Our Thinking
Getting Started with Big Data & Analytics (3/4)
31 July 2018 | Posted by Darragh McConville

Next up, we Enable broader access to what we have built. This post picks up where my previous Build post left off.

Enable

So, you’ve delivered your first use case successfully and other teams are now intrigued as to whether they could also achieve similar successes.  Enablement is intended to support and encourage this intrigue from other teams by enabling governed, self-service data discovery capabilities on your nascent platform.

6. Discover data & insights

Self-service data technologies are long-standing in our industry.  Over the past decade, they have matured and have furrowed their way down into the data platform technology stack, further empowering end-users.  This now means that the end-user discovery experience is richer than ever.

Forthcoming data discovery technologies will incorporate analytics capabilities and search-like functionality to provide so-called Smart Data Discovery – machine-generated insights from heterogeneous data sources supporting user-actioned, drill-through exploration. The purpose of discovery technology is to support the users that don’t know exactly what they’re looking for, until they’ve found it.

Pick a technology that empowers your end users and that they love to use.  Enabling ad-hoc analyses on full-fidelity data sources will further drive new use cases for your platform.

7. Govern data & access

With one central location for all of your internal and external data sources, there will be many voices expressing concern on how access to such a valuable asset will be governed.  A lack of governance is an all-too-common fail factor for data platforms.  Without governance, you’re going to stall sooner rather than later.

Your platform is designed to support the ingestion of any and all datasets, but I recommend militant metadata capture, tagging and classification as precursory steps to data ingestion and access.  This sets the bar from the very first use cases that there is a barrier to entry for data set ingestion and it’s needed to help you deliver success.

End-users must be able to explore and assess datasets without breaching authorisation privileges.  Leverage your corporate directory service to centrally manage user and groups and create roles aligned with those groups with orthogonal privileges – allowing broad and deep access to datasets.

You must strike a balance between good governance and constraining the flexibility and potential of the platform.  A healthy tension should exist between these competing capabilities, however in a self-service world, there will be no innovation without a little chaos.

No comments
Leave a Comment