It’s human nature. Individuals believe that data must be perfect in order to start making decisions. To get around this deeply engraved notion, business needs to shift focus from “I need the data to be 100% accurate if you want me to tell you xyz” to “my tolerance to data accuracy is x%. As long as I can see trends from the data that’s provided, I can start making informed decisions.”

This is not related to blatantly wrong data. If your company is doing a $100 million in revenue per year, but your BI tool is reporting $10 million, that data might almost be worthless. Even then, there are probably still lessons you can take away.

This is also not related to financial reporting, where data does need to be accurate. This article is for decision makers that are convinced that decisions cannot be made without accurate data.

Good Data Doesn’t Guarantee Good Decisions

For the sake of argument, lets say that the data is 100% accurate. What is the point of that data? To start making business decisions off of it. If the decision maker is inexperienced, perfect data won’t make a difference. They won’t have enough experience to know how to comb through the data.

Data is only useful if it helps you make good decisions or decisions faster. What’s more important is that the data is good enough and easily accessible. If you’re pulling data from various systems and it takes you days to compile data to answer your BI needs, how much quality deficit are you willing to tolerate if that data is easily accessible (2 minutes vs 2 days)?

Do you have an analytics expert for your department? Frequently organizations expect IT to extract data and make sense of it. Businesses put too much faith in IT to get them the exact information that they need. IT introduces tools for the organization to use, but that doesn’t mean that IT now needs to use those tools. It’s something new that decision makers need to use themselves. Imagine requesting for Excel to be installed and then demanding that IT creates all of your spreadsheets.

Understanding data is a skill that needs to be learned. I use the example of Calculus. You can memorize all of the formulas, but unless you have experience using those formulas and know how to apply them to a problem, they’re worthless. Companies should focus on creating workshops with existing experts or outside consultants. Workshops get you to a point, but ongoing coaching is more effective in building knowledge.

Looking for Trends

Most times, correct data won’t change the trends observed, it just changes the magnitude of the results. If a customer purchased $100,000 worth of product last year, but your BI tool is reporting $90,000, how likely is it that $10,000 will sway the trend and your decision?

It’s always great to do sanity checks. Pull data from a different source that you trust. Compare the data from your BI tool. Correct the discrepancies but analyze the incorrect information. Does your sales or marketing objective change now that you have correct data? More often than not, no.

Visualize the meaning behind the data. Isn’t this the primary objective? To visualize the meaning of the data in order to make strategic decisions. A simple approach is to graph the data. Add trend-lines to see where the data is going.

With enough data coaching, where you state what the expected results achieved were for a specific segment of data, the organization can even start incorporating predictive analytics. Predictive analytics allow you to see likely future events. It can be scary how accurate these models can get.

Summary

When you start to stress that data isn’t complete or 100% accurate, ask yourself, “what am I trying to achieve with this data?” If the answer to that question is related to strategic decision making, you most likely have enough information already.