Data-driven business and data-driven culture emerge in many organisations at the moment. The amount of data increases enormously and at the same time, data processing methods and technologies are significantly developing. Companies that are able to collect and combine data from external and internal sources and enrich it to new meaningful information are the future winners. Are you dreaming of or planning the roll-out of a data-driven culture in your organisation, but don’t know where to start?
This series of blogs focuses on the six phases that can help your organisation’s culture to become a data-driven one. This post focuses on the first of the six phases, which is recognising your analytics maturity level.
What is the Acquired Potential of Your Analytics?
The maturity level of analytics describes an organisation’s ability to make use of analytics in its operations. The first steps of analytics are familiar support functions, like Excel reports and occasional customer segmenting experiments that are used to try different kinds of new campaigns and operations.
At its most mature level, analytics has gone from being time-to-time trials to becoming a competence that guides your business. This means your company will provide a faster, more personal service to your customers with the help of real-time analytics and products that are tailored to the customer’s needs.
Make Use of Existing Analytics Maturity Models
When mapping your current level of analytics, it is easiest to start by relying on an existing model. Mirror the model to your operation, and you will have a tool that is easy to start with. Repeat the exercise a few times and you can modify or even change the model.
There are various analytics maturity models. At its simplest, the maturity level can be condensed into a single number like in Forbes’ 2012 example with game producer Zynga and vending machine company Redbox, but for this exercise, the intent is to divide our services a little bit more finely.
Gartner divides analytics maturity levels into four different classes, from least to most mature:
- Descriptive: What just happened?
Report that describes history events. Often a numerical report vs. traditional Business Intelligence.
- Diagnostic: Why did something happen?
What was the reason for an event? Interprets and explains past events and phenomena.
- Predictive: What will happen?
Predicts future events based on expertise and/or historical data.
- Prescriptive: What must be done for this to happen?
Attempts to influence the future by modifying it into a more desirable one.
The higher your analytics maturity level is, the closer to making and implementing a decision you can get without human effort.
At its most mature level, analytics is automation – when the information required for the decision-making process is fully known, human input can be disregarded completely.
Tips For Identifying Your Analytics’ Level of Maturity
- Start small and be less self-critical. Don’t worry if you can’t think of anything. Every company has analytics, insofar as statutory finance and work time reporting is concerned.
- Analytics doesn’t have to be performed by a machine. Traditionally, financial reports were prepared by the financial department for the benefit of management.
- Divide the analytics into parts if it seems too difficult. If you find it difficult to get going, make a list of services you work on and consider the level their operation is measured on.
- It doesn’t matter if you cannot find the right category. Every operation will not simply fit into a single analytics category, as the example below shows.
- No company or service will simply fall into a single level of analytics maturity. If your list seems to be on a too high level for maturity classification, divide it more finely into core functions. That will help you move ahead.
Maturity Level Mapping – Example Case: Smart Insurance
Smart Insurance is a new, imaginary insurance provider. It provides smart, scalable-pricing insurance policies based on the usage time and location of the insured device. Smart Insurance has decided to map its own level of analytics maturity and has, for the sake of the exercise, listed a few of its core analytics-based operations.
Listing analytics services and determining the maturity level:
- Level: Descriptive
- Statutory financial statement figures are calculated by hand at the financial department. This is a traditional report on historical details, the development of which is followed by management.
Customer churn forecast
- Level: Diagnostic
- Customer churn is tracked by attaching an outbound customer’s last actions as supplementary information to their survey. This information is used to make an automatic assessment of how the customer relationship ended.
New insurance policy sales forecast
- Level: Predictive
- Smart Insurance has an algorithm that predicts, based on internal and external sources, the amount of new insurance policies sold. Making use of this figure is the responsibility of company management.
- Level: Pre-emptively descriptive
- Clear cases
- Level: Decision automation
- Compensations are decided on automatically, when the claim includes the issues required for classification and does not contain factors that are marked as risks for automatic classification.
- Unclear cases
- Level: Decision-making support
- An algorithm generates a suggested solution to a claim that can be approved or rejected by a human operator after reading the claim and the machine-generated automatic classification risk factors.
Automatic insurance pricing
- Level: Descriptive and Decision automaton – No simple fit to one category
- Insurance prices are generated automatically based on the device’s location and time history. Despite insurance pricing making use of advanced analytics, it does not neatly fit into any of this maturity model’s classifications
After the maturity level has been determined, the next phase is to identify existing successes, weaknesses and future opportunities. I will focus on these in my next article with tangible hints. Start thinking about your organisation’s analytics maturity level and contact us, if you want to get a head-start in its development.
Olli Lindroos works for Solita’s Agile Data team. He is a passionate learner and proponent of data-driven culture, with an interest in technology in all of its forms, whether it’s about the user experience, technical implementation or business strategy opportunities. Olli describes himself as a dad and a nerd, as well as a food and drink aficionado. In his free time, Olli likes to try out all the newest trends as a consumer and building the IoT equipment he needs by himself. Olli is on Twitter as @ollilind.