Data is everywhere. Every decision we make takes in whatever data is available – including our own perceptions and preferences. Your alarm goes off in the morning and you look at the clock to decide if you are late, early, on time – or to hit the snooze button. And so the data analysis begins.
That’s my morning anyway. The funny thing is that when I think of that daily pattern of mine, I realize that I frequently start my day with doubt of the data. “Is my alarm working right? It can’t really be time to get up, can it?” Then there’s the bathroom scale. Clearly that’s not right.
Doubting data seems to be an ever-growing phenomenon. It seems like most of the headlines over the last couple of weeks were all about data and trust. Thank you Facebook and Cambridge Analytica.
But trust it or not – we need data. More importantly – we need the information that data creates.
Over the last few weeks, I’ve been showing you my data geek side with thoughts on the differences between data and information, the importance of a business case and approaches to improving data quality. Today, I want to look at data from the Use side.
Build Usable Information
Data alone is an element of information and not necessarily useful by itself. For example, say I had a list of ages. That list tells you nothing by itself (other than I seem to have too much time on my hands). But if I add to that list that those ages came from people who live on my street, then add education levels, income levels, activity levels and so on – the result is the beginning of demographic information for my street.
The key is to ensure that the data that you’re assembling makes sense for the questions that you’re trying to answer. This is where your Business Plan comes back in. What are those questions? Maybe you are trying to determine if adding a function or feature to your product improved sales. Or maybe you want to know if a change in a process improved X. Whatever it is, make sure the scenario is defined clearly so you get all of the data that you need.
I recently participated in an interesting webinar put on by the IBM Analytics team that talked about the importance of smarter analytics and data. Their guest, Joel Shapiro, had some interesting insights and anecdotes about what he’s come across over the years.
Mr. Shapiro gave an example regarding a trucking company that overhauled their internal GPS tool so all drivers would be using the same tool (instead of some using their cell phones). The company made a significant investment in the tool, deployed it and gathered six months of data before beginning to analyze the results.
At first they were shocked to see that it looked like accidents increased under the new tool. They pulled it from the field right away and then continued to dig into why there were more accidents. Long story short – they didn’t have an accurate comparison in their data – so their results were skewed.
The new data presented on all drivers – previous data presented on drivers using GPS tools (not all used a tool). So while it looked like there were more accidents, the accidents actually went down when you looked at the overall pool before the new GPS and the overall pool after deployment.
Moral of the story – make sure you have all of the relevant data elements that you need to answer the question posed.
If people are going to trust the information that you’re presenting, they (and you) need to trust that the data it’s built on is solid. That’s not just ensuring that the data elements themselves are true – it also means making sure you have the right mix of data. Just ask that trucking company’s GPS team.
What’s also critical is how you present the information. Even if the data elements are correct on their own – and you have the right mix of data – the presentation may not give a true picture of what’s really going on.
It doesn’t matter how you’re presenting the data (dashboard, report, presentation, etc.). What does matter is that it presents a clear message. People will take action based on information – and they will do so quickly.
The trucking company pulled the GPS tool right away – which makes sense. Why would they risk more accidents if the tool is a contributing factor? But the reality was that impression from the data was wrong – it wasn’t the GPS tool. So there may have been accidents that resulted from pulling the tool.
What’s more – information that is presented in a dashboard or a report is often viewed quickly – and without a narrative around it. So the viewer or reader is taking whatever lists/graphs/charts and processing that information with their own personal filters of understanding and experience and then making a decision on what that information means.
Think about it. How many times have you read a headline and started to form an opinion about a situation? Then, when you read the article and dug into it more, your opinion changed. We’re human. We are constantly assessing and processing data – and frequently acting on it (without more digging).
The good news – you can always update and improve datasets, dashboards and reports. The bad news – you now have more to overcome in getting people to trust that the information that you’re presenting is accurate.
Moral of the story – make sure that the information that you present is painting an accurate picture (and as unambiguous as possible)
Lather, Rinse, Repeat
Data and information quality takes time, but it is SO worth the effort. Quality information can improve efficiencies, customer experiences, budget investments and so much more.
My advice? Continue to work on the quality of your data elements, the assembly of data to usable information for your business plan and the clarity of how you convey that information (and then tweak/adjust/continue). If you do that, you will be well on your way to building trust in your data and information. Not to mention reaping the benefits that good data offers.
Anne Hale is the Director of Client Services at HL Group, Inc., a premier provider of mobile inventory management and warehouse solutions. She manages our client engagements, works with Wes Haubein on sales and marketing and is unusually preoccupied with good data.