15 lessons for internal data product managers I learned the hard way
After years of working as an internal data PM and teaching others, I've distilled some crucial lessons that can help anyone managing internal data products. Here are the key insights that could save you years of learning things the hard way.
FWIW I got a great e-mail course (free!) with 20 great lessons and lots of to dos for you ready at “Product Discovery for Analytics Teams”! Go check it out.
But now let’s dive into the quick short lessons I got ready for you today…
1. Users ≠ Customers
Your users aren't necessarily your customers. While a business analyst might use your BI tool daily, the real customer is the decision-maker who acts on those insights. Example: In sales departments, individual salespeople might use your reports, but the Head of Sales often determines what data actually matters.
Why? Because at the end of the day, management discusses your budget, and in those meetings, the Head of Sales needs to know you’re delivering, not the analyst.
2. Follow the Decision Chain
Always trace the path of how your data leads to decisions. If someone can't make decisions based on your data or take action with it, they're probably not your primary customer. They're likely serving someone who is.
So keep on asking “how does this data change your decision? What different actions would you take then?”
3. Look for Hidden Customers
Some of your most important potential customers might not be using your products at all. Like a marketing department that isn't using your BI system not because they don't need it, but because no one ever showed them how to connect their data.
I still remember that one marketing department that hid from me, because they were all in on Google Analytics and didn’t love Tableau at all (the system we were pushing). Well guess what, a year later we had all the Google Analytics data in our system, and the marketing department suddenly was one of the biggest customers.
4. Challenge Requested Solutions
When someone asks for an OLAP cube, they might really just need a CSV export. Don't get trapped into thinking that complex solutions are always better. Sometimes the simplest solution is the most valuable.
5. Break Out of Expected Patterns
Internal teams often get cast into a shell of delivering results in predefined formats (reports, dashboards, SQL queries). Don't let these expectations limit your solutions. Your team probably has capabilities far beyond what people typically request.
6. Split Discovery Conversations
When talking to stakeholders, split the conversation into two parts:
First half: Focus purely on problems and current workflows
Second half: Discuss potential solutions
Don’t mix the two, of you’ll fall back into the expected patterns.
7. Ask About Workarounds
One of the most revealing questions you can ask is: "What would you do if we couldn't build this?" People often have existing workarounds that can teach you about their real needs.
This question isn’t meant to ignore feature requests, it is meant to understand value!
8. Follow the Export Trail
If people consistently export your data to use it elsewhere, that's a red flag that you're missing something important about their workflow. Ask why they need to export and what they do with the data afterward.
Then be prepared to help them export faster! Join better, or replace their existing solution with something even more amazing.
9. Look for Information Asymmetries
Your most valuable opportunities often lie where there's an information gap between different parts of the organization. These are places where data can make the biggest impact.
10. Distinguish Between Three Customer Groups
Always categorize your customers into:
Non-decision makers (users)
Decision-makers (true customers)
Not-yet customers (potential opportunities)
11. Start with Problems, Not Solutions
Even if you only get one meeting with a senior stakeholder, spend the first part understanding their problems before jumping to solutions. You might find that their requested solution isn't what they really need.
12. Map the Impact Chain
Always understand how your data flows to actual business impact: Data → Analysis → Decisions → Actions → Results If you can't trace this chain, you might be building something that won't get used.
13. Question the Defaults
Just because "we've always delivered reports this way" doesn't mean it's the best approach. Internal teams often inherit processes that could be radically improved.
14. Find the Real Bottlenecks
Sometimes, what people ask for (like timestamp formatting) isn't their real problem. Dig deeper to find what's actually slowing them down or preventing better decisions.
15. Build Long-term Relationships
Unlike external products, internal data products benefit from ongoing relationships with customers. Use these relationships to really understand their needs, but don't let them limit your solutions.
Final Thoughts
Being an internal data PM is uniquely challenging because you're balancing deep technical knowledge with organizational dynamics. The key is to never stop questioning assumptions and always look beyond the immediate users to find your true customers.
Remember: Your job isn't just to deliver what people ask for, but to help them make better decisions with data. Sometimes that means challenging their requests and finding completely different solutions than what they initially wanted.