Throughout 2018, we’ve chosen to feature a specific theme each month surrounding data science, machine learning, and artificial intelligence (AI). In June, we want to talk about becoming a data-driven organization — both what it takes and how to get there.
As it turns out, June is the perfect time to be talking about data-powered organizations. First, because it’s nearly halfway through 2018, and that means time to check in on those resolutions or year-long KPIs. If “be more data driven” was on your list, how is it going? If you’re anything like most businesses, it might not be going so well — it’s a lofty goal and one that takes specific dedication and a significant amount of organizational change to achieve.
Secondly, June is the month of the first EGG Conference of 2018, a day-long event in London devoted to the very question of building and moving toward a data-driven organization at scale. You can check out the killer lineup (including leaders in data from ING, ASOS, Uber, Daimler, and many more) as well as snag tickets here.
But back to business — if you’re nowhere near your data-driven goals for 2018, what can you do (aside from attend the EGG Conference)? Here are three concrete places to get started:
Choose an Initial Project
This may seem obvious, but many go wrong here by starting with planning, scoping, and theorizing instead of doing. But the best way to become more data-driven is to choose a project and dive in — use advanced analytics (predictive modeling or machine learning) to achieve a specific business goal, and see the project through to completion (read: production — more on that later).
Doing an initial proof of concept (POC) can open doors for other larger and more impactful projects once it’s successful. Once other people or teams in the business see a successful project, transformation will begin from the ground up as more and more want to start solving business questions of their own. It will also lead to more support and resources.
Have a Clear Path and Plan for Operationalization
The market is already shifting from BI to AI. That means that using data to gather internal insights retroactively is no longer enough — it’s now about real time and putting data projects into production.
So if you have a data team but they’re not able to work with data real time and are constantly working with stale data, start by addressing that. If they aren’t able to easily move out of a sandbox environment and put projects into the real world where they will have real business impact, make it happen.
This might mean investing in tools that help facilitate movement from a design to production environment. But it also is about smoothing out processes that act as blockers along this process (namely the work between data teams and IT teams, which can be so tricky to navigate that projects get stuck).
Empower People
Being a data-powered organization means that everyone — no matter what their role or team — should have appropriate access to the data they need to do their jobs and make decisions based on that data. Many of the world’s top enterprises are moving to a more self-service model when it comes to data access. This model empowers people to use data in new and creative ways.
Of course, it’s easier said than done. But a good start is to move to a center-of-excellence model where people from across the company can access data and then consult with data experts who can help ensure the data is used correctly as well as to apply more advanced techniques (such as machine learning) as well as, of course, put those models in production.
Ultimately, getting people to work together — those with the business expertise along with those that have data science and machine learning expertise, for example — is what makes the difference.
These three tips are just the beginning — stay tuned in to Twitter and LinkedIn to follow our conversations and get more insights this month on transforming into a data-driven organization. You can also check out what some of the top leaders in data have to say about this subject here.