Predictive Analytics Preparation: The Journey
This week, I barely beat out a more worthy contender to advance to my fantasy football league’s championship game. Why am I telling you this? It’s not just to brag. Every week, I, along with more than 40 million other players (according to the Fantasy Sports Trade Association) spend time and often money on setting the team roster. I make my decisions on who to start, who to drop and who to bench based on past performance, but also largely on the readily available projected fantasy points. I am accustomed to seeing this number, this little bit of predictive analytics to make my decisions. All the information on past performance of the individual player, the team, the opponent, etc. boiled down into one easy to understand number is amazing! I started thinking about how this idea can help our association clients. What if we could predict how many people will attend an event if we have it in Austin, open registration 30 weeks prior and send 30 email marketing messages? What if we only send out 15 email marketing messages and do 5 mailings? It’s easy to get caught up in the excitement and rush off to answer these questions using some piece of technology. However, before we rush off to implement predictive analytics, we'll first need to do some proper planning. This will ensure reaching optimal potential long-term.
Predictive analytics can be equated to running a marathon. Most of us know how to run, so why not just buy some new running shoes and start? I think we can all see how that approach will yield a bad result, likely causing us to injure ourselves, yet it can be tempting to rush into predictive analytics the same way.
Here are the steps to ensure your analytics efforts are set up for success. Just picture yourself at the finish line with both arms raised in victory!
Four Steps to Predictive Analytics
- Predictive Modeling Conditioning. Make plans for a sustained effort to develop, test and refine predictive models to ensure they generate findings in line with your association’s goals. "Organizations run into trouble when they expect to [automatically] get amazing results," said Doug Laney, an analyst at Gartner Inc. "Predictive analytics projects are iterative and involve processes that need the regular testing of models."
- Gain executive buy-in. The process of implementing and sustaining a predictive analytics program is inherently complex. You’ll need endorsement from top level management to endorse and value the creative thinking that is an important part of the analytics journey. The most successful organizations have leadership that encourages fresh ideas and making data-guided decisions. See our CEO, Debbie King’s blog on aspects of change for analytics.
- Identify the budding analysts within your organization. Not everyone within your association has the experience, knowledge or drive to produce their own analysis. It’s unreasonable to expect that technology will be able to or should magically deliver the answer to all your burning questions. Technology is the tool to help the people answer questions and make informed decisions that lead to success. But it's the people that decide what to analyze and people who take action on the results. Find, encourage and train those in your organization who have an aptitude and desire for analytics. It will pay big dividends.
- Healthy data. Just like you can’t expect to fuel your body with donuts and ice cream while training for a marathon, you can’t expect to use unclean, incomplete data to accurately predict outcomes. Keep in mind that for predictive analytics, we don’t need perfect data (which actually doesn't exist outside a laboratory), but you will want to take steps to ensure it is as healthy as possible.
So, use that excitement for using predictive analytics and focus it on preparing your association for success. Just like I am doing with my Fantasy Football team roster.