I'm Patrick Morris. I ran data for ESPN+, leading the analytics organization as the service grew to 25 million subscribers. Before that I built the experimentation platform at The New York Times. I work with a small number of heads of data, analytics, and product on the decisions that run on their numbers.
I built aftergrid because agents got good at fetching the number before they got good at the work that follows. The answers reaching product and growth decisions were fast and confident. They reported a rate without sizing it against the goal. Some were simply wrong in ways an experienced analyst would catch in a minute, like recommending a rollback because of a traffic shift, or celebrating revenue growth that hid churn.
The skills are free and MIT. Consulting is how I help directly.
Background
- ESPN+ at Disney, 2020 to 2025. Director of Sports Streaming Analytics. Led a 20-person analytics organization as ESPN+ grew to 25 million subscribers. Built the unified metrics framework behind billion-dollar sports rights evaluations, and a self-serve reporting suite that cut ad-hoc requests by 60%.
- The New York Times, 2013 to 2020. Product lead for the experimentation platform, used by more than 100 teams. Testing grew from 10 to more than 100 experiments a month, and I introduced dbt for analytics code.
- Now. I build AI products on live data. I run Sporting Works, which makes Spot Sports and its AI agent Jerry: a daily briefing that publishes about 290 fact-checked stories each morning for under $3 in model cost. How Jerry was built.
Consulting
Engagements for VPs and heads of data, analytics, and product. Each one starts with the decisions your leadership keeps re-litigating.
- Analytics leadership advisory. A senior sounding board for the person who owns the numbers: metric strategy, operating model, team design, and hiring.
- Metrics and decision review. Choose the metrics that matter and define them so a second analyst gets the same number: numerator, denominator, eligibility, time rules, counter-metrics. Then review the analyses reaching leadership against the goals they serve.
- Experimentation programs. Design or reset a testing program people trust: platform choices, guardrail metrics, a review step, and a cadence that gets from a handful of tests to a hundred a month.
- Analytics agents your team can trust. Put the aftergrid practice into the agents your team already uses.
make-the-callstates results against your targets,analysis-reviewchecks numbers before they reach a decision, and a capture-and-retrieval loop means the fourth analyst does not start from the same trap.
Your data stays in your repository. I do not approve your metrics; the owner does. With your consent, each engagement contributes one sanitized observation to the public examples.
Email me the meeting you are trying to get right.
Sponsorship
Sponsors fund the public skills, worked examples, and open-data Findings. In return: a named line on this site and in the docs, and early word on what ships next.
Rules that do not bend: nothing commercial appears inside a Finding. No sponsor touches a metric approval or a capability status. Placements live in site and docs chrome only.