Every Org needs a Director of Analytics and Insights
If your team members spend even 20-mins per day collating and reporting data and you have 50 or more people in your org, then you should hire a Director of Analytics and Insights.
Data is everywhere and there have been amazing technological advancements in the ability to capture various signals throughout the life cycle of a business, from Marketing to Sales to Customer Support, Operations, Finance and so on. But the way we draw insights from these expanding data sets and convert them into actions has by and large remained stagnant. Most of the journey from Data to Insights to Actions is heavily weighted on gathering and sharing “what” happened — in the last quarter/year/time-period — and most teams stop at ensuring that the facts are accurate, available and accessible. A majority of the cumulative time spent across the orgs on analytics related activities is still in the context of technology rather than business outcomes.
The need to know the facts is generally widespread and in response some orgs invest in hiring talented analysts directly on the functional teams (like Marketing team or Sales team etc) to harmonize data from various sources and build informative charts and reports. If you have a mature data warehousing team that maintains a unified, modeled and discoverable data foundation, then the analysts work off of that data foundation. This is a necessary step in the right direction but is usually not sufficient for the org to operate in the higher tiers of the Analytics Competency Model (Fig. 1) because these individual analysts are likely fully occupied just recording and reporting the facts about the “what”.
By hiring a Director of Analytics and Insights and staffing that team with the right number of analysts, every org can provide the necessary support and institutionalize the commitment to reaching the top half of the Analytics Competency Model (Fig. 1). Such a team along with the Director can ferry the org across the competency chasm and elevate the org to a state where a majority of the time is spent on quantifying “why” the outcomes happened and deciding “how” to amplify the good outcomes and remedy the bad ones in future time-periods.
This analysis has to be led by someone that’s organizationally in an influential position, to counteract the possibility of confirmation bias and to increase the likelihood that your org will always think about the “why” and the “how”. With the term “Director” the emphasis is for them to be a senior and influential designation in your company. In some industries this may be a Senior Manager and in other industries it may be a Vice President or SVP. Unless of course your company is truly rank agnostic. If you have a “… best idea wins and source doesn’t matter…” culture in practice, then the designation is not as important. Either way, the important thing is to staff these embedded, functional analytics and insights teams in a manner that makes them organizationally influential (business and technical acumen are minimum requirements and assumed).
The role of the data warehousing team becomes even more important in a federated ecosystem (Fig. 2) like the one that starts to emerge when there are multiple functional teams for analytics in a company. The data warehousing team needs to be at the center of this federated ecosystem (Fig. 2) to orchestrate activities such as data governance and data cataloging. They need to own and maintain a common set of technologies that all teams use for storage, transformation and visualization. For the data warehouse team to play the role of a conductor in this analytics-symphony, they need to follow regimented processes to mitigate development risk and employ data-mesh and data-fabric type technical architectures as they build out and evolve the unified data foundation.
A big advantage of this federated ecosystem (Fig. 2) is that while the data warehousing team institutes processes to manage overall risk, it gives functional teams the liberty to work off the data foundation to run experiments, iterate quickly and fail fast. This flexibility is required for an org to get to and continue to operate in the top half of the Analytics Competency Model (Fig. 1). Once there is feedback and validation that a set of metrics are relevant for company-wide use, the new metrics can be made widely available by the data warehousing team by adding them to the unified data foundation (implementation of the unified data foundation may be conceptually one-foundation but realized via architectures like the data-mesh).
The absence of such functional analytics teams forces other team members in the org to become those data analysts which takes time away from their core jobs and competencies. So, when an org decides to invest in such analytics teams, the return on that investment (see Break-even analysis Fig.3 and Fig. 4) should be assessed in terms of the time savings from team members in other roles who no longer need to train to be data ninjas. All artifacts pertaining to tiers 1 and 2 of the Analytics Competency Model (Fig. 1) should be automated and just be available so that the rest of the org can work on applying the insights, implementing data-informed actions and realizing the desired business outcomes.
To all the direct and indirect teachers in my life. Also to Chris Chan and Nick Kogelman from Cisco for sharing their experiences being on functional analytics teams.