Garbage In, Garbage Out - The Importance of Data and Storytelling

Like Frank Sinatra famously sang about love and marriage, the same can be said about data and storytelling -- you can't have one without the other. Or at least you shouldn't. In 2018, we have gotten really good at collecting data, analyzing data, and automating reporting off of that data. Technology has enabled us to get the big picture like never before, and have insights literally at our fingertips. But what about when the data is incomplete? What if the data is inaccurate? What if there is more to the story than the cold numbers being rolled up the chain to management? You might think that in this day and age we have found a way to totally automate this roll-up, totally account for any discrepancies in the data, and tap into enough data sources where we can automatically get the core of the truth. You would be wrong. We have not found a way to truly replace the human element of telling the story of what the data actually means, nor is it clear that we should want to (for more on this topic, click here). The largest, most successful companies of our time rely heavily on the anecdotal stories from their customers and employees to shape their decisions, in addition to cold hard data. In fact, Jeff Bezos (Amazon) says that he has " noticed when the anecdotes and the metrics disagree, the anecdotes are usually right. That's why it's so important to check that data with your intuition and instincts, and you need to teach that to executives and junior executives."

This brings me to the topic of garbage. All too many times I've worked with companies that will create a system that is set up to fail before it begins. They will create a seemingly endless array of KPIs, all of which (or most of which) require a human to fat finger data into a system. They will say, "ok, now we've created the 100 fields we need to collect to gauge this project's success. Now if only those project managers would enter the data and keep it up to date. This typically leads to PMs burning daylight trying to keep up with the requested data and heated phone calls from leadership when things are not up to date (which is probably daily). In Texas, that's called a big ol' goat rodeo. This is true of Project Management, Sales, Customer Success, you name it. There are always infinite possibilities of what data you could collect. And yes, if your employees always entered every one of those 100 bits of data you could probably get a pretty accurate picture of what is going on. But this is far from reality. Back in the real world, it requires a significant time commitment, and a significant amount of discipline to maintain the system. If a project manager invests in the discipline piece, they can find themselves spending so much time updating the data that they've ceased managing projects. If they are so heavily invested in delivering actual results that the time commitment to maintaining the system slumps, it can be seen as a lack of discipline. The problem is the economics of time and the reality of human imperfection. If the quality of your inputs is seriously lacking, it follows that the quality of your outputs will also be seriously lacking. What's worse, is that it can cause you to zoom in and focus on the wrong thing, and totally miss the big picture. It's like looking at a couple of green dots and making a decision on what those mean when you are actually staring at a Georges Seurat painting. If (like me) you couldn't remember that he was the guy who painted this, go ahead and ask Google.

If your data inputs are garbage, there is minimal value in generating reports off of them. Don't sweat the small stuff, focus on what you need in order to see the big picture, and make it easy for employees to collect that information and roll it up.

So, how do we correct this? Well, you could get strict about enforcing discipline. I've actually heard the words "any rep who doesn't keep their deals updated in Salesforce will be fired" on more than one occasion. Not exactly the way to a happy and motivated team though. Now, I'm certainly not arguing against discipline - on the contrary, personal discipline is (in my opinion) one of the key defining factors of success. But it's important to understand that management's lack of clean data may not be due to a lack of discipline, as discussed in the project manager example above. The solution lies in making it easier to get data in the system. The first piece of making it easier is to get the system working for you, instead of just you working in the system. Having the system automatically pulling data from different sources, and rolling it up into reports certainly solves some of that problem. It reduces the margin of error when employees no longer have to enter the same data into multiple different systems (spreadsheets, CRM, Project Management tool, etc.) and swivel between them. The other piece is simplifying the information you are looking to collect. Management should take a hard look at the information they are asking for, and ask the question is this truly important? In general, we've found that focusing on a few key bits of manually inputted information (while automating the rest if possible) and additionally focusing on the project manager's story around the project health leads to better results. There's a degree of trust in this equation - you have to trust that your employees are still doing the smaller less critical items, even if it's not being entered in a system. In this situation, employees are happy because they are not spinning cycles entering data instead of executing, and management is happy because employees are getting them the data that actually matters. It's easier to enforce discipline on a few major items, than across a million tiny details.

The final piece to this is the storytelling around the data. The Project Manager or employee will always have access to data sources that haven't been plugged into the system. The system doesn't contain every detail of their real-world knowledge, even if they are keeping all 100 KPIs up to date. The employee might know something the system doesn't - for example based on the data in Salesforce we are projecting to hit our company sales target this month, but the sales rep knows from talking to customers that they are getting hit by the blizzard of the century in Minnesota, and no one can get to the office to send a purchase order (for more on visibility across Sales, Marketing, Customer Success click here). So it is critical that we incorporate the stories of our boots on the ground, and their individual know-how into company decisions.

If your data inputs are garbage, there is minimal value in generating reports off of them. Don't sweat the small stuff, focus on what you need in order to see the big picture, and make it easy for employees to collect that information and roll it up. It doesn't have to be rocket science (unless you're a rocket scientist, and your project involves sending something into space) so don't make data collection more complicated than it has to be.