Understanding your analytics maturity: The first step to success
Let’s say you landed a great job as a data scientist at a big company — and you cannot wait to put your skills to the test. Your mission? Building one of the most accurate and mathematically intricate machine learning algorithms in existence. No one has done this yet – you could be the very first. You have the solution drawn up in your mind and can’t wait to get started. All you need is the green light. Then, the age-old question: Just because you can, does it mean you should? If you haven’t considered your analytics maturity, the answer may well be “no.”
What is analytics maturity?
The analytics maturity model is a hierarchical measure of how evolved your organization’s analytics capabilities are. Some companies have a suite of Power BI dashboards, while others rely solely on Excel to carry out their daily reporting. Misjudging this can have serious consequences on whether or not your solution delivers long-term value.
There are many different interpretations of the analytics maturity model, but the following levels of maturity as defined by Altexsoft are widely recognized:
Why is maturity important?
Suppose you have developed a sophisticated data-hungry Random Forest machine learning solution that you believe can deliver value to the business.
If your organization’s maturity is a 2 (descriptive analytics) on the scale above, the solution will be counterproductive. While it may provide an initial “wow factor” when demonstrated to your stakeholders, that excitement will soon fade when attempting to integrate the model into processes insufficiently mature to support it.
Without the requisite skills, training or infrastructure to fully leverage the model, it will soon become clear that the effort to create it was wasted and similar initiatives may meet the same fate. On the other side of the coin, providing colorful bar charts and visualizations (a staple of level 2) to leaders of an organization already working with predictive analytics (level 4) will hardly add any value either. In analytics, the key to success is to take the time to evaluate where your organization sits on the maturity ladder. Only then can you develop solutions that enable the company to move up to the next level – without skipping any steps.
In analytics, the key to success is to take the time to evaluate where your organization sits on the maturity ladder. Only then can you develop solutions that enable the company to move up to the next level – without skipping any steps.
For example, if you have an on-premises hosted SQL database of product transactions, you may find value in exploring machine learning to determine trends in item sales. However, your infrastructure may need some upgrades before you can host a computationally demanding deep learning solution.
It’s imperative that any data scientist takes the time to review the strengths, weaknesses and analytics resources available within the organization before proposing a solution.