How to rationalize data in your product
It’s not simple to extract data from your product, but you should take the first steps
Hello, I’m Tiago Ferreira, Sr. Product Manager in Brazil with +6 years of experience crafting products. With The Next Movement, I want to share part of my product management experience with the whole world, but also talk about career more broadly, technology, good books, and - why not? - philosophy, music, culture, gossip, just like an open diary. If you enjoy reading my article, subscribe and share it with your friends 🤓
Every product needs effective data collection, but the reality often falls short. Product Managers frequently find themselves in situations where there's no data collection in place, hindering any important decision-making.
Before venting frustration to your boss, consider taking the initiative to rectify the situation.
A product needs data to understand its value generation. This process can start gradually: from scribbling on paper, transitioning to a digital board (Miro, Mural, etc.), engaging in conversations with data analysts and scientists, and eventually arriving at a tagging plan.
But before all that, it's crucial to have clarity about what is expected from your product.
Below is a basic guide to constructing your product's data rationalization. This will require a considerable amount of time from your schedule to structure your thoughts, engage in discussions with experts and managers, and even adjust the prioritization of your backlog.
Understand What Your Product Does
It might seem straightforward, but Product Managers don't always precisely articulate what their product does - sometimes due to being immersed in numerous meetings, juggling various tasks when dealing with stakeholders, or simply because their backlog is overflowing and urgently needs attention.
Taking the time to contemplate what your product does is crucial. A tip: block off your schedule for these moments. And don't think of it as a solitary task: discuss it with your direct leader, understand how stakeholders perceive your product, and start shaping that vision.
Document this, whether visually on a Mural/Miro board, in Confluence, or even in a PowerPoint presentation. Align this with all stakeholders and management to ensure your product vision is solid.
One way to achieve this is by cascading down from the company's objectives. If you work for a retail company, for instance, focused on GMV, and you're responsible for the cart stage, it's worth rationalizing: how can my product contribute to increasing GMV?
This will provide a good direction on how to start structuring the metrics.
Apply a Metrics Framework
Frameworks exist for a reason: to help us understand our product. For this, there are various frameworks for different situations.
Start with the simple ones. First and foremost: what is the main business objective?
Let's say you work with an API that allows banks to transact cryptocurrencies.
The company's goal is to increase the number of banks and fintechs plugged into this API.
As the person responsible for API activation, you need to ensure that customers activate the product. To do this, you need to keep an eye on:
Number of times the API is called;
Failed calls;
Transactions made through the API.
By looking at these metrics, you'll understand where there may be friction and can delve deeper.
A very useful framework is the KPI Trees, which, in the words of product coach Petra Wille, help to:
Generate alignment between teams, establishing a bridge between user behavior, product metrics, and company goals;
Understand how you're progressing with your strategy, measuring the right things, and establishing connections between what you're measuring and your strategy;
Help PMs understand metrics and visualize the big picture of their work.
In summary, this framework allows you to begin rationalizing metrics from the company's strategy to unfolding your product's actions.
What data does your product deliver?
Your product rarely delivers all the data you'd like to evolve it. In this case, it's your role to ensure that the product collects and records this data.
I recommend considering what data it delivers after rationalizing your product's most important metrics.
The point of attention here is: don't overthink it. If you need to talk to GPMs and other leaders, try to expedite this process, so you don't get stuck with your product.
If you don't feel any openness, take the next step and move forward to understand what data your product is delivering (while studying the best time to align with managers and stakeholders, if necessary).
Many times, a product was designed and built to solve a specific problem. For various reasons, it may not collect the necessary data for its monitoring and evolution.
Check if there's any data analysis tool plugged in - like Google Analytics, Adobe Analytics, Amplitude, etc. - and study the data already collected. For this, you can use a sheet of paper, an Excel spreadsheet, or whatever you want to jot down the data being delivered.
The golden tip here is to dedicate a good amount of time in your schedule to these studies, to extract as many insights as you can.
From personal experience, I often grab an Excel spreadsheet and start extracting different types of queries. I begin by organizing things on sheets of paper, even before having a first structure in Excel.
The important thing is to start, no matter how disorganized and 'ugly' it may seem at first.
And if the product isn't generating data, well, it's time to roll up your sleeves.
Structure Your Product's Data Collection
Not all products collect the necessary data for their monitoring. Not every company has dedicated data analysts. And not every Tech Lead or CTO develops with data in mind.
PMs may face difficulties when this happens, but fortunately, it's possible to address this situation by taking the lead.
With a Miro or a Mural, you can map out your product's journey and insert the data that can be collected. This will greatly assist the Engineering or BI team in conducting a POC with the best Analytics solution or even structuring the database for the product to perform the necessary collection.
Before that, though, you need to know your product's objective.
Let's use a hypothetical example of a Mercado Livre (the biggest retail website in Brazil) team responsible for activation from the home page.
Before knowing what to metricize, they need to know which business objective it responds to.
When you know your product's objective, it's easier to establish a focus on what needs to be extracted.
This is important because we have ingrained the idea that we need to metricize everything but, when it comes to data collection, this rationalization is harder than it seems.
Bring a Data Scientist or a Business Intelligence Analyst into this discussion - if possible, of course. The Tech Lead of the team is also an essential counterpart. What matters here is to take the first step, so that your product starts collecting data.
Still using Mercado Livre as an example, knowing the need for:
Smooth purchase: that is, without friction in the journey, ideally leading to a product page that allows the purchase to be made without any doubts;
Buying a product on sale: from the home page, making it easier for the user to find the best offers on the site.
There is a direction of what to prioritize for metrification. Using Mercado Livre's home page as an example, it is important to understand how it influences the buying journey.
In a nutshell: how the home page contributes to the user making a purchase.
Since we are talking about the home page contributing to the purchase process, it is necessary to go further. In the end, we want to know if the user who purchased on Mercado Livre passed through the home page.
Understand your product's intersections
From this perspective, the ideal journey would be:
Accesses the Mercado Livre home page → Clicks on an offer and accesses the offer's product page → Navigates the product page → Clicks on the buy button → Completes the purchase.
But you, as a PM, are not responsible for the entire journey. You need to make sure the home page directs the user to the purchase page - which may be the responsibility of another product team.
In this context, you would need to know:
Of the purchases made on Mercado Livre, how many started from the home page?
Did users type mercadolivre.com.br in the browser?
Did they come through search engines? (Google, Bing, etc.)
Did they use any Mercado Livre widget (whether through a browser or an LP, for example)?
Of the users who passed through the home page and reached the product page, how many progressed in the journey?
How many clicked on the "Buy" button?
How many abandoned the cart?
How many completed the purchase?
With all this properly tagged, you can already structure a purchase conversion metric from the home page (I made this emphasis to remind that purchasing a product on sale would be the most important metric in our example).
With the evolution of this tagging and analysis, you can see where the user goes, where they tend to get 'stuck,' and even know which payment method is most used. In the future, this information can provide valuable insights and contribute to a backlog that generates more value for the business.
Dashboard? Start with a spreadsheet
Once you can see your product's data, the next step is to have at least some monitoring.
PMs dream of beautiful dashboards in Power BI or Tableau, but the reality is a bit harsher: sometimes you need to search for this information in a database or take a deep look at a data analysis platform, like Google Analytics, Amplitude, among others.
Since the reality of those structuring data in a product is rawer, don't be afraid to start small, with Excel or Google Sheets.
Create your columns, extract as many tables as necessary from the database, and always try to go further. The support of the Data team is crucial, but don't forget that it's your responsibility to understand the product, question more, and, if necessary, get it to extract more data over time. Involve your leaders in this construction
As a PM, you need to have a vision of how your product generates value for the business. This importance is even greater for the leader, who can very well contribute to this construction.
Even starting small, align with your direct leader what you are mapping about data. Most of the time, they have important considerations and can bring a business perspective that you haven't captured yet.
If there is pressure from stakeholders for new features, count on your leader to emphasize the importance of making your product generate data. Clearly state the benefits of this vision and share your first drafts, so that this construction is richer.
In addition to helping deal with stakeholders, leaders can mobilize teams and data specialists in your company for this construction.
With this alignment, the dream of having your product dashboard may be closer than you think.
Conclusion
Understanding your product's data is the first step in understanding how it truly contributes to the business. When you don't have this visibility, the PM can take the first steps, sketch out which data they would like to collect and create a data analysis routine.
Before that, however, the PM needs to understand the business expectations surrounding the product they work on, so they can focus on what to extract as a priority.
Don't be afraid to ask for help from the Data references in your company and, especially, from your direct leader. Also involve the technical team and UX, who can help generate an understanding of what needs to be collected from the product.
The beginning may be a board full of post-its. However, down the line, your product may deliver the data you and your team need to effectively contribute to business evolution.