Our previous level of “we get the gist” clearly hadn’t cut it. This time, we knew we needed to dig in until we understood the complex data more deeply.
We met again with the product manager and the data science team. We needed them to help us understand what data customers care about when trying to perform audits.
We left that meeting prepared to brainstorm how the newly-explained data would translate into a homepage that met the design principles we’d constructed.
But we had a problem. We still just had a bunch of data. We didn’t yet have information, and information had to be the minimum designable unit if we had any hope of providing real value with our designs.
Time to make sense of the data.
After an afternoon slicing and dicing, pushing and pulling, combining and deconstructing bits and pieces of the sample data set we had, insight struck.
Now, we no longer had “a few nuggets of wisdom we could throw on a homepage and trust the system will generate more”, as we had last time.
Instead, we discovered 3 flexible, sustainable, meaningful ways of categorizing both current and future data emerged from our analysis that would allow customers to know where to focus attention.
- Systemic problems – problems that impact at least 20% of the organization, and need holistic, widespread solutions
- Repeat offenders – problems that impact specific locations month after month, requiring individualized solutions
- Emerging patterns – issues that don’t yet exceed defined thresholds, but that could become problems if they continue
While each company’s actual problems would differ, as would the thresholds of what constituted a “problem”, the categories were solid.
After validating our newfound understanding with the data science team, we finally had accurate, unambiguous content feeding our design brainstorm.