The classic male dilemma: ask for directions or press on knowing you are, in all likelihood, lost. Though men probably get lost a lot more often than they should, a few careful glances at a map could probably spare them a lot of trouble. Similarly, when you’re looking at an analytics roadmap, there is a wealth of information that can keep your organization from wandering in the proverbial desert.
Not everyone is sitting in the same room in today’s strategic meetings, but digital transformation can’t stop. If you’re sitting at home, and all the stakeholders are meeting with you remotely, an agenda is never more important in guiding the conversation where it needs to go. You’ve gotten decision-makers to agree to hear out the plan and vision for a digital transformation, AI/ML, or an analytics project…now what? Here are the essentials for your agenda.
Define analytics use cases
There are generally three types of approaches for building an analytics roadmap. The first is the use-case driven approach. This is when the stakeholders have specific problems/opportunities and seek data and analysis for addressing them. The second is to develop a larger analytics framework for an entire department. For instance, if the department was Marketing, you would have to consider all the data you need to track and analyze for optimizing all marketing decisions across the board. Third, is to build and implement an entire analytics strategy built around your organization’s strategic priorities: everything you do is linked back to the company objectives and how data can be used in all decision-making processes.
Who you bring into a specific project will depend on the approach. The most important first step, however, must be a connection to the business strategy. What are the most important goals or objectives for the organization and how can data support that? It can be limited in scope like a use-case driven approach or it can be organization-wide analytics transformation, but the objectives need to be clearly defined or you risk a poorly defined roadmap.
Assign clear ownership
Once you understand the scope of the roadmap you would like to develop, you need to understand who will be taking ownership of the project beyond the initial planning phase. For instance, in the early days, ERP implementations used to be run by the IT department and not everyone was involved resulting in poor ownership and adoption. The same thing happens with analytics initiatives today; not everyone in the organization is using or caring about these solutions. That’s why you need to expressly seek out the support of all the stakeholders that will be impacted by the initiatives, not only from the business, but also from the IT, data and analytics organizations. This goes beyond just the CIO; ask yourself who owns these initiatives at a granular level, so that we can get everyone to be a user, producer, or analyzer of the data.
Define key metrics and decision processes for success
What are the key decisions you are trying to make and how will data and analytics impact the outcome metrics? Is it reducing customer churn, increasing marketing return on investment, driving product revenue, etc.? Once you know the decisions and metrics, you need to understand what current processes are already in place to make these decisions. Are there any gaps/issues in the process, from a metrics definition, decision-making or communications perspective, and how can they be addressed? Sometimes you have bad data and it leads to bad decisions; sometimes you have good data, but your models don’t help you extrapolate that into better decisions. You need to spend serious time evaluating your metrics and decision-making processes to ensure that the conclusions that you are reaching are the right ones.
Plan for analytics gaps and data visualizations
Of course, as businesses increasingly use sophisticated analytics, there’s going to be a gap between an organization’s ability to come up with high quality analyses and its ability to apply them effectively to business decision-making. There is a need to get the analyses in front of the right people and deploy the models into production. This is where you need to make your dashboards and models flexible enough to address the changing business circumstances.
Beyond just having the right models, you’ll need high quality visualization to help convey the analyses in the most compelling way. It should be able to speak and provide insights without needing a team to explain the intricacies. Finally, use this visualization to disseminate the data directly to the decision-makers. Information is that much more powerful when it is in the hands of those who are making critical decisions. If you want your analytics roadmap to be successful, you’ll need to show real analytics results to the people at the top or you might risk being ignored.
Building the roadmap
All the stakeholders in the meeting need to understand these issues, contribute to a root-cause analysis of current gaps and provide insights that can help address the gaps. Take large issues and break them down into smaller constituent parts, so that you can start understanding how to obtain the data and do the analysis required for solving them. If the data required is already being collected in the organization, how will you get it where it needs to go?
Having an agenda for tapping the right approach, stakeholders, metrics, and visualizations while addressing the analytics gaps as they emerge is key for building an analytics roadmap. Ultimately, once you have developed the map, be sure to consult it regularly to ensure you’re staying on the right path as you move from point A to point B in the most efficient way possible.