Building Insights with Product Usage Data
A process for delivering informed product decisions.
As humans, we are easily influenced by our own biases. Often mislabeled as “product intuition”, product managers can easily fall victim to unvalidated instinct. A sound decision-making process relies heavily on the quality synthesis of the information that you have at your disposal. Strong investments in both your team and product have the potential to raise you to the top but without relevant and reliable data, it becomes difficult to know if your investments will truly enable you to execute your goals.
From a digital product perspective, insights derived from product usage data can serve as an antithesis saving many from costly mistakes formed solely on gut instinct and anecdotal evidence. While requiring effort to manage, prioritize, and analyze, establishing a process that delivers accessible data with foolproof analysis tooling can enable you to effectively carry your team forward with confidence.
Below, a fictitious coffee franchise, The Coffee Caché, has been created to help illustrate a path that can lead your team from analytics definition all the way through to analysis and insight.
Welcome to the Coffee Caché
Highlighted as an example, the Coffee Caché is a fictitious coffee house franchise in the Minneapolis metro area. Brewing upmarket coffees & teas, the Caché team focuses on serving adventurous caffeine devotees with unique beverages from a frequently rotating menu. Hoping to drive incremental connections with their guests, the Caché franchise made the decision to bring technology further into the mix. By launching a mobile-first loyalty program lined with incentives, they hope to engage more frequently with guests while fostering a deeper love for the coffee craft.
The Process
Business transformations of any size are packed with decision points. Previously, many Coffee Caché franchise decisions were based upon the anecdotal feedback received from customers, baristas, and franchisees. With the launch of this new mobile product, they quickly found that usage data is a powerful tool they can count on when evaluating their next strategic investments. However, without a team dedicated to this analysis work, they knew they needed to explore how they can execute regular analysis economically.
Easing into and operationalizing a data analysis function requires care. Using The Coffee Caché as an example, these steps can be followed to help identify which key items you may wish to track and how you might go about analyzing your data without a degree in statistics.
1. Start asking questions.
Begin by identifying the questions that tie your product, customer, and problem space together. Identify the leading indicators that will influence how your product meets your goal or vision.
For example, The Cachè team may be looking to build incremental sales by increasing redeemable rewards. Some of the questions they may ask could be:
- How many guests are converting and becoming accessible to us on the new platform?
- Which type of guest is most active within the app?
- How far away are users traveling when they order?
- How do wait times impact ordering habits?
- Which new products are guests interested in?
These questions will help establish your data points of interest. Holding a narrow focus on these questions will ensure that you can effectively glean insights from your product and protect your team from getting overwhelmed by sifting through every micro action a user may be able to take.
2. Move from points of interest to critical events.
Using your team’s questions of interest, begin by creating a concise list of the key activities that users will perform ensuring that they roll into each of your product questions. Quickly you will find that many of your questions will be able to be answered by looking at just a few events and the associated metadata. Looking at the Coffee Caché example, they will likely be able to answer their questions by tracking :
- When orders are placed
- The user's distance from the shop at order time
- The average wait time for an order
- When menu items are viewed
- How users locate the shop before ordering in the app
Having identified the events that will feed your team, prioritize this development effort early! As soon as you launch without analytics tracking, you will be left yearning for the data you decided not to collect.
Caution: Clean data collection is key! It may not be user-facing but the need to sanitize bad data weeks down the line can quickly erode trust in your data. Place an emphasis on ensuring that events are accurately reported, naming conventions are formatted appropriately, and metadata is uniform.
3. Create and visualize.
Synthesizing data in the pursuit of insight is a learned skill. Without making your data easily accessible, this process gets become difficult when you try to involve others. Relying on data lake querying only works if everyone has an accurate understanding of what event data represents, and how it should be reviewed. While your team may be sharp, asking folks to query and review the results on their own often results in a considerable amount of time being spent validating that everyone is reviewing the data in a consistent fashion.
Introducing a tool like Google Data Studio can help limit some of this stress. Prebuilt visualizations that clearly represent your data will ensure that everyone is reviewing the same information and remove the need for all to be intimately familiar with exactly when and how an event is triggered.
Getting started in Google Data Studio, you will be able to establish a connection to your live data. Connections can be established via Firebase, 3rd-party databases, or a host of other analytics reporting tools. (A list of all Google endorsed connections can be found here.) Google then provides you with a simple yet powerful WYSIWYG interface to begin creating visualizations. Allowing you to blend sources, clearly articulate the intended meaning behind events, and share results with a wider audience.
4. Learn to share.
If kindergarten taught us anything, it is that sharing is caring. Data Studio provides easy sharing and subscription settings. Whether you grant view access to everyone, add stakeholders to a weekly status email, or download reports and regularly distribute them in a deck, this tool can quickly become the first point of contact when you have questions about product performance and user behavior. Regardless of your visualization tool of choice, establishing a product dashboard will enable the rest of your team to help spot patterns and create insights that refine your product.
For example, check out this sample report built for The Coffee Caché team below.
So what did The Coffee Caché team learn going through this process? To date, franchisees were under the impression that their customer base was hyper-localized to their neighborhood. After reviewing their data, they have found that their Loyalty customers are willing to travel much further to indulge in their drinks. With this insight, teams began focusing on finding ways to convert customers and grow awareness of the program in order to cast a wider net of customers.
More generally, it can feel needlessly frustrating when you don’t have clean product data at your fingertips. Without carefully crafted data collection mechanisms, decision-making quickly becomes based upon a collective gut feeling. Investing in a flexible product dashboard can make data analysis more manageable and accessible to your team. Whether the focus is on tracking operational efficiency, strategic performance, or user behavior, centralized dashboards will empower more folks to contribute meaningfully and instill confidence that you are headed in the right direction.
As seen in the Coffee Caché example, Google Data Studio offers an approachable process that can assist teams looking to place an emphasis on shared data. If you are interested in learning more, check out their introductory video series!
Jack Cunningham is a data-driven Product Manager helping teams build informed digital products with Livefront.
