Relationships – we’re constantly looking for them whether consciously or subliminally (many of which we may go on to regret). Sometimes we don’t see the relationship right in front of our eyes, and other times, we see relationships that just aren’t there. So it makes sense that in world where everything that surrounds us is data, our instinct is to connect pieces and seek relationships. For example, a company does a big online promotion to boost e-commerce sales. That week, they get new customers. They walk away thinking ‘this worked! We should allocate a larger portion of our marketing budget to e-commerce.” Well, maybe. The win here is that the business now has credible data on online purchases and can capitalize on the insights hidden within this data. But without a solid approach to analyzing that data, the company can make misguided business decisions that can have serious negative effects.

Harvard Law student, Tyler Vigen, hilariously illustrates this point on his website Spurious Correlation. A person looking at the data in the graph below would likely conclude that keeping the age of Miss America (how is that still a thing) under 21 means fewer murders by steam (why is that even a thing).

Spurious Relationships Research Strategy
At Compass(x), we think that getting the right answers has to do with asking the right questions and thinking things all the way through before you start. Here are the four things we think about before fielding research or beginning data analysis.

Building a Data Analysis Plan to Make Better Decisions
1. Know why you are asking
It’s likely that you have a number of unanswered questions that could benefit from further research. So, start with identifying a specific problem – what are you trying to solve for and why do you need to know this? Defining the issue to solve or the question to answer will minimize the time you and your team spend wandering around, lost in the woods.

2. Develop your hypotheses
In order to tackle the bigger issue you’re trying to solve for, you have to break it down into smaller pieces that can be tested – your hypotheses. These are the statements that will either be validated or disproved through your research and will guide the specific questions that are asked.

3. Determine how you will measure your hypotheses
For each of your stated objectives and/or hypothesis, identify the variables that need to be collected. For example, if you want to know if your new content plan is generating better quality leads, have you defined what “better quality leads” means? How will you know if you are getting them?

4. Develop an analysis plan
The most exciting part of research is analyzing your data (we think)! It involves thoughtfully unpacking data in order to find meaningful patterns. Now that you know what you are trying to learn, and how you intend to measure it, you are ready to think through all the ways that you are going to look at the data. If you want to understand how average transaction value changes with frequency of trips, think through how you want to look at it – a line graph that will plot the two against each other? Using an effective analysis tool allows you to visually engage with your data, makes it digestible and allows you to tell a story.

Choosing the appropriate analysis framework also involves deciding on a measurement method that you will later cite as as evidence. As Vigen’s graph illustrated above showed us, even commonly used metrics like correlation are not always appropriate.

Once you have the components identified above, your Data Analysis Plan should lead your team throughout the research process. From using it to delegate responsibilities amongst your team to writing your research questions, integrating the Data Analysis Plan into your project ensures your asking the right questions and getting data that is actionable.

While the urge to look for relationships will still exist, you now have an objective analysis plan that supports your research goals and gets you closer to solving for your business issues. Without a clear approach to fielding research and analyzing data, you risk living in fear that the world will be steamed to death by a pool of aging Miss America contestants. And no one wants to imagine that.



Sehare is a Strategist at Compass(x) Strategy, a Chicago brand and growth strategy firm that creates sustainable growth for passionate companies.

Pin It on Pinterest