Digital is perhaps the most measurable form of media that ever existed. The temptation to measure everything is great because essentially everything is measurable. But before adding to the expense of analytics, it is important to challenge yourself to commit specific actions you will take based on the data.
I’ve seen far too many instances of clients buried in information overload, receiving in-depth comprehensive reports on all aspects of a campaign, but when asked what they are doing with the information, they don’t have an answer. Yet pharma has an unquenchable thirst for data, even though much of it often goes untouched and underleveraged. When a deep analysis is undertaken, pharma often overanalyzes and over does segmentation. Not much can or should be gleaned from a segmented sample size of 12, but there are instances when valuable time is wasted exploring and debating the findings of insignificant segments.
The scenarios above create a continuum from too little analysis to over analysis, so what is the right balance? How much should you invest in analytics? As with any strategy, a good data and analytics strategy begins with the end in mind. Start by asking yourself some key fundamental questions:
- What is the desired outcome?
- What is the most important objective the program should accomplish?
- What metric will best determine how successful the program is at accomplishing the key objective?
- What is the best leading indicator to measure success of the objective? (A leading indicator should be predictive. If the leading indicator trends positive, the metric that determines success of the objective should also trend positive, even though there may be a delay in the data.)
- What is the best lagging indicator to measure the desired outcome? (The lagging indicator should reveal if you are focused on the right objective, as you should see progress toward your desired outcome if everything is trending well.)
- How often will you analyze the data?
- What actions are you both prepared to take, and committed to taking, based on what the data tells you?
- When will these actions be taken? (Consider milestones and acceptable ranges of performance.)
These last two questions get you to set some ranges for acceptable program performance and they force a “plan B” alternative in the event that your program is not delivering on the results you expected.
When you ask these questions and go through the process of determining your answers, it’s important to focus on key indicators. Measure your objective, measure the best leading indicator, measure the best lagging indicator. They will clearly tell you if you are on track to achieve your desired outcome. That’s three pieces of data. Resist the temptation to analyze 100 metrics. More data leads to information overload, and analysis paralysis. The temptation is great to expand to another objective, another leading indicator, etc., and some programs will justify this—but most won’t. The most you should be measuring are three to five key data points, each of which should have a specific acceptable performance range, and each of which has key milestones that if missed, a specific defined action plan will be undertaken. You can’t manage what you can’t measure, and you can only manage a reasonable number of data points effectively.