I’m often asked, “Does designing a BI tool mean utilizing data in your process?” Short answer: Yes. Quantitative data has become the data of choice over qualitative research. Now more than ever, data is playing an increasing role in our design process. To keep pace with the business, we must work with — and make sense of — product data. Analytics is a gift to designers, and there’s an increasing need to embrace it as a tool of our trade. Data also helps design teams marshal their resources and bring everyone in the organization closer to user needs.
At Zendesk, Design teams resemble a structure similar to product development. Within a triad of design, engineering, and business we’ve organized our own; research, data, and design. We apply these three disciplines to every project regardless of size. Designing this way fits with our brand, “Relationships are complicated.” It’s a bizarre assortment of perspectives and methods, but we’re driven towards a common goal: to gain a richer understanding of our customers’ behaviour. At our core, we believe studying the actions of a few will illuminate the intent of many. Data helps our process tremendously by allowing us to accelerate and focus research on the right users, uncover breakdowns in the interface, and model the projected impact of new designs. Below are three ways your next design project can benefit from data analysis.
Set the stage
Historical data is the perfect starting point for improvements as it provides benchmarks for current activity. By identifying the breakdowns in the interface, data can equip teams with a lay of the land and support or refute broken patterns. Using tools like MixPanel, Pendo, or HEAP, funnels can show precisely where customers are on their journey. Are trial customers converting? Are new customers following the intended on-boarding steps? Visualizations like these help designers understand the metrics of most importance and where there are drop-offs, if any.
Find (and recruit) the right users
Data helps designers understand the realistic use of product vs. the hypothetical. A standard deviation can show just how extreme edge case behaviour is. You can solve problems directly for the many while considering the outliers. If you think of your digital product as a ski resort, data helps you understand which users are milling about on greens, polishing their technique on blues, or barrelling down blacks. By identifying cohorts with tools like Segment, designers can test prototypes that cater to some—or all—of their user types. While I don’t recommend validating designs with just one audience, gathering feedback from impacted users usually yields the best feedback. Some tools allow you to send targeted messages directly within the segmenting flow. It makes gathering account IDs and emails for usability testing a breeze. Obtaining realistic use of the product in this manner takes less time than traditional research methods.
Drilling into unexpected behaviours can inspire new product directions that would otherwise go overlooked.
Spice up persona documents
Include product data and visualizations in your next persona document. By adding information like average session times, activity history, and monthly spend next to ethnographic data personas pack a stronger punch. Keeping persona documents in digital format ensures the data is always up-to-date. Apps like Segment take personas one step further by helping designers group users by traits and themes automatically. Rich, interactive, fluid persona documents like this help teams during design sprints but have extended use later to other parts of the organization. Product Marketing can utilize the same document to craft personalized messages to specific audiences or inform word choices in release notes, for example.
While data seems to be the information du jour, qualitative methods still have a significant place at Zendesk. Pairing traditional research with data analysis bridges the gap between the builders and the users better than either method could on their own. Insights beget insights, and the overlaps between qualitative research and quantitative data unlock mountains of useful ideas. The fun part is digging into the outliers to see what is happening. Blips in user behaviour are exciting for product owners and designers alike because it helps parse out anomalies vs. seeds for new features. Drilling into unexpected behaviours can inspire new product directions that would otherwise go overlooked. We verify anecdotes during customer interviews with product data to confirm the behaviour. Likewise, a researcher may investigate anomalies in a data visualization with a survey or usability test. Two approaches, same outcome.
In the end
Design has been historically hard to measure but incorporating more data throughout your process will boost the potency of any sized team. Data helps communicate the value of design in business parlance and appeals to the System 2 thinkers in the room. When done well, Design offers a framework for understanding the information regardless if it’s qualitative or quantitative. Remember it takes a village to build a successful product. Ruminate on data. Be inspired by people. Oh yeah, we’re hiring!