Product metrics: what every founder needs to track
The essential startup metrics: activation, retention, NRR, and CAC payback. What to watch every week, and what only adds noise to the dashboard.
Topic guide
Product strategy from idea to launch: customer discovery and validation, MVPs, stack choices, and when to build or buy.
3 topics · 17 articles
Talk to customers before you build.
The essential startup metrics: activation, retention, NRR, and CAC payback. What to watch every week, and what only adds noise to the dashboard.
A practical roadmap for adding generative AI to your product: embeddings, RAG, fine-tuning, and agents, with the cost and privacy decisions made upfront.
GA4 answers where users come from. PostHog keeps raw events and shows the funnel. Compare privacy, 3-year cost, and the team that reads the report each week.
RICE orders the backlog in an afternoon. Inflated confidence, effort without review, and reach without a number skew the score. Three rules fix it.
Embedded text, Brazilian date formats, and values without symbols block your second language. Split strings, format by locale, use RFC 5646 tags.
Event contract at the source, a queue between product and warehouse, and tested transforms. Without them, every dashboard becomes a number debate.
Ship the smallest product that tests the hypothesis.
Data silos, manual copy-paste between tools, and decisions without context. The invisible bottleneck that kills startups before the product does.
A SaaS MVP is not a landing page with a payment button. A 30-day framework for billing, multi-tenancy, and onboarding that validates without a big budget.
A technical checklist for founders about to hire engineers: what to audit in the code, the infrastructure, and the processes before you lose months.
Firebase ships auth and push over the weekend. Firestore charges per read and exports no data in real time. Validate in 6 weeks and plan the exit before you grow.
Stable assignment, a metric defined upfront, and duration set by statistical power. Without those three pieces, an A/B test produces vanity and no product decision.
Open source charges no license and charges maintenance. SaaS starts cheap and gets costly as the team grows. Add operations and upgrades over 3 years before picking.
Integrations, APIs, and third-party SaaS.
The wrong pricing model kills SaaS early. Compare value-based, usage-based, and seat-based, then pick the model that captures what your product earns.
Every build vs buy decision carries a hidden cost. Four questions that give you criteria and keep pride, lock-in, and future debt out of the call.
An API without versioning is a time bomb for you and your clients. A practical guide to versioning, migration, and deprecation without breaking integrations.
The Stripe checkout takes an afternoon. Proration, partial refunds, and idempotent webhooks take weeks and decide whether billing becomes support or revenue.
Strategies for modernizing legacy systems while the business keeps running: anti-corruption layer, strangler fig, and events, in the right order.
Talk to 10 to 15 potential customers. That is 2 to 3 weeks of conversations, not 3 months of analysis.
The smallest product that tests the main hypothesis with real customers. If it will take several months to go live, it probably has too much scope.
Buy what does not differentiate your product, such as authentication, payments, and email. Build what makes customers choose you.