Unlocking Success: Leveraging the Power of Data in Digital Products
What you’ll find in this article:
A comprehensive roadmap for data-driven digital product development, encompassing the entire process from data collection to the conversion of insights into actionable strategies.
After having several years of product management experience, I’ve been involved in too many feature decisions and had the chance to observe decision-making processes of stakeholders in different ways. Some of them were based on blind insights, some on know-how collected through many years of experience in business, and some on very well-structured analytics. Regardless of what approach is applied, I’ve seen both failure and success stories. Today, I’d like to discuss a common aspect of the best practices for product decision making, which is essentially condensed to “Data-Driven Product Management.”
What is Data-Driven Product management approach?
Data-driven is an approach to product management that highlights the use of data and statistics to assist with and guide decision-making throughout the product lifecycle. It involves using numerous data sources to drive development, optimization, and decision-making, such as customer feedback, market research, product usage metrics, and user behavior analytics.
Here are some key components of a data-driven product management:
1.KPIs (Key Performance Indicators): Creating explicit KPIs that are aligned with company goals and using them to assess the effectiveness and impact of product efforts. KPIs are quantitative measures that can be used to evaluate product performance, validate hypotheses, and track development over time.
In order to measure something, it is important to start by thinking about what should be measured to get a snapshot of the performance. At this point, defining the KPIs for sure is the first step. Defining the correct KPIs helps identify what data channels should be used and what implementation needs to be done to measure them.
2. Data collection: Collecting relevant and usable data is crucial. This may contain user activity data, customer feedback, market research, A/B testing, analysis of rivals, and other relevant indicators.
Usually, Data collection is not as easy as it looks. It is crucial to select the correct tools to collect the data you need. When it comes to digital products, probably Google’s Firebase and Analytics 4 are the most common ones, yet there are a bunch of others.
To get an overview of the performance of any feature, data should be collected at all levels. While many types of KPIs rely on user interactions on the interface, there are some others that should be collected from the product database, user feedback, market stats, etc. The combination of data from different layers helps us gain a comprehensive view. Also, in this way, it becomes possible to validate the accuracy of the collected data. You’ll have a chance to compare the figures from each level and find out if they relate reasonably.
3. Data analysis: analyzing the data gathered to uncover patterns, trends, and insights that can guide product strategy and roadmap. This may involve techniques such as statistical analysis, data visualization, and predictive modeling.
In data analysis, everything starts with the question, “How many?” Once you have all the analytic structure settled, you are able to answer questions such as “How many users do we have?”, “How many items have we sold?”, “How many registrations have we got?”, etc. And then the adventure begins. It wouldn’t be helpful to answer “how many” questions in order to find out how to improve your product. Therefore the follow-up question “why” arises. If you work closely with business decision makers, you must have heard at least one of these questions: “Why sales are not high enough?”, “Why do users not engage with the app/website?”, “Why is the conversion rate that low?”, “Why don’t users subscribe?”, etc. At this point, data analysts do a deep dive into the data. They analyze the conversion, create funnels for each flow, calculate the bounce rate on each step, analyze the user segments, and work on revealing patterns based on user behavior.
4. Product roadmap and prioritization: Creating a product roadmap and prioritizing features and enhancements based on data insights. Data assists product managers in making informed decisions regarding resource allocation and feature priority by identifying high-impact possibilities, potential hazards, and areas for improvement.
After a detailed analysis, the product team gathers the insights. Pain points and possible opportunities are discussed. They convert the insights into ideas. Once the roadmap and prioritization are clear, the design phase begins.
Each improvement should be measurable and should have an objective result. During the design phase, it is crucial to consider “how to measure” the success of the change. Otherwise, it wouldn’t be possible to prove whether the improvement achieved the goal.
Measurement is the key step
Iterative development is a method of product development in which data is continuously collected and analyzed to inform incremental product changes and updates. In this cycle, the measurement step stands to ensure that the product evolves in response to user requirements and market conditions.
Real-life example :
How Data-Driven approach increased user numbers by 26% !
Pain Point: A mobile app’s registered user count was lower than expected.
Data Sources: Google Analytics, AppStore, PlayStore, reviews, and user feedbacks.
Measurement Phase:
· Conversion rate has been calculated between users who have downloaded the app vs. registered users.
It was clearly seen that a large number of users from a certain region could not complete the registration flow after downloading the app. The fact is that registration is mandatory in order to use the functions of this application. So, having a good conversion rate at registration is a must.
· User Funnel has been created in Google Analytics to visualize the bounce rate of the users who have started the registration flow.
All user interactions have been tracked by Firebase Events on each screen view and click. This way, we had the opportunity to see exactly at which step the users were struggling.
Ideation Phase:
After conducting data analysis on the user registration experience, it was detected that there was a dramatic drop between the steps of the registration flow. Especially the drop was happening in a certain user segment. When it was deeply investigated, we noticed that the same user segment was located in one of the developing countries where education is limited and income is low.
The segment analysis showed us that we needed to deep dive into the user experience of that certain user group. We created a funnel consisting of each steps of the registration flow filtered by that particular user segment.
After seeing the funnel above, we thought that this must be an improvement area to increase user registration. During the ideation phase, it was decided to simplify the registration process by reducing the password creation complexity. Additionally, adding new registration methods to the app was prioritized in the backlog.
Design — Development — Roll Out Phases:
During the design phase, we also discussed what kind of additional parameters should be tracked in order to measure the performance of the change afterwards. Once they were all defined, the changes were developed. Development was followed by tests, including the verification of the measurement tools. And last but not least, the version was rolled out.
Result:
After the changes were published, the exciting wait had begun. It was great to see that the conversion rate between the steps was increasing as long as users updated their app version.
There were two ideas that we realized.
1. We simplified the password requirements. This change has reduced our bounce rate around 26%. In other words, our registered user count has increased by 26% compared to the previous app version.
2. We added a new registration method as an option. Previously, users were only able to register with an email address. Now they also have the option to register with a phone number. The addition of the new option has reduced the bounce rate around 24%.
At the end, we also found a chance to evaluate which registration method was performing better in terms of the success ratio of the users who could complete the whole flow. When the user abandonment was compared, the result was significant. As seen below, method 2 was around 15% more successful than method 1.
Bonus : Mitigate the risk
As product development teams, although we usually think we come up with great feature ideas, sometimes things go wrong after rollout. If the change has a possibility of having a high impact on the user experience, then there are three methods that should be considered, which I highly recommend:
1. A/B Testing
A/B testing is a digital product management technique that compares two or more versions of a feature or user experience to see which one performs better. The best part about A/B testing is that, once the most performant solution has been identified, you can easily spread that solution to all users without releasing a new version. Instead of depending on assumptions or opinions, it enables product managers to make data-driven decisions by analyzing the effectiveness and impact of prospective adjustments or new features.
2. Feature Toggles
Feature toggles are mechanisms that enable product teams to activate or deactivate specific features, functionality, or a design within a digital product. They provide a way to control the availability of features. Imagine that you detected an issue with a recent feature you implemented. With a simple feature toggle, you can easily turn it off until the issue is resolved. This way, you can avoid the possible unfortunate impact on the user experience.
3. Staged Distribution
Staged distribution helps with the management of infrastructure and scalability issues. Product managers can examine the impact on servers, performance, and other technical elements by progressively increasing the user base. It enables proactive monitoring and resource adjustments as needed, delivering a seamless experience for all users.
