The analytics and business intelligence spaces are frequently poorly executed spaces. After more than a decade operating in this space, I believe I’m beginning to understand why. With a background in software engineering, and a dozen engagements in the financial, manufacturing, logistics, and resource sectors, the narrative has been strikingly familiar. This is an attempt to better define this space, such that organizations can more effectively place employees, and employees can find satisfactory cultures to work in.
Should you be working for the business or IT?
The first step to understanding the opportunity is to understand what part of the business to role reports into, and the consensus of the organization itself. Typically, BI reports into IT with a dotted line to the business. This may have been the best choice in the past when technical competency was nascent in the business. That’s changed, as evidenced by the growing BOYD trend, and new techno-savvy generations in the workplace.
It’s probable that there’s some level of mistrust between the business and IT. If there’s already a level of mistrust there, then you’re already starting at a deficit on either side of the fence. When you’re engaged with the business directly, there’s the opportunity to observe and infer what the leaders in the business are really seeking. If you’re sitting with the business, it’s likely you’ll be part of the discussion on how and why they are trying to consume the information, and not just what they are trying to consume. If you’re part of the how and why discussion, you’ll quickly be able to see what will be asked, and you can be prepared.
In my professional experience, successful analytics and BI practices are more easily driven from the business. This makes it easier to establish trust, and allows the right questions to be asked.
Does the culture support your initiative?
Top down hierarchical organizations tend to have to force engagement. Information is only useful if it drives change, and leaders of the organization have to be engaging if they want their workforce to embrace change. I’ve worked places where the latest and greatest tool will be our salvation, and we will finally be able to operate more effectively. Truth is, the technology has always been there. The barriers to entry have long been political, organizational, and egotistical. Information easily empowers people, so they hold on to it. If you really wanted to engage employees, why would you choose to spoon feed them information? Part of engagement is autonomy. Few people want to do a bad job, but most people don’t have what they need to do a good job. I suppose it’s easier to blame these failings on a tool rather than an individual.
If this space were to be executed more successfully, a good analyst would have some understanding of behavioral science, performance coaching, and business sense. When providing information to your consumers, you have to be able to articulate why it matters to them (WIFIM). Great organizations will attach their metrics to strategic goals, and will move together. It’s not following the leader, it’s emergent behavior. Understanding the organizational cohesiveness is paramount to where the analytics team should place its effort. If you’re trying to deliver information, and no one agrees on it, you’ll have to manufacture consent. In order to do that, you’ll have to seek consent from leadership, and then find stakeholders where ever possible.
If you are embedded in the business, you have more reach and insight into possible constituent stakeholders. It’s best to be engaged directly with the leadership in the business, so you can better understand and cater to the dynamics of the organization. You can provide information and the correct analysis to foster emergent behavior.
Do they want you to do analytics, or build reports?
I’ve come to realize that there are two types of individuals in the BI space, those who do analytics, and those who build reports and claim to do analytics. The latter is a class of individuals who do what the business asks of them, even if it is the wrong task. A true analytical mind will challenge the question. The report builder may build something that answers a question, but it’s not an answer that will add to the bottom line. The difference is that one understands the business, and can marry that understanding to the data, and the other sees only the data and draws only the immediate conclusion from the information. In my experience, the report builder will stop at this point, with the immediate conclusion. The business leader may not have the time or the energy to take it further, but the analyst will understand that it is the journey, and not the destination.
Most organizations require both. A true analyst will throw spaghetti against the wall and see what sticks, creating technical debt along the way. The report builder should be charged with paying off that same technical debt such that the flow of information is sustainable and will scale.
Is your group composed of the right individuals?
The genesis of BI & Analytics comes from IT, and a prevalent mindset seems to arise from the application development world. I’ve worked in groups that will try to test reports with business users, when the data is out of date. It’s fantastic that the report works, but testing the most current information is more relevant. The best BI environments I’ve worked in have had a copy of production information that is kept current, such that timely information can be tested. If you’ve come from app dev, you’re used to testing functionality (smoke, regression, and unit tests). If you’re really doing analytics, you’re testing the data. That’s part of the journey the analyst should be taking with people in the business. Testing helps build consent, as you’re likely to be getting people in the business involved.
You’re testing information, not functionality.
Does the organization have a holistic approach to BI & Analytics?
At it’s core, analytics serves as a feedback mechanism to the greater business. Every organization has many broken processes, it’s very easy to be sucked into fixing something that adds very little to the bottom line. There will be a few issues that have little cost to fix, that can add a great deal of information. This is one of the reasons I believe the analytics team should be embedded in the business, as they will have more sway to fix broken processes to improve the feedback loop.