Opinions are mine alone, not my employer’s. Typed on my personal Mac.

“It’s not about the players, it’s about getting on base.” That single line from Moneyball explains why baseball changed forever — and why product managers are still stuck chasing batting averages.

In the movie, Brad Pitt had to explain to a room of baseball executives that they were chasing the wrong thing.

They were still obsessing over batting averages, stolen bases, even the looks of the players’ girlfriends. Instead, he argued, the dull, unfashionable metric of on-base percentage was the real driver of wins.

Product management has its own version of batting averages and pretty girlfriends: GMV, conversion, revenue. These are handsome metrics. They sparkle in boardrooms and look great in quarterly business reviews. But they are lagging indicators. By the time they dip, the season is already lost.

And right now, this matters more than usual.

It’s roadmap season.

Every team is scrambling to defend their existence with metrics, trying to look indispensable in a spreadsheet. The problem is most of these defenses are built on the same shiny but hollow numbers — GMV, conversion, retention. They sound impressive, they travel well in executive decks, but they tell you nothing about whether the underlying engine is healthy.

In roadmap planning, this is the equivalent of baseball scouts arguing over jawlines while the real game is happening somewhere else.

The Case for Leading Metrics

The game is to find the boring, predictive signals that silently determine whether those flashy lagging metrics ever show up. That’s the real on-base percentage of software.

Netflix knows monthly subscribers don’t predict retention. They measure completion rate of the first three episodes.

Airbnb realized GMV doesn’t predict trust. So they tracked percentage of listings with verified IDs and 5+ photos.

Amazon doesn’t just worship GMV in WBRs. They drill into detail page views per session and percentage of items delivered next day.

Duolingo doesn’t care if you downloaded the app. They care if you start a streak — that tiny act predicts lifetime value.

None of these are glamorous metrics. You don’t win applause in a town hall for reducing checkout latency by 200 milliseconds.

But these are the boring truths that win the game. What about intuition?

Intuition vs. Data

The software industry is full of inventions that began as intuition, not dashboards:

Gmail came from Paul Buchheit’s hunch that email should be searchable and limitless.

The Facebook Like button wasn’t demanded by data. It was a gut belief that a single-click gesture could change engagement.

Slack was Stewart Butterfield’s conviction that work chat should feel like game chat, not a boring inbox.

Git was Linus Torvalds’s refusal to tolerate clunky tools, not a market analysis.

But intuition alone is dangerous. A hunch without data is called “vision” when it comes from an executive and “insubordination” when it comes from a PM.

Some inventions were born purely from data:

Netflix Recommendations came from the Netflix Prize — math alone spotted patterns no human guessed.

Google PageRank proved backlinks were the real predictor of relevance.

Microsoft’s 41 shades of blue were not aesthetic genius, just A/B tests that made millions.

TikTok’s For You Page is entirely data-trained on watch time. No human curator knew what teenagers wanted.

If intuition-born products were gut punches, these were cold, machine-calculated body blows.

The Real Skill of a PM

Too many PMs, including the author of this post, either: Worship dashboards blindly — data first, story later. Or worship their gut — story first, proof never. Both are useless.

The truth is in the balance: intuition points you to the lever; data tells you if it’s moving.

Personally, I’m in a phase where I trust my intuition enough to recommend actions even without estimates — and I often believe the results would bear me out. Unfortunately, I’m not an executive to make such proclamations stick.

And maybe that’s how it should be. Because if every PM ran on unchecked instinct, companies would drown in false positives.

Executives, of course, enjoy this privilege. They can declare bold visions without data, and the math often checks out because they’ve marinated in their space for decades. But for the rest of us, telling stories without data is not “visionary.” It just makes us look like a lousy product manager.

Daily Exercise to Sharpen Data-savviness

Talk to a Dataset Like It Owes You Money

Open any dataset, a Kaggle CSV or last week’s dashboard. Ask AI what’s wrong with it. Not what’s in it, but what’s missing, what’s suspicious, and what a cynic would say about it.

If AI ever returns a clean bill of health, you’ve either picked the wrong dataset or the wrong AI

Treat Every Metric Like It’s Lying

Pick a metric — churn, CTR, GMV, anything your colleagues weaponise in meetings — and ask AI how it can be gamed.

If you don’t know how your own metric can deceive you, congratulations, the metric already has.

Practice Causal Storytelling (A Skill Most People Think They Have)
Give AI two numbers that move together and ask for five possible causal stories — and five ridiculous ones.

Half the fun of data is discovering how often the ridiculous ones are true.

Summaries That Hurt Feelings
Paste a long WBR, seller deep-dive, or a heroic Jira comment written at 2 a.m.

Ask AI to summarise it in ten brutal lines — each line a decision, not a paragraph of throat-clearing.

If someone’s feelings aren’t hurt by the summary, you’re being too polite.

The Daily “What If Everything Breaks?” Drill
Pick a scenario: delivery delays spike, search collapses, carts get abandoned like New Year’s resolutions.

Ask AI for the ripple effects — the real ones, not the ones in the deck.

Data intuition often begins with imagining disaster and then calmly predicting the sequence of misery.

Daily Exercises to Sharpen Intuition

Silent Forecasting: Predict yesterday’s numbers before opening a dashboard. Track your hit rate.

Metric Substitution: Take one lagging metric (GMV, conversion) and invent three leading ones that might predict it.

Micro-Observation: Read user complaints or reviews for 15 minutes a day. Spot patterns before the analysts do.

Reverse Dashboard: Pick a single metric and ask: If this were the only metric I had, what story would I tell?

One Hypothesis a Day: Write one prediction daily, serious or absurd. Validate later with data.

So maybe the lesson is simple: don’t chase the handsome metrics. Don’t become the PM of pretty dashboards. But don’t become a prophet either.

Play Moneyball. Find your on-base percentage. Use your intuition to spot it — and your data to prove it.