In May 2016, Infinity Insight had the opportunity to participate in the largest Qlik conference of the year, Qonnections. We had a great time networking, learning, and partying!
For the first time ever, this conference was open to both partners and customers; the result was a really exciting and collaborative atmosphere that benefited everyone. It was particularly Qool (yes, this is a word) to see customer application showcases, where customers got to present the solutions they have implemented with QlikView and Qlik Sense. A notable superstar was Johns Hopkins, with their COPD and Perioperative services apps. These applications stood out because not only were they designed in a visually appealing way, they displayed KPIs that were actionable. This is the holy grail of business intelligence, and a valuable lesson for all of us in the world of BI, especially those who are more experienced—when faced with a business challenge, too often we find ourselves creating applications the same way we have for dozens of other clients, in an almost automatic fashion. We should always ask ourselves (and our clients, because we don't have all the answers, despite what we may wish!):
- Who will be using this application?
- What kinds of decisions do they need to make on a day-to-day, week-to-week, or month-to-month basis?
- How can I use their own data to help them more effectively make these decisions as quickly as possible and with minimal pain?
Here's an example from the world of big pharma, one of Infinity Insight's core specialties. When we, as BI developers, show up to a pharma client, let's face it: we pretty much know in advance what we are going to do. We are going to create line charts that show Total Prescriptions (TRx) and/or Market Share for our client's product and for their competitors' products over time. We are going to create a list of prescribers that show these same two metrics. We are going to show territory/region/area/zone comparisons to see how our sales reps are performing compared to one another. If we want to get really fancy, we will incorporate a geographic map of some sort. And that's all fine, it really is. But if we stop there, we are doing our customers a disservice.
The 'I' of BI stands for Intelligence, and the purpose of non-self-service business intelligence tools should be to help the business run intelligently. In my opinion, this does not mean creating an application that requires users to spend hours digging for meaningful data points, problem areas, or opportunities. For one thing, you risk that these insights will never be found, or at least not found in time. For another, unless all of your users are business analysts, every hour that they spend in your tool is an hour that they are not doing their primary jobs.
Bring problems and opportunities front-and-center so that users see them instantly and can react in a timely way. Use your Dashboard sheet to show actual time-sensitive/business-critical data points; think of a car's dashboard and you'll get the general idea. Going back to our pharma example, we can use the exact same measures we always do, but what if we plot our territories in a slightly different format?
Now we can clearly see the problem: quadrant 3. All my home office managers have to do is lasso these three territories and zoom in to see the high-level business problem.
But you also need to design for all users. The above chart is great in an app for managers, but what do my sales reps care about? Most BI developers don't know; if that's the case, it is imperative that you interview the man on the ground. After all, it is impossible to hit a target you cannot see, so opening your eyes is essential. But let's assume that I do, in fact, know the answer to that question: my reps care about achieving the sorts of measures on which they are compensated—typically, meeting a certain threshold of sales or market share in their territory. As a BI developer, my job is then clear. I need to quickly show my reps the individual data points than are affecting these measures the most, so they can react quickly, and in the most impactful way possible, before the fiscal quarter ends and the opportunity is gone forever. Perhaps one of my largest prescribers has had a statistically significant dip in new prescriptions (NRx) of my product in the past week, putting my volume goal in jeopardy. Perhaps another prescriber just wrote unusually high NRx of my competitor's product, putting my market share goal in jeopardy. Both of these data points need to be highlighted prominently on my Dashboard.
There will certainly be insights that only humans can derive through deep-dive analysis. But let's not make our users work too hard when there's a better way. Sales reps are busy, but they should be busy selling. Managers are busy, but they should be busy managing. Let's take a page out of Johns Hopkins' book and create easily actionable insights whenever possible.