I want to jump right in:
If Sir Francis Bacon was right and knowledge is power,
product analytics is the key to making excellent products.
Product analytics helps you to understand exactly who is using your product, why, and how, and gives you the data that you need to make decisions based on real-world information rather than speculation.
What this means in practice is that product managers need to add yet one more feather to their caps with a good working knowledge of product analytics.
But where should product managers that wish to become analytical product managers begin?
In this article, we are going to go through the basics to get you started. We are going to start with what exactly product analytics is and what it can tell you. We will then look at both the hard and soft skills that product managers need in order to leverage analytics, tools to utilize, and pitfalls to avoid.
What is Product Analytics?
Product analytics is all about gaining useful insights from the first-party data that you acquire from the users of your product.
First-party data is the information that you yourself collect about your users.
So, this will include things such as:
- registration data,
- the data of what they actually do in your product,
- what pages they visit,
- what functionality they utilize,
- what they buy and so forth.
This data can be anonymous, such as the kind of data that you collect through Google Analytics, or it can be data that you gather and link with user accounts, for example using a CRM, a customer relationship management tool.
While analytics can reveal a wide range of insights, and they differ from product to product, there are some insights that are universally considered valuable.
- Analytics can reveal who is using your product and why, what is likely to make them continue using the product, and what factors have the greatest influence of user churn.
- It can show you potential customer bases who aren’t using your product but should be.
- Analytics can show the methods of communication that most effectively recruit new customers and get them to an “aha moment” within the product that makes them valuable customers.
- It can reveal exactly what point within a product a user needs to reach in order to see its value and want to continue using it.
- It can show the roadblocks that trip customers up and prevent them from being able to do what they want with the product.
- Analytics can allow you to segment your customers to identify the most valuable customers that need to be prioritized.
Product analytics can be achieved using a variety of tools, but it always combines methods of gathering the data and visualizing the data to give it meaning.
Product analytics is a quantitive analysis of your product, as opposed to surveys and interviews, which provide qualitative analysis of your product. While both sources of data are valuable and important, product analytics is considered “hard data” because it is based on actual events. Survey and interview data, however, lack this “hardness” as people are notoriously terrible at accurately self-reporting on their actions and behavior.
How to Become an Analytical Product Manager?
Being an analytical product manager is a mindset more than anything else.
It is a commitment to making decisions based on real-world data rather than speculation. It is also the ability to link business goals to data, and knowing which parts of the business can most benefit from a data-driven approach.
This requires a range of hard skills and soft skills. While there is no limit to how deep you can go, I’d suggest starting with the following.
Hard Skills Required for Analytical Product Managers
1. Basic Data Competency
While, as a product manager, you might not do much of the actual data-wrangling yourself, you should have a basic idea of how to obtain and analyze data so that you know what is possible, and therefore how to ask the right questions.
It means that you can have meaningful conversations with programmers, analysts, and providers of analytics software.
The most committed product managers who want to be able to analyze their own data without always having to rely on a member of their team will also want to learn SQL. The majority of data analytics tools are SQL or SQL-based.
2. Advanced Excel Skills
Excel might not be new or fancy, but it is still one of the best tools available for managing and crunching data without coding.
All data is exportable into excel.
You won’t achieve this with your basic spreadsheet, you will need to learn how to create pivot tables and use macros to make your data dynamic and create meaningful visualizations.
Learn advanced Excel at the Udemy Academy.
3. A/B Testing
Learn how to use hard data to choose between various options by using A/B testing.
This is basically where you publish both options, letting some users see option A and others see option B. You then analyze the result of each option to choose the best one.
Consider investing in tools such as Optimizely.
4. Prototyping and MVP
Analytics product managers need to understand how to develop prototypes that are a minimum viable product (MVP) to test with users.
This means a prototype that gives enough of the final product that it is meaningful to users, and therefore produces meaningful insights, and doesn’t leave users distracted by what is missing.
These are essential for collecting data on new product ideas, rather than just user feedback. As we have already said, people are terrible at self-reporting, so analytics from the actual use of the prototype can tell a different story from tester feedback.
You can make clickable interactives with Protio.io.
Soft Skills Required for Analytical Product Managers
The soft skills required to be an analytical product manager are really an extension of the general soft skills of product managers.
Among the most important is an intimate knowledge of the product and the business, so that you know what questions are worth asking of the data.
You also need the ability to tell stories with the data. Not everyone can look at a spreadsheet or a chart and understand the information it holds. You need to be able to tell the stories of the data to stakeholders and team members in order to make the case for action and investment.
Tools for Product Analytics
There are hundreds, if not thousands, of tools out there that can be used to delve into product analytics.
But remember that analytics relies on the data that you can collect about actual users and what they are doing. Before you have users onboard you are doing exploratory research.
When you do launch a product and start acquiring users, it is often a good idea to start out with free tools to get an overview of what is happening with your users and product, before investing in more comprehensive and expensive tools for more sophisticated analytics.
This overview analysis can help you to identify which tools are likely to be of the greatest use for your product and business.
Some of the top product analytics tools on the market include:
Amplitude – allows you to collect and analyze data on what your users do within your product so that you can get a clear idea of the user journey, both on an individual and collective level.
Google Analytics – a free web analytics tool that allows you to see who is using your site or product, how they were referred to it, and what pages they are looking at.
Heap Analytics – code that tracks everything event that happens within your product without you specifically needing to designate it for collection. It can then run reports off retrospective data, so you don’t need to wait for reports to pay dividends.
Kissmetrics – collect data on what users are doing within the product and send them automated notifications based on specific actions.
Mixpanel – similar to Google Analytics, but instead of tracking pages, it tracks events, such as clicks and scrolls. It is also possible to link this to individual user accounts if this data is available.
Segment – allows you to streamline your data collection so that you can use a single code base to collect data and send it to all your different analytics tools.
Pitfalls to Avoid
When you are new to working with data, there are a number of common pitfalls that product managers need to learn to avoid.
1- Confusing correlation and causation
When you see a correlation between elements of the data, it can be tempting to always assign causations.
This is a mistake I’ve stopped making not a long time ago.
For example, you could identify that long-term users, who use your product for more than a year, are more likely to use a certain feature than short-term users, that use your product for less than three months. You might then be tempted to say that this feature is a major contributing factor when it comes to retention.
But this is not a given.
Perhaps long-term users, because they spend more time in the product, are more familiar with features and therefore more likely to use them. This could mean that the correlation is reversed and investing in this feature will do very little to help you retain customers.
Project managers always need to be on the lookout for red herrings that can fruitlessly divert time and resources.
2- Underprioritizing visualization tools
Visualizations aren’t just about selling the insights from data beyond the immediate team, although they play a vital role here as well.
The right kind of visualizations can provide new and unexpected insights.
The big mistake that many people make is to use visualizations that they are already familiar with or appeal to their aesthetic, rather than considering the visualization that is best for the specific data set.
Visualization should be seen as an essential tool and not a piece of marketing at the end, and therefore should be chosen at the beginning of the process based on a defined goal.
3- Failing to revalidate data models
Often we come up with approaches to analyzing data which we think explain the data that we have, and then we leave them in place.
But the veracity of models needs to be constantly re-evaluated in light of new evidence and new data.
If you don’t do this, the data models used lose their value as they are no longer an accurate predictor of the behavior being measured.
4- Measuring everything
Sometimes, especially when you have a dozen of different metrics at hand, you just want to crunch all the data.
I know how tempting that is!
And you might believe if you do that, insights will surface.
But that is not the case.
While a high-level look at the data can help give you context and point you in a certain direction, data works best when it is being used to prove or disprove a hypothesis.
You need to be using the data to test the dual hypotheses, such as that x has a significant impact on y, or that there is no significant correlation between x and y. Only in this way can data be appropriately leveraged to provide useful insights into the product.
5- Underestimating the power of integrations
I am aware that setting up a working product analytics structure and adding all the integrations is difficult.
It’s basically torture.
But it rewards you and your business. How?
Integrating the different platforms you use to manage user experience and customer relations with the analytics platform you use doesn’t only make the insights you’ll get more valuable than ever, it makes them as precise as possible.
For example, say you are using a no-code user onboarding software such as UserGuiding to create user onboarding experiences with various elements such as interactive walkthroughs, checklists, and in-app messages. (without any coding BTW)
And UserGuiding provides you with insights on the performance of these elements. Yes, you can use this data to improve these elements.
But you can also integrate UserGuiding with the analytics tool that you use and see what the user onboarding experience is affecting in other parts of the user journey.
The integration gives you the eye of the Argus, and what you do with the insights you see is up to you.
The ability to gather data about who is using your product and how can take a lot of the guesswork out of many product-related decisions.
It allows product managers to dive into the minutiae of what is happening in the real world, and therefore make decisions based on evidence, rather than informed guesses based on experience.
But while data and analytics present an important opportunity, it is also easy to get it wrong and waste a lot of time and energy asking the wrong questions and drawing the wrong conclusions. That is why product managers need to prepare themselves to work and thrive in this new data-driven world.
Frequently Asked Questions
Why is product analytics important?
You can’t improve what you can’t measure. To make sure that your product is getting increasingly successful and address the problems if it isn’t, you need product analytics.
How can I measure product analytics?
Setting up and maintaining product analytics is not as hard as it sounds. Finding out the right analytics tool can provide you with all the metrics you need related to your product’s performance.
Which product delivers all analytics capabilities?
Although Google Analytics, Amplitude, Heap, and Mixpanel are great products for analytics, SAP Analytics Cloud is famous for enabling all analytics capabilities in one product.