The decision to make your company more data-centric is definitely a big one. Whether you operate in insurance, banking, retail, or almost any other industry, your company is almost certain to benefit from decision-making based on evidence and hard facts, rather than relying on the whims of fate.
If you’re looking for a way to establish your company’s data foundation, you’ve come to the right place. Here’s how you can get your company to love data and more importantly, how to benefit from it.
BE CAREFUL WITH WHAT YOU MEASURE
There’s an old management adage that goes, “If you can’t count it, you can’t manage it.” This saying supposedly underscores the importance of having data at hand as a management tool for performance. Though there is still some truth to this saying, it should probably come with a caveat: be selective about what you count. In theory, almost anything could be included in the data pool you’ll be building for your company, but every good manager knows that not all inputs are equally valuable.
Given this, take the time to figure out what really matters to your organization, then find a way to measure that and include it in your data pool. Certain business functions will be more important and valuable to your operation than others, so identify how to quantify those and include them in your data pool. You can leave out the less important items for another time.
Another oft-overlooked benefit of being selective about what you decide to count is that you prevent your team from feeling like they’re being over-managed. All employees should understand that some performance metrics will be necessary, but at the same time, they wouldn’t want to be nickel-and-dimed at every turn by their employer.
KEEP CLEAR OF DATA SILOS (ESPECIALLY WITH YOUR DATA TEAM)
Once you’ve decided what performance metrics you want to include in your dataset, the next step is to assemble your data team and put them to work. There are many ways to do this, and how you execute your data operations within your company will depend on your specific requirements. That being said, one thing that must be avoided at all costs is accidentally walling your data team into their own silo.
To expect that your data set will be perfect from the get-go would be considered unreasonable by most experts. In order to arrive at a complete and accurate picture of your operations, the information you choose to include in your data foundation will have to evolve over time. By placing your data scientists in a silo, you prevent them from gaining on-ground insight into the raw data that they gather.
At the same time, insulating your data team from operations will also hamper your growth as a data-driven company. If the way reports are assembled is not transparent, employee trust in them will likely become quite low.
MAKE YOUR DATA SELECTIVELY AVAILABLE
Different departments will need varying levels of access, and as such, the data available for their perusal should be filtered accordingly. For example, members of a bank’s marketing team will need access to their company’s most important investment products. However, they will probably not need the background check results and information that account officers will use to vet a potential client. Deploying sufficient access controls in this way ensures that team members are presented with information that is relevant to their function, without inundating them with a deluge of data.
Obviously, this bit of advice does not mean you should simply give just anyone access to mission-critical data. Standard security measures still need to be implemented to ensure the data foundation’s integrity. However, once the basic protocols and correct access settings are in place, you also have to ensure that their access is unimpeded so that they can do their jobs properly.
GIVE ERRORS A VALUE
To expect that your dataset will be flawless and will lead to perfect decision-making every time is simply unreasonable and unrealistic. Your data should be the foundation of your company’s decisions, not its end-all-be-all. Over time, certain business functions might become more valuable while others may end up obsolete. Having to restructure your dataset around constantly changing variables will compromise the quality of your data and, by association, the quality of your decisions as well.
At the same time, there’s bound to be a level of uncertainty present in your data. Where this uncertainty comes from differs depending on your operation and industry. A country’s political climate, macroeconomic or monetary policy, and even the weather are all factors that are difficult to pin down. However, any or all of them are likely to impact a company’s quarterly forecasts, no matter how solidly supported by data those forecasts happen to be.
Smart companies know that this is simply the price of doing business, and have learned to quantify the values of these errors so that they are budgeted and accounted for. Being proactive and attempting to see shortcomings in one’s own dataset results in better bottom-line reports and fewer surprises when it’s time for the annual review.
Establishing a data-loving culture within an organization starts at the very top, so company leadership must advocate for it. Actively demonstrating why establishing this kind of culture is in everyone’s best interest will go a long way towards winning buy-in from team members and eliciting the change that you want for your organization.