You Presented the Data. You Forgot the Recommendation.

The metrics review meeting is on the calendar. You spent three hours pulling data, triple-checking queries, and arranging it all in a beautiful dashboard.

You share your screen. "As you can see, sign-ups were up 7% month-over-month. Time-on-page for the new feature is 2 minutes and 15 seconds, which is solid. However, our conversion rate from free trial to paid is down 3%."

You pause, waiting for applause. Instead, you get silence.

Finally, your head of product speaks. "Okay. Thanks for pulling this. So... what do we do about it?"

You freeze. You thought your job was to present the numbers. The clean, objective, unassailable facts. You were the messenger. Now they expect you to have an opinion?

Yes. That is the entire job.

The Data DJ vs. The Product Leader

When you only present data, you're acting like a data DJ, just playing the numbers. You’re asking your stakeholders—who are juggling a dozen other priorities and have a fraction of your context—to do the hard work of synthesis. You're outsourcing the most critical part of product management: forming a judgment.

This leads to two bad outcomes:

1. Analysis Paralysis: The team stares at the numbers. Someone asks for a different slice of data. Someone else questions the cohort definition. The meeting ends with a vague "let's keep an eye on it." No decision is made. Nothing changes.
2. Gut-Feel Decisions: The data is treated as interesting trivia. The highest-paid person in the room ignores it and makes a call based on an anecdote from one customer. Your hard work was just background noise.

A product leader doesn't just show the numbers. They interpret the numbers and recommend a course of action. They connect the what to the so what and the now what.

A Framework for Data-Backed Recommendations

You don't need a data science degree to do this. You just need a simple, repeatable process. Stop starting with the dashboard. Start with the question.

1. The Question: What are we trying to decide?

Before you write a single line of SQL, state the business question you're trying to answer. Frame it as a decision to be made.

• Weak: "Let's look at the onboarding funnel metrics."
• Strong: "Should we invest in improving the onboarding flow next sprint, or is it good enough for now?"

This focuses your analysis. You're not just exploring; you're on a mission to answer a specific question.

2. The Data (What): What are the objective facts?

This is where your dashboard comes in. Present the key metrics clearly and concisely.

• "Our funnel shows a 50% drop-off between Step 2 (Create Account) and Step 3 (Invite Teammate)."
• "Users who do invite a teammate have a 2x higher retention rate after 30 days."

This is the objective foundation. No opinions yet. Just the facts.

3. The Story (So What): What does this data mean?

This is the bridge from data to insight. You need to connect the quantitative facts to the qualitative user experience. Why is this happening?

This is where you bring in other evidence:

• Session recordings: "I watched 10 recordings of users who dropped off. Seven of them hesitated on the 'Invite' screen, scrolled around, and then closed the tab."
• Support tickets: "We have 25 tickets this month with the phrase 'don't know who to invite' or 'skipping this for now.'"
• User interviews: "In three interviews, new users told us they don't understand the benefit of inviting teammates so early in the process."

Now you have a narrative. You can say: "The data shows a major drop-off. Our qualitative evidence suggests it's because the value of inviting a teammate isn't clear at that moment. Users feel forced into a social action before they've experienced the core value of the product themselves."

4. The Recommendation (Now What): What should we do?

This is the final, crucial step. Propose a specific, falsifiable action. Don't be vague.

• Weak: "We should improve the invite step."
• Strong: "I recommend we run an A/B test. The control is the current flow. The variant will make the 'Invite Teammate' step optional with a prominent 'Skip for now' button. My hypothesis is that this will increase funnel completion by 15% without hurting long-term retention, because users who want to invite still can."

This gives your team something concrete to debate and act on. You’ve defined the action, the hypothesis, and the metric for success.

Your New Presentation Template

Throw out your old metrics meeting agenda. Use this one instead.

• The Question: We need to decide if we should change the 'Invite Teammate' step in onboarding.
• The Data (What): We see a 50% drop-off at this step. However, invited users are 2x more likely to be retained.
• The Story (So What): Session recordings and user feedback show that new users feel confused and pressured at this step, as they haven't experienced the product's value yet.
• The Recommendation (Now What): I propose we A/B test making this step optional. We'll measure the impact on funnel completion and 30-day retention.
• The Ask: Do you agree with this plan? If not, what is your concern?

Your job is not to be an impartial observer of your product's metrics. Your job is to have a strong, data-informed opinion on how to make the product better. Presenting the data is just the entry ticket. Presenting a recommendation is how you lead.