During our recent Metabase × Productside webinar when I teamed up with Tanya Aulachynskaya (Product Manager, Cloud, at Metabase) half the audience admitted they juggle “multiple options” for analytics: Amplitude for product, Excel for finance, Salesforce for sales.

The predictable result? “Reports in different numbers” even when everyone is “trying to report on the same metrics”. Trust erodes, sprint reviews derail, and product velocity flat-lines.

The Before-and-After Story: Heavy Stack ➜ Self-Service Product Analytics

Before: At Tanya’s previous unicorn, data lived in a “very massive and cumbersome” stack: ETL pipelines, Snowflake, Tableau and Amplitude. Annual tool cost: ~$150 K. 
Head-count just to wrangle reports: ~$800 K. And every new dashboard sat “a month” away in an analyst backlog.

After: At Metabase, the same insights come from a lean team (“one data engineer and one analyst”) because PMs and success managers build the dashboards themselves “in 5, 10, 15 minutes”. That is the promise of self-service product analytics: insight at the speed of curiosity, minus the seven-figure overhead.

A Three-Step Playbook to Build Better Dashboards in 90 Days

Follow the timeline and you don’t just build better dashboards. You build an always-on decision engine.

Real-World Patterns You Can Reuse Today

The webinar demo shipped an end-to-end SaaS dashboard in real time. Steal these patterns straight from Tanya’s screen:

1. Overview tiles: sign-ups, conversion, MRR, customer count, so execs grasp health at a glance.
2. Acquisition funnels: chart every drop-off step (e-mail verify, pricing wall) and target the gaping holes.
3. Activation curves: track the % of new users who perform the “aha” action (e.g., create first chart) within 48 h and correlate to retention.
4. Feature leaderboard: rank events by usage and by associated revenue to spot zombie features and premium darlings.
5. Timeline annotations: log outages and pricing experiments so you can answer, “Why did activation dip in April?” with evidence, not guesses.

Together these components transform a static board into living self-service product analytics.

Quick Dos & Don’ts (All From the Webinar Floor)

Dashboards Are a Product (Treat Them Like One)

Tanya’s build-live demo ended with a killer reminder: you can save the board and share by link, then iterate. That mirrors any good MVP cycle:

1. Define the user need. PM? CFO? Success? Different questions, different boards.
2. Ship the smallest lovable version. One activation chart > 40 vanity graphs.
3. Instrument your own dashboard. Track views, drill-downs, and filter clicks; sunset zombie widgets.

Do that and you build better dashboards and run them like a real product.

Put the Insights to Work

• Watch the full session to see Tanya build, slice, and drill into cohorts live.
• Download the Metabase chart cheat-sheet (link in the webinar recording) and pin it next to your monitor.
• Join Optimal Product Management to master Outcome Canvases, prioritisation, and how to use self-service product analytics in six momentum-packed days.

Ready to start? Consolidate the data, give PMs the keys, and watch decisions accelerate. Because the best dashboard isn’t the prettiest… it’s the one your team trusts enough to bet the roadmap on. Connect with us on LinkedIn to share your insights with us!

Let’s make product data work like they should for product managers.