Using the Right Data: 7 steps for improving your data driven decision making
Making data driven decisions is all the rave these days, especially in operationally focused fields such as sales strategy, sales operations, and performance marketing. Nearly any job posting in these fields will demand some type of experience with this skill set. In this article, we’ll review some simple steps you can take to raise the bar on your data driven decision making game.
A common data driven decision making process might look something like this:
- Identify the scope of the situation
- Understand the desired outcome
- Understand the variables influencing the situation
- Measure the variables using available data
- Interpret the data
- Make a decision
This may look straight forward, but the truth is that it can get tricky along the way. Even the best of us get sidetracked and derailed. Some of the most common pitfalls in this process include:
#3: Understand the variables influencing the situation
When starting out with a fresh problem and evaluating metrics, more is better. It’s easy to get led down the garden path of “magic KPIs”. Gold-standard KPIs are a good thing, but don’t forget to gather as many metrics as possible when beginning your data driven journey. You can later group these metrics together, identify influential hierarchies, and then eliminate for significance.
#4: Measure the variables using available data
This can quickly turn into your worst nightmare of a black hole time suck. It’s easy to get caught in the vortex of “but the data is dirty!” or “but we don’t measure that!” or “our sample size isn’t big enough to show significance”. After analyzing countless data sets, I can tell you with 100% confidence that everyone faces the same challenge: there is no such thing as perfect data. Make note of deficiencies along the way and prioritize improvements into your backlog.
#5: Interpret the data
Cole Nussbaumer Knaflic does an excellent job providing clear direction for data driven professionals in his book, Storytelling with Data, describing the important distinction between exploratory and explanatory analysis. As analysts, we might geek out and want our audience to enjoy every awesome thing we found in our exploratory journey. Using our explanatory chops, though, we should select only a few most valuable points our audience needs. To help guide this process, I ask myself, “If this data could tell me only three things in concise sentences, what would they be?”
To help bolster your data driven decision process, I recommend using the following 7 steps to save time, money, and headaches.
1. Use what you have
“Take what is offered, and that must sometimes be enough.” – Altered Carbon. Most definitely an awesome show, but for the context of this data driven decisions, it’s a good rule to live by. Stop wasting time fretting about what data you don’t have or how dirty it is. Accept what you are given and build from there.
2. Find out what other people know
One positive externality of an increased focus on data driven decisions across the globe is that benchmarking and comparable analyses are relatively easy to find. Look for similar businesses with public financial statements, research papers published on Harvard Business Review or other reputable sites and make note of anything you find that might be useful when applied to your data set and situation.
3. Remember what you already know
You didn’t get as far as you did in your career by accident. Chances are you’ve got a pretty good head on your shoulders, s don’t take that for granted. Forget about the data for a minute and ask yourself, “What do I know?” What do you know about your business? About the world around you? How might these things influence the data you’re analyzing? Write down your observations to all these questions, clearly marking them as such assumptions. These assumptions will be imprecise, but the directional accuracy they provide is invaluable.
Whip out your Excel chops, start with a blank worksheet, and start punching in the metrics and assumptions you’ve collected. Identify everything that’s an input along the way. You’ll start to be able to see holes in the story your data tells. From there, start making calculations using multiple data inputs to extrapolate the ones you don’t have.
Turn your three concise sentences the data would tell you into a paragraph. Tell a very short story that looks at all the information you’ve collected. Remember that stories have a beginning, middle, and end. What was the situation? What did you find along the way? Which way does the data point?
6. Topsy turvy that shit
There’s a great scene in Curve your Enthusiasm where Larry David is giving advice to Leon about how to take advantage of an interview by turning the tables. Similarly, take your data, your assumptions, and your story, and flip it. Spend some deep thought looking for any other combination of stories or explanations you could tell that might fit. During this process you’ll be able to identify potential weak points, assumptions you should change, or KPIs you might omit, preparing you to be mindfully cautious when moving on to the next step.
7. Make a decision
You’ve front loaded the work and now your mission is to pick a direction and stick with it. Remember that when making data driven decisions there no single right answer. Most often direction accuracy with high confidence level is the best we can expect. Ask yourself “Is this the right direction?” If it is, the data has gotten you this far. Now it’s time to go with your gut and stick with it.