Businesses can’t afford to wait weeks or months for the centralized engineering team to approve, scope, build, test, and deploy requested changes. The opportunity will evaporate before the necessary data or analytics make it back to them. To the extent an analytic model is disconnected from the evolving business context, it becomes little more than an intellectual exercise.
Even so, the work must get done. In a top-down engineering environment, analysts are often forced to go it alone. This is the reason Excel is the most popular business tool on the planet. It’s also the reason companies are plagued with orphan analytics. When analysts work in isolation, you end up with disconnected, inefficient, inconsistent data silos representing hundreds of versions of the truth.
For all the drawbacks of centralized analytics, this approach was formerly the only option. The old world of expensive, fixed server architecture demanded it.
Not anymore. The old requirements for pre-defined server capacity have given way to dynamic cloud-based scaling. Highly elastic data platforms can be scaled up or down within minutes or seconds. The capabilities and economics of capacity-on-demand have changed the game.
The technology has evolved. The analytics function must also evolve to meet the needs of the business and realize the benefits of this new data mesh architecture.
The future of analytics is decentralized, empowered… and collaborative.
An organization’s analytics program must keep up so business teams can respond quickly to new data challenges in order to drive outcomes.
Collaborative analytics recognizes that the experts on the business are those sitting on the front lines, and as such, the overall structure of an organization’s analytics must be organic to take advantage of this knowledge and evolving business needs.
Collaborative Analytics brings engineers and analysts together to deliver outcomes and increase competitive advantage. In this new world, successful analytics organizations focus on outcomes, then organize & empower cross-functional teams around those outcomes with as few dependencies outside their team as possible.
It’s telling that Amazon, one of the most successful analytics companies in the world, has no central analytics team. Instead, engineers are embedded alongside analysts within autonomous business teams organized around specific business objectives.
This is a vast improvement from the “collaboration” which often occurs in top-down analytics organizations. An analytics leader at a Healthcare company recently lamented that because inefficient queries are such a drag on their data warehouse, a group of his DBA’s are forced to spend a better part of their day killing bad queries. Analyst writes a query; DBA kills it. Not a great working relationship!
By embedding data engineers within business teams, you get the best of both worlds. Engineers and analysts do their best work together as they figure out solutions collaboratively. One is an expert on the technology, the other an expert on the business context. An empowered analytics culture pushes responsibility to empowered, autonomous teams who have outcome objectives. Business objectives are set centrally, but the decision process becomes local to the autonomous team.
Centralized analytics continues to exist under this new model, but as a support function providing resources, capabilities, training, and tools to these business teams. The role of the data architect shifts to that of reviewing and curating the best ideas from the edges of the business, and then cataloging those as standards for reuse by all.
Collaborative analytics promotes efficiency and consistency while enabling the business to make confident decisions for the best outcomes.
At Aginity, we work to make this process seamless. See for yourself by downloading your free trial of Aginity Pro, our highly functional SQL Management software for individual analysts, and a steppingstone to Aginity Premium, our collaborative analytics platform.