In today’s tidal storm of data, we are constantly combating analysis paralysis. We invest so much time into analyzing, evaluating and validating data-driven decisions, and not enough time taking action with it.
We need to rethink the way we make decisions and let go of analytics perfection.
Coming from an engineering background, I revel in absolutes. However, instead of suffering from data overanalysis, we need to accept that "good enough" data analytics is the new great and free up resources to focus on action.
Finding the 'Good Enough' Ideal
Relying on good enough analytics depends on the pedigree of the data and what action you are going to take. We see good enough analysis all the time. The first step in knowing you can rely on it is asking questions about the data.
Take this example: If data shows 70 percent of your customers are wheat farmers, then you may propose specializing your entire business around wheat farmers — a seemingly great idea.
However, would you still pursue building a company catered to wheat farmers if the percentage was based on a sample size of 30 customers? Probably not. What if it was based on 100 customers? Maybe. What if it was based on 100,000 customers? Likely.
Somewhere between 100 customers and 100,000 customers there is a level of ambiguity — and that in-between phase represents the notion of good enough.
Understand the Data Pedigree
All metadata is important, but in todays’ decision-making climate, not everyone is taking that into consideration. By asking the following questions you can ensure your good enough analytics is that much better:
- How timely is the number and what data is it based upon?
- How relevant is the number to your business objective(s)?
- What do you hope to achieve with this number?
Don’t be stymied by the actual number and the presentation of it. Rather, it’s more important to understand all of these conditions and a way to track iterations of the data than it is to be fixated on the number itself. Taking this course of action will help you achieve more with your analytics.
Once you understand the lineage of data, assess the risk. The risk will vary based on whether you are massive company like Google or a mom-and-pop shop. So take an honest look at the metrics you are working toward and your criteria for risk.
Additionally, consider the risk of the specific action you plan to take with the data. For example, in the case above, if you discover 70 percent of your customers are wheat farmers you could take a low-risk action like proposing to dedicate eight hours to building a marketing campaign targeting this customer segment or a high-risk action like pivoting the entire company to focus on solutions for wheat farmers.
Ask yourself: Is this a behavior we afford to change, and can we change it with confidence? Know these answers before blindly taking a risk.
Discover the Power of Good Enough Analytics
We’ve seen more and more customers combat analysis paralysis by embracing good enough analytics. After all, companies invested in advanced analytics to drive innovation and make data-driven decisions. The problem is you can’t take action on those decisions if you are too caught up in overanalyzing.
Now is the time to make good enough analytics the new great.
Learn how you can join our contributor community.