October 4, 2018
What Good is Data Without Insights?
Adopting the Four Commandments of Data Analytics, From Deliberate Collection to Confident Predictions
We collect so much information – there are data points everywhere, all filed into neat reports that you can color code or export to Excel or drop into a PowerPoint deck. You’re collecting more data than you could probably ever use or know how to interpret. “Data-driven” is certainly the buzz word of the moment, so the more you collect, the better you feel.
But collecting data just for the sake of it is a waste of time. Collecting it without putting it in the hands of trained analysts? Also not a great use of time. Data is supposed to make us smarter, more efficient, better able to perform our jobs—but only if it’s collected logically, interpreted thoughtfully, and turned into actionable insights. From collection to prediction, following the Four Commandments of Data Analytics below will help you put your data to work and deliver actionable answers to the questions that may be impacting patient recruitment.
Be prudent with the data you are collecting. More is not better – in fact, more data can end up clouding valuable information and bog down your analysts. You should have a vision of how you want to use the data you collect. Before we even begin to think about building a data collection plan, we spend time working to understand what matters most to our sponsors. This way, we ask the right questions and find the answers they care about. The only way to find a balance between data collection and the time it takes to analyze that data is to recognize priorities.
“The only way to find a balance between data collection and the time it takes to analyze that data is to recognize priorities.”
Know the Problem
It’s one thing to report the data. It’s another thing entirely to understand the story it’s telling. Showing that fewer referrals were screened than predicted is only helpful if you’ve taken the next step and dug around to figure out why – is the site understaffed? Out of materials? Are the referrals largely unqualified or is there a new competing trial? Problems vary, and variables interact. Turning data into insights helps us develop and recommend unique solutions for whatever problem the data uncovers.
Patient recruitment should be objective. There will always be variables we cannot control – if a patient drops out of a trial because they had to move across the country, there’s nothing we can do about that. But what we can control and analyze, we should. Concept testing our creative materials, going through validation with target audiences, refining user experience with tactile materials and websites – these methods all produce a huge amount of actionable, objective data. When sponsors demand “data-driven” solutions, it’s because they are eschewing guesswork, and rightfully so.
“Turning data into insights helps us develop and recommend unique solutions for whatever problem the data uncovers.”
Predict with (Cautious) Confidence
Modeling and forecasting begin with data, but the analysts turn it into insights. Without trained analysts, we are bound by what has happened in the past. Knowing past behaviors is important, absolutely. It can help us adjust projects and refine tactics so as to not repeat mistakes. But we want more than to analyze the past – we want to anticipate the future when we can. Predicting when and how potential patients will travel through a funnel – where they are, when they will screen, how many will randomize – lets us make adjustments before anomalies grow into problems. It’s how our analysts use data to forecast outcomes that gives our sponsors confidence in our models.
Ultimately, these four commandments come down to one thing – understanding the difference between data and insights. Information is not, in and of itself, valuable. It needs context, parameters, interpretation. Employing trained analysts who can serve as translators between the information and the story contained therein is where the real value lies.