Why Product Managers should drive analytics

In this post I'll be sharing how to best structure analytics capabilities within a product company and how to collaborate better with analysts

Only 10% of product teams are able to validate their most important product decisions with data, according to the latest State of Product Analytics report from Mixpanel and Product School. Over the last few years I have definitely noticed a big increase in awareness of the importance of data, but despite it being talked about more, analytics systems are not being implemented effectively yet as this study shows.

I believe there are several reasons for the lack data driven decision making in product teams. In one of my recent articles The gap between product and analytics I shared my observations about the lack of resources, conference talks, or training opportunities for Product Managers to learn analytics skills. The lack of knowledge ultimately means analytics is often put into the “too hard basket”, and I don’t fully blame them knowing how long it took me to get to an acceptable level.

The other big reason, and the one I want to explore in a little bit more detail today, is how we structure analytics ownership within an organisation.

The importance of data for Product Managers

As Product Managers we are responsible for making sure we build products that bring value to customers and are viable for the business. To evaluate whether the product is on the right track to have an impact on the customer and the business, the Product Manager needs to have continuous data inputs.

If the Product Manager is not equipped with sufficient data insights, product prioritisation decisions will typically be based on gut feeling or opinions from the loudest stakeholder in the room. Critical insights may also be missed that could uncover new opportunities for the product like a new customer segment with high engagement or unexpected usage patterns of your product.

Something I hear very often when interviewing Product Managers is that data analysis is being “outsourced”, usually to a centralised data and insights team within the company that splits their time across many product teams. Unfortunately I have never met a Product Manager who truly developed a deep understanding of their data when it was being presented to them in a summarised Powerpoint once a month.

If you want analytics and data to become part of your day to day decision making, you need to be the one actively driving the creation and aggregation of data insights. That means collaborating a lot more closely with data analysts, defining what events are being tracked, driving implementation of events and dashboards, as well as facilitating regular sessions to share and discuss data insights with the product team.


Another responsibility on the Product Manager plate?

One comment that I hear very often is that Product Managers don’t have enough time to drive analytics implementation and monitoring the data. If your company is big enough to have data analysts or insights teams available, great - definitely use their help as they are absolute data experts and can be an incredible help for diving deep into your data.

I have usually found it pretty doable to take care of analytics myself if I prioritize it enough and put aside focus time. Sometimes I find Product Managers spend time on things that are a lot less effective and helpful than this, like wireframing or QA testing their product themselves. I would strongly recommend to fill those gaps first with the right people, or fix any other inefficient processes, so the Product Manager can focus more on key activities like defining and measuring product usage.

Insights teams are still an amazing resource to have though - if I can complement my analytics work with people who can go way deeper than I can, run custom queries on raw data or setup more reporting automation, fantastic. But I do expect Product Managers to be able to define what we need to track, create basic reports and facilitate the sharing of insights to truly build an understanding of their product usage data.

How we can work better with insights teams

We’ve long been advocating for the close collaboration between product, design and tech. We know cross functional product teams work better than if we would keep them in separate departments, so why are we still keeping insights teams in silos?

It may seem more effective to have shared analyst resources across the company, however this setup comes with a few downsides that can have a big impact on the ease of access and quality of insights for product teams.

While Product Managers often lack a deep understanding of good data practises, insights teams typically lack context from us to produce truly useful insights. Just imagine what you could achieve through more frequent collaboration, and sharing the current product challenges, upcoming goals and discovery activities with an analyst who is actually embedded in your product team!

Especially in larger companies insights teams also often have a large backlog of requests to work through, and it is not uncommon to wait weeks to get a report with the requested insights. This in turn has a big impact on how quickly product teams can work through discovery activities that rely on this data.

The worst thing we can do is to ask insights teams to just look into the data and “try to find something interesting”. There is a case where you want to explore your data to look for the unexpected, but that’s the next, more advanced stage once you have your basics in place, so let’s not worry about that yet.

If you’re lucky to have a great team of data analysts in your organisation, here’s how you can setup a more effective structure:


Step 1

Try to make analysts part of your product team. Ideally you have one dedicated person, even if part time, who can gain real product knowledge over time and build strong relationships with the team.

Step 2

If you have managed to secure a dedicated analyst for your product team, include them in your daily standups, your discovery work, product strategy, and planning sessions. The more context you provide, the more the analyst will be able to see how they can contribute to your product with their insights.

For example, the analyst will always be clear on what’s coming up next and can make sure the reporting is setup in a way the team can measure the impact of the upcoming features from day one. In general, you will find the reports and data insights will become more and more valuable as the analyst will get the benefit of the full feedback loop working closely with the team over time.

Step 3

Sharing, sharing, sharing. Now that you have established a data capability within your product team, the key is to not shield the information from the rest of the team either. While this has nothing to do with your team structure, I thought it was crucial to mention in this context that the Product Manager and the analyst need to make the insights available to the whole team.

Give everyone access to your analytics dashboards, put them up on big screens for everyone to see, allow everyone to ask questions about the product usage. My favorite meeting of the week is usually a weekly data insights session with the team. We run it on a Friday, where we review how our key product metrics have developed throughout the week, and plan meaningful actions based on those insights.

Final 2 Cents

If we want our product teams to truly operate in a data driven way, we need to integrate analysts closely into our teams. If there is no insights team available, the absolute worst thing a Product Manager can do is to use this as an excuse for why they don’t have any analytics setup.

The best Product Managers I know have found ways to make time and figure it out themselves as they truly understand the importance of data, even if it means some trial and error. And whether you have analysts to help you or not - I strongly believe the Product Manager needs to be the overall driver to gather and share product usage insights.

On a last note, doing an analytics plan, tools and reporting setup at least once or twice yourself is sometimes the best way to really understand the true nuances of product analytics.

Edits April ‘21: added embedded analysts graphic and explanation


Want to read more of this?

If you want to read more on the topic of analytics, here are some more articles you may find useful:

The big gap between product and analytics

Why we need to stop tracking vanity metrics

We need to stop tracking too much