How to Begin an Analytics Transformation in Your Organization, Part 1 - The Stages of Transformation
Updated: Aug 15
As a consultant, I often find myself in a room with clients that tell me they are ready to use “advanced analytics” to realize their organizational goals. Often, they are already into the fiscal year, and something like “leverage analytics capabilities to drive revenue growth” was added to the list of goals for the year with no real idea if that was possible. I tend to think of this stage through the lens of what I have come to see as the standard growth strategy for any organization seeking to effectively leverage their data. The graphic below is one that I provide to clients during that discussion.
The first thing that I want to make clear about this strategy is that, ideally, it does NOT start with the entire organization. In a company of any considerable size, there will be departments with wildly different concerns and workflows. Rolling out an analytics product to all of them at once is a complex task that will require considerable effort to gather requirements, plan, develop the product and test. I always encourage my clients to embrace Agile, start small (adoption), and incrementally enhance (evolution).
In the Adoption phase, the best way to ensure success is to locate a group who will be more receptive to a new way of doing things. Your initial partners will require existing data to gather useful insights and, most importantly, the initial set of users should have a sufficient amount of credibility within the organization to ensure that their advocacy results in support amongst their peers in other parts of the company. A reliable way to identify such a group is often as easy as finding a team that uses cumbersome analytics processes by necessity. In such cases the team members in question have two jobs: creating the reporting to help them function and successfully executing the operational tasks they were hired to do. In those cases, I find it useful to describe the process using the four icons listed under the adoption phase.
Take some time to find out what reports the team is creating, who is doing it and how long the tasks take every day/week/month quarter. Find the easiest reports to duplicate that will have the highest impact and engage in a freeform brain dump with the relevant team members. The key in this step is to engage your knowledge resources and assure them that your goal is to automate this process in order to allow them to execute their primary duties as much and as efficiently as possible.
The highest performing architects I have ever known all had a great talent for leveraging and empowering the experts on the team. Those experts are not only a wealth of knowledge, they also have the credibility needed to persuade the rest of the team of your solution’s potential. During this stage it may be necessary to make some compromises. As much as its pained me, there have been situations where I simply replicated a tabular Excel report in Power BI to get those experts on board and to start creating a situation where the source of truth is the new analytics product. If possible, it is always good to agree upon the addition of a companion visual that clearly shows how efficient the analysis could be without Excel as our primary source, but the main goal of this step is to generate some consensus and ensure engagement. So, regardless of how effective you think the format to be, it is important to remember your main goals and see that they are prioritized.
Regardless of who your new users are, they will need training. This should consist of one or more actual training sessions followed by working sessions that allow the team to learn by doing, which usually proves to be the most effective means by which to transfer knowledge. Your objective in this stage is to nurture a sense of expertise among those people who will be your test case, your first users and, if all goes to plan, the advocates disseminating information about the benefits of your new system to the rest of the organization.
With your requirements gathered, your users trained, and a rapport built with your pilot group, the final product that THEY helped design is the creation of a limited reporting suite. This should consist of automatically generated versions of the reports you are replacing, with some upgrades to leverage the power of visualizations (if possible). It is important to note that the reports should also be highly performant, stable, and should be updated as much or more than they were created manually. This is the first impression that your product gets with this team of respected experts, and you only get one shot at a first impression. That means your focus should be on usability, accuracy, and availability.
Assuming this process has been executed successfully and lessons were learned, you should close out this first set of features with the knowledge necessary to replicate the process throughout the organization within an Agile framework. It’s critical to focus on replacing manual processes, increasing accuracy, and providing users with the tools they need to do their job better. Eventually, this stage of growth in analytics should be complete and your client, hungry for more efficiencies and greater insight into their business, will seek to enhance the existing set of tools.
Stay tuned for part 2 of this series, during which the evolution stage will be our focus.