Data warehousing is the least valuable thing for your data

A data warehouse is a terrible place for your data. It goes against everything we have heard for the last two decades, but it is true that warehousing is the least valuable thing you can do with your data. Information Theory proves this contrarian view and, along with Lean practices, suggests a better way to leverage your institution’s data. That said, it is important to note just how important data is to the future of your credit union. Data can make the difference between your credit union thriving and better understanding your members and being one of the financial institutions that close their door or are merged every single day. So, let us dig in on what Information Theory and Lean practices tells us, and more importantly how you can do better than warehousing when it comes to your data assets.

View your data warehouse as an actual warehouse. It is helpful to understand why it is the least valuable thing you can do. In this virtual warehouse sits hundreds of thousands of square feet of stuff in boxes. Stuff just sitting there, doing nothing. Anyone trained in Lean knows that this is a recipe for disaster. Lean tells us that we only produce as much as we can reasonably expect to use in a short duration. Holding as little in inventory as possible reduces costs, defects, and waste. Defects is an important aspect that we will revisit later. So why would we treat data any differently than any other inventory?

Some might argue that a data warehouse is different than a cavernous physical warehouse filled with boxes. But is it really? Lean tells us that everything should be contributing directly to the value chain. So, what is value of the terabytes of data that is either wholly or mostly going unused? The answer at many credit unions: nothing. That is of course, unless your credit union is currently using that data to tangibly improve the member experience, reduce costs, prevent fraud, or improve the financial lives of your membership. However, most data warehouses are not doing this, they are simply storing that data – metaphorical boxes sitting in a warehouse generating cost, introducing risk, and creating little to no value.

The data warehouse metaphor gets even more interesting when viewed through the lens of Information Theory. Firstly, we will step back and briefly talk about Information Theory. Imagine you are in a completely dark room you have never been in before. You want to leave the room, but you do not know how close you are to the door or even what might be in the room. Information Theory says the information that is valuable is only what tells you how to leave the room. Getting an updated baseball score while in the room is not valuable because it does not help you leave the dark room.

Thinking back to the data warehouse, where all that member data is stored in metaphorical boxes, but not being used violates Information Theory. Data collecting digital dust in a warehouse violates the theory because it is not being used to inform decision makers in the credit union. It is not informing how to leave the dark room or in our case, improving the credit union’s situation. For data to be information, it must add to knowledge. Data at rest does not do that, it is just a cost.

It is interesting how Information Theory and Lean both agree here. Product, or in our case data, not directly in the member value stream is waste. For most financial institutions data is being somewhat used, at a clip of 5 to 10 percent, which means that 90-95 percent of the data warehouse is producing little to no value. When a credit union is spending hundreds of thousands of dollars a month on servers, salaries, and bandwidth, this becomes very expensive waste.

Taking a moment to get perspective about how data can produce value at your credit union is the first step. Using the analogy of the dark room with your membership is helpful. What does the senior leadership team need to know to make strategic decisions in the interest of the membership and the future of the credit union? What does the digital banking team need to know to make product and offering decisions to improve the member experience? What does the branch team need to know to improve the branch experience or increase product adoption? Asking these and other questions are critical to begin to formulate what data is valuable and what is not.

Once these questions are answered, the input data has been inventoried, and the systems identified that deliver and create the information, a credit union is ready to approach data. Not as a warehouse full of metaphorical boxes, rather as a value-producing asset – this is what data should be. Data that contributes to answering these questions gets used and is no longer wasteful. Further, because the data is being used, quality issues with data get quickly noticed and the trust in the internal teams goes up. Once everyone is pulling from a single source of truth the internal data culture also begins to grow as does your data’s value. Data increases the efficiency, accuracy, and precision of a team’s decision-making process as a whole – both on the larger, strategic scale and in solving day-to-day operational problems.

Once an institution has focused its effort on only investing in the data that is usable, both in terms of content and age, the institution can focus on getting out of the dark room. Once the volume of data is manageable, instead of storing the data in sealed boxes in a warehouse, a credit union can spread the data out onto an active workbench. A workbench that can be used to craft insight, prove hypotheses, and generate tangible results from an otherwise intangible asset. Data work-benching, instead of warehousing, also implies a level of transparency, accessibility, and supports people frequently revisiting the workbench with new questions and ideas. This level of data interaction is not obtainable with a data warehouse.

Having a data warehouse is not a bad thing, but a credit union needs to get value out of it. Credit unions that view their data warehouse as an asset to help them navigate the analogous dark room are getting value out of it and likely improving the lives of their members. However, if there is no clear vision or understanding of how it contributes to servicing the membership, it is most likely nonvalue-added cost. Information Theory helps leadership determine what in their data warehouse helps them make better decisions and what is only incurring cost. Put your data to work in service of your members, they will thank you for it.

Co-authored by T. Buck Strasser