Graham Sharpe, director strategic solutions at Exasol, unveils the five levels of analytics maturity and explains why it’s so important to identify where you sit on the continuum before you can move through its stages.
The last eighteen months have seen a major shift in perceptions of how important data is to all sorts of organizations. Now that data has become a recognised resource, a similar storyline is developing in analytics. This makes sense. After all, if data is the “what” then analytics is the “why”. Data is nothing without the analytics to create the insight and generate benefits.
What this means is that analytics is switching from a supporting function to a driving role. As such, organizations need to understand and define how mature or sophisticated their data analytics operations are.
Analytics maturity model defined
The simplest way to think of analytics maturity is as a measure of how well an organization uses its data. Those with a high maturity have data deeply ingrained in the organization and across all decision-making and departments.
The analytics maturity roadmap is something of an industry standard – a way of better understanding the different levels of maturity. The basic idea is that you’re on a journey over time, starting with a small startup/entrepreneurial approach. As you gradually add more governance, control capabilities, people and functionality, you’ll move up the maturity scale. The benefits of moving through the levels of maturity are high, but the road is full of potholes and detours. At every level there are common challenges.
What are the different stages in the analytics maturity model?
Although there are many ways the model is represented, there is usually consensus on five levels based around following:
- Level 0: Entrepreneurial analyst
At this level data usage is very spreadsheet-driven and IT independent. As such, some of the common dysfunctions include a lack of trust and governance, and a lot of subjective ‘truth’ guiding ad-hoc decisions.
- Level 1: Business governance and controls
Here we see the development of departmental ‘marts’ and dashboards based on static data. The dysfunctions at this stage include departmental silos, slow proliferation of tools and poor performance because of high latency.
- Level 2: State of the business
At this level we start to see the emergence of data warehousing and business intelligence (BI), and a more central organizational data and analytics capability. This is when IT capacity can become overwhelmed by the scope of what’s required. Data is still not a business priority and performance often limits adoption.
- Level 3: Agility to adapt
Here we see data becoming more of a strategic asset and the emergence of cloud economics. Low latency data pipelines and strong SLAs are driving new Loading...data science capabilities. However, there are still governance gaps and IT capacity is often still a struggle. AI/ML is only adding very limited value at the edges.
- Level 4: Business innovation driver
At the last level of maturity you start to bring Loading...data science to the data, rather than the data to Loading...data science. AI/ML are being applied at the operational core with query performance, concurrent usage and high availability key components. But it’s not nirvana. There are still dysfunctions to overcome in terms of AI/ML being seen as a poorly understood black box. And the challenges around explaining results hinder adoption.
Why identify which level you’re at?
The need to know your baseline analytics maturity comes back to the need to have a clear data analytics strategy. You can’t put together a proper strategy unless you know and understand what your maturity level is. If you don’t know where you are starting from then you can’t define where you want to go, or confirm when you’ve got there.
Your data analytics strategy will and should be multi-dimensional. A linear strategy that attempts to move from A to B will ultimately fail, as your organization will have multiple data-related priorities in various areas of the business. Instead, you’ll need to work on multiple fronts. There will always be things to be done immediately, such as a high-risk compliance problems that need solving. Then you’ll have some quick wins: probably tasks that are easy to pick up and use (these can help you the most in showing the data team is adding value). Meanwhile, your underlying long-term strategic target to build all your real competencies and grow your baseline maturity will be bubbling away in the background. At some point, those quick wins will disappear, the urgent fixes will be gone, and you’ll focus on the strategic baseline work 80% of the time. That’s the theory, anyway.
But, it’s complicated
It’s probably worth clarifying at this stage that it’s very rare for one data maturity level to apply across a whole business. Normally different parts of the business will start at different levels and probably have different aspirations. For example, maybe HR doesn’t want full AI/ML and just needs a couple of reports. But the marketing or finance teams have bigger aspirations.
It’s also important to understand that the journey through the maturity levels isn’t always straight. Sometimes you might even go backwards, for example when you have to re-platform, or when you first move to the cloud. You’ll also spend a lot of time going sideways and rebuilding to align with changes in your organisations enterprise strategy.
The good news is that regardless of your maturity level, your roadmap or the issues you’re facing, Exasol can help you grow and move through the levels – to ultimately help you achieve your goals via the quickest, straightest path possible. There are no barriers to entry. Exasol is a fast, flexible and proven platform used by some of the most ambitious businesses in the world. Working on-premises, in the cloud or in a hybrid role, Exasol provides the analytics solution that grows and adapts to the aspirations and direction of any organisaton.
The truth is that everyone would like to move up to the highest maturity levels, but in most cases, they don’t have the tools, the people, or culture in place to make it happen. Although there are a lot of ways to become a data mature business, it’s not always a linear path and it’s not going to happen overnight. With each level of analytics maturity reached, you’ll need to support data democratization by enabling more people to not only access data but empowering them to do something smart with it.