Blog post

The End of the D&A Center of Excellence?

By Andrew White | June 12, 2022 | 0 Comments

DecentralizationData and Analytics OrganizationData and Analytics LeaderChief Data and Analytics OfficerCentralization

News last week: Meta shunts their AI hubs into their business product Units. That’s the message reported in the papers Tuesday.  See Wall Street Journal and Meta shakes up AI unit Amid drive for faster growth.

The Trouble with Organizations

The problem or trigger is not unknown: a centralized team tends to specialize in skills or capability but it’s often remote from the business roles who need that capability. The functions are centralized and operate as a shared service to business functions or units.  This creates one of the oldest customer-service arguments; does the centralized service meet the needs of the distributed ‘customer’?  It’s an organizational gulf that is hard to cross.

In the past a customer mentality, not unlike “real” customers outside your organization, has been used to help drive success.  This worked in only a few places.  The application of lean, agile and DevOps has been used recently, with mixed results. Not least because such practices evolved for other challenges.

The Rise and Fall and Afterward

Over the years we have seen the rise and fall of centralized teams. Rather than rise and fall, it’s more like a seven year itch. Organizational structures tend to vacillate every few years. It would seem that we are in a new cycle where the focus is remote, distributed capabilities rather than fatter, centralized structures.

The case reported in the WSJ article suggests that time to value is hard to reduce in centralized teams.  The inability to organize the central resource with the remote “customer” needs is the great challenge. Shifting skills to the edge and away from the center should put the capabilities in more direct control of business or “customer” needs.  The result will be that some capabilities will need to be duplicated across business functions or units.  This will likely increase costs and duplicate investments. As such the key here is not really where and when to centralize or distribute. The real challenge here is coordination. And that is the real battle field. Meta May experience shorter time to value in their next cycle. But their costs will increase.

Rather than assume your organizational decisions are resolved by shunting the team skills to the edge, real success over time will be in how and who coordinates all the work. This is where your Chief Data and Analytics Officer (CDAO) role is key. The right personality, the right skills, will orchestrate and progress value delivery better than any one dedicated organizational approach. So don’t fret the Meta change: focus on the hinge or fulcrum that connects all the piece parts.

For some related research: Where to Best Organize Data and Analytics.

Comments are closed