In the last decade, product managers of digital products have managed to create mind-blowing smartphones, self-driving cars, and virtual reality devices without product analytics. So, why would product managers need to change their way of working now?
Product analytics , the activity of measuring how end users engage with a product on an ongoing basis, can dramatically impact the daily work of a product manager. This article focuses on how product analytics influences two key responsibilities of every product manager: feature road mapping and software lifecycle management.
Feature Adoption Analysis
Wouldn't it be great to know how and how much customers use your product? As a typical product manager, you collect input for your roadmap from multiple sources, including your customers, competitors, market analysts, and internal stakeholders such as developers, support engineers, and sales and marketing folks.
You probably match the input against your product strategy (or the next quarter's sales quota if you are more in a survival mode) and draft your feature roadmap. But, all this input information is biased. It doesn't matter how much time you spend with key customers, power users, or novice users of your product. The information will always depend on a subjective opinion as long it represents the opinion but not the actual usage patterns of selected users. It is more likely that end users will highlight what is missing or doesn't work as expected than bring up the things that work already and they use every day.
Feature adoption analysis with 24/7 tracking of the actual usage of your digital product opens an entirely new perspective into what your customers use, when, and how often. Product analytics solutions, such as Qt Insight, record and visualize feature adoption over time across geographies, user groups, and different devices. Feature adoption analysis records an event for every user interaction on product elements you marked as relevant to be studied. These events are uploaded to a cloud solution anonymously and then aggregated to meaningful charts. Suddenly, you will understand whether a feature is used within the first 2 minutes of the customer journey, is used 20 times a day, is used a second time (ever), or is used only once a year.
Image: Screen capture of Qt Insight Beta Usage Report
A feature being used only once a year (or even less) might not be bad if that's what you want. Let's imagine you have a feature that allows your product to be reset to its factory settings if the end-user changes. You really want this feature to work to give your product a long lifetime value and support a cyclical economy, but you wouldn’t expect it to be used frequently. Whether or not the frequency of product usage meets your design expectation is what matters.
Features that you thought deliver a lot of value but which do not find much adoption in everyday life are such that deserve your attention. If users do not find or understand functionality, you must ensure that you place or communicate it better. Alternatively, the feature might not be as valuable as intended, and you need to reconsider whether you defocus the enhancement of this feature and reposition the feature in the UI navigation not to take up unnecessary real estate. You do not want to build a bridge to nowhere by extending a feature that nobody uses.
In the fight against becoming a feature factory, product managers can use feature adoption analysis to make road mapping driven by data. Feature adoption analysis serves as a complementary input for making product decisions.
Software Lifecycle Management and Dead Code Removal
Would you believe me if I claimed that 25% of code in enterprise software* is never used? Could you imagine that 80% of the code in cloud solutions** is rarely or never used? These results are based on actual data researched by the Technical University of Munich and Pendo.io, a significant player in the cloud solution analytics market.
As product managers, we are not eager to produce features that nobody uses. If 25% of the software code we manage is not used, then we carry a significant amount of dead weight with us. Unless dead code is identified and removed, we (or, to be precise, our software developers) keep maintaining it. The maintenance includes upgrading third-party libraries, scanning the unnecessary code in security and penetration testing, keeping its unit and graphical user interface test cases updated, and modifying it for newer programming language versions or compilers. So, even if nobody reports any bugs for dead code – which isn't really a surprise because nobody uses it – there is still a significant amount of work related to dead code. Not to mention that the dead code bloats the software, which causes unnecessary storage or memory consumption.
Image: Screen capture of Qt Insight Beta User Flow Visualisation
Product analytics solutions typically provide means to show where users navigate in the graphical user interface identifying views and UI elements that remain underutilized. Product analytics is a powerful approach to fighting software erosion over time.
Beyond Feature Management
Product managers get more out of product analytics than the above-mentioned feature management tools. Optimizing the onboarding experience, improving time-to-value, identifying potential premium features, adjusting the UI navigation, and A/B testing of new features are only some of the other relevant product analytics capabilities.
Image: Screen capture of Qt Insight Beta Monthly Dashboard
The examples in this article around data-driven roadmapping and more intelligent software lifecycle management stress the relevance of product analytics for today's product managers. So yes, product managers need product analytics to create even better products than they did in the last decade and stay relevant in their jobs.
Qt has started to help companies with its own product analytics solution, Qt Insight. If you want to learn more about the benefit of Qt Insight or try it out, please check for more information on our web page here.
Interested in our eBook "Product Manager's Guide to Analytics"? Download your free copy here.
*Source: Research of Technical University Munich, “How Much Does Unused Code Matter for Maintenance?”, 2012
**Source: Study of Pendo.io Inc., “The 2019 Feature Adoption Report?”, 2019
