“Data” is one of those words that strikes fear into the hearts of many. Often, people have a visceral reaction to the word–they liken it to “math” and opt to ignore it by pretending it’s not there. This is a huge mistake though, as proper analysis drives optimal decision making. Ignoring data is like piloting a plane in the dark–if you aren’t watching your controls and dashboard, how do you know where you’re headed?Ignoring data is like piloting a plane in the dark–if you aren’t watching your controls and dashboard, how do you know where you’re headed? Click To Tweet
So, if you’re running a business without direction, or if you’d like a refresher course on data analysis, this one’s for you.
Steps for Analyzing Business Data:
1. Define Your Objective
To begin, you need to define the focus of your objective. What exactly is the problem that you are trying to solve and what type of data can you gather to help you address it?
For example, imagine you had a subscription model business where your customers received a series of packages from you; one new package every month. You might decide to analyze what stage in the process your customers were ‘quitting’ the subscription. In order to maximize your profits, it’s in your best interest to identify the stage where your customers are leaving so that you can fix the issue and keep them happy.
2. Determine the Data Required to Measure the Problem
Now that you’ve identified your problem and objective, you need to determine what you will measure to help you improve. In keeping with our previous example, you might decide that to determine the stage where your customers are no longer renewing, you would need to track the order history of a customer over time. The most pertinent data points to collect would be (1) customer name, (2) customer first order date, and (3) customer last order date. Form this, you could determine how many months (on average) your customers stuck around.
To add additional color to your analysis, you might also consider including (4) the reason for non-renewal and (5) any comments/objections made by the customer. Generally speaking, it’s easier to provide analysis around quantifiable data. As this fifth data point isn’t easily quantifiable, you might consider ignoring it. However, as a rule of thumb, it’s always better to collect more data than not enough.
3. Analyze the Data as You Collect It
In many cases, you will have the opportunity to analyze data that you’ve already been collecting for some time. Data collection generally happens as a matter of necessity in today’s world. For example, if you have a website, process payments online, or advertise, you already have a cache of data waiting to be tapped.
Generally, you can begin analyzing this data immediately. Keeping with our example, if we took orders for our product on the internet, we’d have the majority of the data we need waiting for us already.
Side note: this is called “dark data” –any data that you’ve collected, but that isn’t being used to drive any particular initiatives or decisions.
So, let’s imagine that we already have our (1) customer name, (2) first order date, and (3) last order date. Most platforms would allow us to export this data to Excel, or we might elect to use a more sophisticated tool for this analysis. Regardless, the function is identical–we need to take the raw data and paint a picture that tells us when our customers quit our subscription. Ideally, I’d hope to see this both as an average and also represented graphically–imagine a bar graph, showing cancellation months across, by quantity vertically.
Graphs are often the quickest way to identify outliers. For example, imagine a graph like this:
In this example, we can quickly see that non-renewals are happening at a higher rate in the 5th month. This is driven from the following data set:
However, an important thing to ask yourself when analyzing your data is whether you’ve painted a complete picture, or if you are limiting the information because you didn’t dig deep enough. In this example, I would hope to see a more robust analysis, including the percentage of customers leaving in a given month.
In data analysis–percentages are always your friend!
So, including the number of remaining subscriptions allows you to determine the percentage of your customers that fail to renew at any stage–a much more valuable metric!
In this case, we clearly see that the 5th month has the highest cancellation percentage (4.06%). However, something that wasn’t apparent in our graph is that the 8th month also shows a high cancellation percentage, relative to the other months. It would be advisable for us to explore both of these months to determine how we can improve the customer experience.
4. Track Data Over Time
After our initial analysis, we will want to make tweaks to improve (we hope) our business. But don’t think your analysis is complete! In order to know if the changes you are making are, in fact, improving your business, you will need to continue to monitor these metrics. Ideally, you’d like to track them on a consistent timeframe, so that you can avoid hurdles and verify that your improvements are performing as anticipated.
For example, if the cancellation percentages we calculated above were for January, we’d want to calculate the cancellations for February as well, and then compare the two months. We would present that data in a format like this:
“M/M” stands for “Month over Month” –we are simply taking the difference between these two time periods. (Alternatively, we could take a percentage change.)
Examining the above data, we would want to observe two key issues:
- The 5th month saw an increase in cancellations–whatever we did to “fix” this didn’t solve the problem.
- All other months saw a M/M change of under +/- 0.20%. This suggests that the other months are relatively consistent, which lends more weight behind our first observation.
In short, the key takeaway for data analysis is this: something is better than nothing. As Peter Drucker famously said, “If you can’t measure it, you can’t improve it.” There has never been a time in history where tracking data is as automated as it is today–don’t let a fear of the unknown keep you from growing a more successful business.
Start a 14 day Free Trial and streamline your business with PaySimple:
Start My Free Trial