How to Make Predictive Analytics Work for Your Business
Q: How can I incorporate predictive analytics into my business?
A: Say I run a pizza shop. If I knew exactly how many cheese, pepperoni or veggie pizzas I was going to sell on a given shift, I could have those ingredients on hand, and maybe even make them ahead of time, so the customers get their pies fast. And if I knew that half my customers order a large soda with their pizza, I could offer them a pizza/soda special to keep them coming back.
The goal of predictive analytics is to take the guesswork out of understanding your customers so you can maximize every sales opportunity, according to Mike Leikin, manager of data and cloud operations for iSeatz, a New Orleans-based company that provides custom online travel and entertainment solutions. We asked Leikin for his take on the promise of data-driven crystal balls.
How do analytics work?
Using your pizza restaurant as an example, let's say we discover that on Fridays we sell twice as many pizzas compared to Wednesday. With that information, we can use this week's Wednesday sales numbers to predict Friday's order with greater accuracy.
It's not complicated, but it can become infinitely complex if you're trying to analyze too much data at once, or simply don't know what to look for. Instead, stay focused on incremental improvements.
What software can help me?
You'll need a Business Intelligence tool. There are free, open-source tools out there, like R from Revolution Analytics, but they require sophisticated programming knowledge to use. Tools such as those from SAS and SAP provide plenty of enterprise-level power but are expensive and complex to operate. If you just want a way to visualize your data to make it easier to understand and follow, there are inexpensive dashboard tools like Geckoboard and Leftronic that you can try out for less than $100 a month.
Can I do this on my own?
Sure, as long as you ensure that the data going in the system is accurate, the analysis being done is correct and the outcomes are actively tracked. You may want help from an analytics consultant to set all thatup, though. And you'll want marketing experience--or business knowledge--to help you parse through the data and translate the findings into changes that bring in more revenue.