs tri-track now, across Investment Management, Digital Execution, Platform Capabilities, and smaller parts of the tech and product org. Hundreds of clients saw the Onyx vision at the Global Client Summit, and their feedback fed directly into shaping. The same teams onboarded our first clients and counterparties to Onyx's derivatives trading products, and are preparing a broad launch for clients to manage their debt and assets in the platform.
We stopped being territorial. Designers ship code. PMs show ideas as coded prototypes built with our design system. Engineers make design calls without a Figma file, because the design guidelines live in the frontend monorepo. Delivery teams run with more autonomy, which frees shaping to move faster and spend more time on the vision. I brought data from our UX Health Log to the vision track.
Reflection
What I Learned
Dual track was a highlight of my career, and a stretch where I was improving at a rapid pace.
Two champions didn't make it formal
Our COO and Chief Strategy Officer championed dual track. When we pitched our CTO and CPO, they thought it was fine. Both were new to Chatham and had their own vision for how teams should work, and they brought good ideas. Dual track never became the formal process. Looking back, I'd pitch them hard, with data beyond our hit deadlines and anecdotes.
Spreading a process takes protected time
We ran two or three squads and tried to coach others on the side. Evangelism reached the willing and stopped there. Tri-track started with leadership behind it.
Don't hand a process to people still learning the practice
Shaping stalled when it moved to PMs alone. Tri-track spreads it across the quad.