Find the work that helps with your next decision
This library is broad. These four paths are deliberately narrow: five articles, in order, for one kind of reader and one practical outcome.
You do not need to finish a path. Start with the question closest to the decision in front of you.
Lead an experimentation program
See what changes when experimentation becomes an operating system rather than a sequence of isolated tests.
What you should leave with: Build executive trust, make better tradeoffs, and scale without losing rigor.
- Inside 200+ A/B Tests Start with the patterns that only become visible across a large portfolio. Read article
- How to Build a High-Impact Experimentation Program in a Year Move from individual wins to the systems, relationships, and cadence behind a program. Read article
- Decisions with Incomplete Data Learn how senior operators decide when the evidence will never be perfect. Read article
- The Politics of A/B Testing Recognize how incentives and interpretation can distort otherwise sound results. Read article
- Speak the CFO's Language Translate experimentation into risk, capital allocation, and business outcomes. Read article
Run better experiments
Strengthen the judgment around test setup, evidence quality, and shipping decisions—not just the mechanics.
What you should leave with: Catch weak tests earlier and make decisions that survive scrutiny.
- Getting Started with Optimizely Web Experimentation Establish a practical setup and QA foundation before interpreting results. Read article
- Why Most A/B Tests Fail Diagnose the program conditions that produce weak tests again and again. Read article
- Speed vs. Rigor Choose where to move quickly and where methodological shortcuts become expensive. Read article
- Post-Hoc Metric Shopping See how flexible interpretation turns noise into a persuasive but unreliable story. Read article
- Flat Primary Metric, Positive Downstream Work through a difficult shipping decision when the metrics disagree. Read article
Use behavioral science responsibly
Apply behavioral principles as testable hypotheses while protecting trust and staying honest about the evidence.
What you should leave with: Design interventions that are useful, testable, and defensible—not manipulative decoration.
- A Practitioner’s Guide to Behavioral Economics in Marketing Begin with a working model for turning principles into measurable decisions. Read article
- Behavioral Economics After the Replication Crisis Separate durable evidence from findings that deserve more caution. Read article
- The Endowment Effect Goes Bad Examine where a useful principle crosses into a trust-damaging dark pattern. Read article
- Price Anchoring Without Tricking Buyers Use comparison and context without hiding the real economics from customers. Read article
- Why More Social Proof Sometimes Backfires Learn why a familiar persuasion tactic can reduce confidence instead of increasing it. Read article
Study the founder field notes
Follow the constraints, experiments, and distribution choices behind building products with a small team.
What you should leave with: Find practical operating ideas without mistaking one founder’s path for a universal formula.
- How Jobsolv Grew to 30K Users with $0 Ad Spend Start with the distribution system that turned product value into sustained organic growth. Read article
- From Services to SaaS See how service work exposed the recurring problem worth turning into software. Read article
- Managing a 27-Person Team as a Solo Founder Understand the operating model, delegation choices, and limits behind the headline. Read article
- The Solo Founder’s Playbook for Growth, Experimentation, and Shipping Connect product decisions, distribution, and a sustainable shipping cadence. Read article
- Build What You Use Use proximity to the problem as a research advantage without confusing intuition for proof. Read article
Turn the reading into a decision
If the question involves an active experimentation program, I can help diagnose the constraint and identify the highest-leverage next move.