Becoming a data-driven organization has many pitfalls along the way. We provide the five proven steps your organization can take to expedite the process and encourage mass adoption.
Consensus can be hard to find in complex scenarios like large organizations, but the flywheel effect is extremely powerful when everyone is pushing in the same direction. IT departments know the challenge of consensus as they get pulled in a dozen different directions that are all the top priority of the requestor. We've found that the fastest way to consensus is to focus on already defined high-level goals that your organization has likely defined during its strategic planning process. If that doesn't work, you can start with common goals like increasing revenue or improving profitability to get the conversation out of the weeds. After defining the highest level goal, objective, or vision, the next question is "what drives this result?" For example, one of our customers has plenty of demand but limited supply. So, growing revenue isn't about more customer orders but rather how they can execute their sales and marketing strategy more efficiently. A manufacturer may want to improve profitability, but a quick chat with Finance may find that inventory costs are having the most significant negative impact on profitability. So, the short-term goal turns from an ambiguous "improve profitability" to "how can we improve inventory turnover?" The next question is, "what insights would help us improve inventory turnover?" This top-down approach will get your team to a consensus much faster.
You may notice that the examples above didn't include a group of techy people and data scientists isolating themselves to develop the solution for the organization to become data-driven. That process seems right because they understand technology, but it's much more efficient to understand and gain consensus around the business needs and let the technology organically meet the business need once it is clearly defined and the value is understood.
If there is anything you take away from this blog, let it be that the technology selection is the easiest part. Today, many organizations have easy, inexpensive access to business intelligence platforms like Power BI or Tableau, yet still, 73% of employees claim not to have actionable insights to make decisions. The survey data reaffirms what we know to be true: the biggest challenge in becoming a data-driven organization is centralizing your data into an easily accessible location.
What does "Easily Accessible" mean? Your employees should be able to log in to an easy-to-use platform with Google-like search capabilities and drag-and-drop report building that anyone can do themselves without advanced technical knowledge. Admin permissions should be easy to adjust, and employees should feel equipped to make quick data-driven decisions.
I say "Non-Negotiable" because this is the most commonly eliminated step when organizations seek to become more data-driven. It is eliminated because technologies have created workarounds that work early on, but problems with the workarounds are exposed in time. The data will eventually get corrupted, and the complex cleaning and modeling will break. It's not a matter of "if" but "when." The worst thing to happen to an organization seeking to become data-driven is when the users lose faith in the tool and data. This will have a long-term negative impact on your organization that you can't afford.
Your IT team is in a constant fight to protect your network. A study by Positive Technologies found that outside cybercriminals could penetrate 93% of organizations' networks, so security must be the top priority. Most IT departments also don't have data scientists, API-integration experts, or artificial intelligence experts, so adding business intelligence platforms to their plate is likely a bridge too far. The last thing you want your IT team to feel is like they are on an island by themselves, fighting off cybercriminals while you break down their walls with requests to use new outside tools. So, one of the most important steps is to ask your IT what security standards and data structure requirements are non-negotiable.
Most technology providers are prepared to meet the strictest security requirements you can throw at them, and there are dozens of ways to structure your data. Some organizations require no copying or moving of their data, so some technologies visualize your data so it can be used and seen but not moved or manipulated. I will warn you, meeting the most extreme data management requirements (Government, Finance, Health Care, etc.) will add cost, so setting and sticking to a reasonable standard will save you money. If you do this step right, everyone will feel good that we are solving valuable problems while maintaining a secure network.
There is so much ambiguity around technology, especially artificial intelligence, that it is almost impossible for most people to understand the differences between platforms. We suggest focusing on practical execution expectations instead of the core technology. Below are a few good questions to ask your team before looking at technologies:
"Do we want to hire full-time employee(s), a contractor, or do we want a DIY tool our employees can use after basic training?"
"Are we okay waiting for a new report to be created, or do we need live data look-up and report creation?"
"Who needs access to this, and what are the admin permission levels (edit vs. read-only)?"
"How are we measuring success?"
"Do we want all the general data insights they can give us, or do we want a specific answer/direction/decision?"
When you answer the above questions, it helps you define exactly what you need in a language everyone can understand. It will also weed out the vendors that don't have the features you need for successful implementation.
If you start by gaining consensus around your goals and solving valuable problems, you should be well-positioned and gaining momentum within the organization. This should result in what Kotter defined as "Enlist a Volunteer Army." Once again, it's not a team of people isolated in a room trying to solve the company's problems. Instead, keep it simple, keep it in the open, and empathize with people's needs. After all, people are happy to change when they feel heard, understand why, and know it will make their lives easier.
Now, it's finally time to evaluate technologies. If you're like me, you get excited about new technologies, but new innovative technologies might not be best for a first step. The right first step is likely a proven, known technology that you can trust. I say "technology" not "company" because many great young companies will do a 10x better job solving your problem. We like working with vendors that specialize in the specific application we have. Generalists are fine if you want general results, but specialists typically understand your application better and have designed a technology to maximize your ROI. Meet with at least three vendors and pick the one that meets all your non-negotiables and adds more value than expected.
If you would like a personalized plan for becoming a data-driven organization, reach out, and we will provide a free discovery process and project plan.
.svg)
.svg)
