My 3 hard earned lessons for analytics impact

The goal is action, not analyticsSam Daish, Xero

Most organisations are looking to become smarter or more data driven, and analytics or data science is presented as a crucial part of that. I agree wholeheartedly with that ambition, although I’d add that for Xero, our direct customer conversations can add richness and colour to data that is invaluable for us.

However this drive to be ‘smarter’ ignores or minimises an essential part of having an effective data science function.

It’s pretty well accepted today that an individual’s IQ is not enough for success. Without qualities like perseverance, interpersonal skills and drive, any smarts you have will fail to have much of an impact. I believe this is behind what many organisations find baffling about their data and analytics efforts - we have all this data, we’ve employed data scientists, we talk about being data drivenhellip; we even have lots of insight packs that tell us really interesting stuff - but we’re not getting the success we thought we would.

Now I don’t want to downplay the technical and capability challenges around data science, there are many and the science ‘half’ of data science can sometimes be treated too lightly. However I’d like to touch on three ways of approaching data science that I’ve found useful.

The goal is action, not analytics

I hear the phrase ‘actionable analytics’ often, and it bugs me. If this is your goal then you’re not actually aiming to change anything, you’re just making it someone else’s problem. Take responsibility for getting analytic results deployed in operations and/or in front of customers. Your job isn’t done until there has been an action that changes the organisation in some tangible way.

Part of a team with a shared goal

For the most part you’ll only be able to achieve actioned analytics if you are part of a team with capability and responsibility broader than analytics. Ideally this will be a project or virtual team focussed on a business outcome rather than functional delivery - i.e. a team tasked with reducing churn rather than a marketing or product development team. If those teams don’t exist, then you should go about creating them.

Minimal viable experiment

I only came across this phrase recently and it’s a great play on the more common minimal viable product. Ask yourself, what is the least effort real world test we could do to see if this analysis should really be actioned at scale. One reason I like this because it connects so well with the previous two points - it really pushes you to get action and to make it happen you’ll almost certainly need to work with others.

Analytics can be a catalyst for change in an organisation. Exploring data to create insights is interesting and usually fun. But to get to real value, aim for action and work alongside those who can achieve that.

Sam Daish is head of data innovation at Xero.

Related reading:

Follow CIO New Zealand on Twitter:@cio_nz

Sign up forCIO newsletters for regular updates on CIO news, views and events.

Join us on Facebook.

Copyright © 2017 IDG Communications, Inc.

7 secrets of successful remote IT teams