Business operations and strategy lead. Ten years getting Product, Finance, GTM and Engineering onto one set of numbers, then building the cadence, reporting and software that keep them there.
Built with the teams rather than handed to them. Annual planning and budget for a $B business; the OKR architecture a COO adopted into the next year's operating plan.
Fifty-six initiatives across four product areas, through weekly forums and VP QBRs. On the merger, none of the 16 VPs reported to me.
The SQL and Looker layer that joined pipeline to product adoption, and the eight figures of incremental revenue that came off it. A Tableau compliance layer across 60+ projects.
When the system a plan needs is not there, I build it. All four are below.
The same sequence on a merger, a cost program and an annual planning cycle.
Name the decision actually being made, and who owns it.
Turn the ambiguity into constraints, resources, dependencies and dates.
Find the binding constraint, and what it costs while it holds.
Compare the options on impact, on cost, and on when the effect lands.
Hand it over as ownership, cadence, and a number that keeps arriving.
The simulator below is this method running end to end, on a fictional company's numbers. Open it →
When the system a plan needs does not exist, I do not wait for a software team. Three I built, shipped and open sourced.
Systems thinking · Interactive model
Can this organization actually execute the plan? A month-by-month model of a fictional company: which team hits its capacity limit first, which initiatives are at risk, and what each fix changes, on what date, for how much. Runs on your own numbers too. I built it because it is how I actually work.
Open the simulator → Automation and data integrity · Open sourceAny system that reads your mail and decides what each message means will quietly start lying to you. This one doubts itself: 886 hand-labeled messages every change has to clear, and a command that re-reads a year of records and reports where they and the rules disagree.
Read the write-up → Operational tooling · Open sourceGoogle Sheets keeps comments where you cannot sort, count, or export them. I built a browser tool that fixes it, then shipped the same thing as a Google Workspace add-on: one button, and every thread lands in a tab with its cell, its replies, and whether it was resolved.
Read the write-up →
Here to solveStand up a new line of business, and make the cost of running the work visible early enough in the year to act on it.
Scope
Impact
Here to solveGet a fragmented regional partnerships organization planning, selling and reporting as one business.
Scope
Impact
Here to solveBuild the operating machinery five years of integration would run on, before the deal had closed.
Scope
Impact
Lean Six Sigma Green Belt · Google Project Management Certificate
What was actually in the way, how I worked the constraints, and where I got it wrong.
Unit cost arrived from finance late in the year, when the only lever left was people. It became a number three functions signed and saw every month.
What I got wrong: built the complete version first and the one-screen version second, which is backwards.
Read it → Program governanceSeveral hundred people kept building through months of regulatory hold, on a deal that might not have survived review at all.
What I got wrong: did not standardize intake before a dozen vendor teams each built their own tracker.
Read it → Go to marketSellers already had the list. They could not tell a publisher what it was worth, so it kept losing to pitches that arrived with a number.
What I got wrong: scoped the queries to my own region, then could not validate the two that adopted it.
Read it →Operating problems that usually get settled by instinct. Each of these has something you can play with.
Operating model designWhen a support function should sit central and when it should split into dedicated pods. Your shared person would have to be 73% slower on every single request before pods become cheaper, and almost nobody guesses anywhere near that. The number moves with client size.
Read it →I built intake the way I needed it rather than the way anyone would fill it in, and then blamed the stakeholders when nobody used it. Forty-two fields, three of them ever read, at a company whose CRM already held most of the answers.
Read it →
If you need someone to take an ambiguous cross-functional problem, structure it, put numbers on it and leave behind the cadence that keeps it moving, that is the work I want. Particularly the kind that arrives without a clear owner.