Let me start by saying that I am a firm believer in leveraging data to help drive innovation. What drives me crazy is when organizations get caught in the cycle of looking for the PERFECT data. Guess what? It doesn’t exist! In searching for the ‘sasquatch’, we lose valuable time that we could be using to get our idea out to the market or organization faster. So how do we balance the need for data with the need for speed (if we want to be ‘first’)? Here are five tips to help.
Tip #1: Accept that we will NEVER have perfect data. The reality is that while we all have access to reams and reams of data, it will never be pure or in the perfect format, cross tab, reference table, query, report, etc. that will allow us to feel 100% confident that we have the information we need to eliminate ALL risk. We need to have enough data to back up our decisions around innovation, while recognizing that sometimes 85% confidence is good enough to invest in prototyping innovations in products, services or systems.
Tip #2: Identify what data you need early. As you define your innovative idea, take a bit of time to determine what kind of data you will need to help prove out your concept. From there, identify what you currently can access within your organization or industry and where there are gaps. Where gaps exist, identify proxy data or experiments you can run quickly and cheaply to get some directional data. Proxies and experiments are not perfect, but even if they add a 10% confidence factor to what you have been able to gather via existing data, it is better than nothing.
Tip #3: Do a gut check. When someone on your innovation team or a senior leader asks you for more data or suggests you dig ‘deeper’, ask yourself (and them), “Is the time and effort collecting or analyzing this additional data going to give us a substantial increase in confidence? How likely is it that the data would significantly change the direction we are pursuing?” If the additional data has limited potential to impact the project, stop the madness! If it is going to cost $100,000 and six months of resource time to get to 90% confidence levels from 85%, is that the right thing to do?
Tip #4: Seek validating sources of data. Sometimes we pull ‘data’ out of our systems and apply it without considering how valid the data may be. It is important to look at what the data tells us and seek to validate it in some way. Does it make sense? Is it consistent with other related data or does it appear contradictory? Who can we talk to about the structure of the data to ensure it is capturing what we are looking for? Are there other factors we need to consider that can be impacting the data? We know the data won’t be perfect, but it is helpful to understand where it is flawed so we can establish and seek ways to test assumptions.
Tip #5: Recognize that qualitative data is helpful. Yes, it’s true that qualitative data is more difficult to sift through, but it can be highly valuable in the innovation process as it allows us to dig into the ‘why’ of the data. Here’s a great example of what I mean:
If you are looking at a new product, service or system, a common concern is customer or employee acceptance of the innovation. Often, innovation team participants and/or leaders jump to the need for a comprehensive survey that requires an external consultant and a large cheque, only to get results that are quantitative but not helpful (“I love it” or “I hate it”). I know this sounds crazy, but what if we went and talked to people? What if we could show them a prototype (even if it is just a diagram!) and asked for their input and feedback? The ‘data’ we collect from those interactions is more helpful to iterate our idea than a nice survey report that leaves us wondering why people responded the way they did. The good news it is that not only is that data more valuable, it is often cheaper and faster to get our hands on!
Most innovations shouldn’t take ‘years’ to get off the ground. If you are finding that your process is a lot slower than it should be, take a look at how you are leveraging data. Are you getting caught in endless data digging or analysis paralysis? If so, stop, take a step back, revisit your approach, and repeat after me: “There is no such thing as perfect data”.