One thing we've noticed from working with clients at different stages of their digital transformation is that they're unaware of where they are against where they are trying to get. And sometimes what's even possible with data.
The four phases of data maturity
We've put together our view of a data maturity model to outline:
- The four distinct phases of data maturity
- How to characterise them
- The questions you might be asking
- The goal you're trying to achieve at each of them
While the key areas we'll be focusing on within these phases are:
- Technology
- People
- Data Quality, Management, and Governance
Technology relates to how advanced you are at using different tools, services and technology to manage the data in your business.
People focuses on what your internal capabilities/skillsets are like and how both management and colleagues are adopting data within decision-making and processes
Data quality, management, and governance are relatively self-explanatory. Do you trust the data? How are you managing it? Does it conform to GDPR compliance? And what sort of control do you have in place to make sure you can continue to trust in it?
Phase 1 - Foundational Phase
Let's consider this the starting point. At worst, you're here, but the good news is that you're also at the point where the smallest changes can result in the most significant gains.
You're probably still heavily reliant on Excel for reporting and data manipulation. Everybody has a different number for "how much did we sell last month?" and there's no real process behind the production of reports. You probably have no idea whether the reports you're producing are even being read, let alone used to make decisions.
Internal expertise is limited. You probably have the "data guru" who gets inundated with Excel-based requests, and if you ask a question about some sales figures, you might get passed on to a specific person because "they have the answer". If this is the case, then it also means that you have many single-points-of-failure, i.e. the individuals have all of the knowledge.
In times like these, where staff are on furlough and people are being stretched beyond their means, relying on individuals for critical information is a difficult situation to manage. And you should avoid it at all costs.
You have a mixture of internal systems, 3rd party suppliers, and external data sources producing information and trying to aggregate them is an almost impossible task. If you can do it, it takes somebody a significant amount of time each month to produce them. And given how manual the process is, you have to question the accuracy.
A data discovery workshop can help you catalogue and organise your data, processes, and systems.
If you say NO to any of these questions, you're likely in this stage:
- Is my data accurate?
- Do I have access to all of my data?
- Do I know when data is missing?
- Do I know when data is erroneous?
Phase 2 - Reporting Phase
At this stage, you're moving in the direction and using data to understand what has happened. You're moving towards using modern technology to handle the automation of data processing and producing dashboards that offer high-level reporting to different departments in this business.
You might have a small internal team that has helped put this in place, but their experience is limited to doing what's required to get the usual reporting completed.
The good news is that people recognise this work as useful, and department heads are using it to steer their actions for the next period. This consistency and value in the data is sparking more interest, and people are now trying to use data to assist with other projects and increasing the appetite for more business intelligence overall.
Now that you're managing data centrally (maybe in the form of a data warehouse), it's much more reliable, and people trust what the figures say. There may be some governance in place for data sources based on who the internal experts are for each data set.
Questions that you're focusing on most in this phase are:
- Can I spot trends in my data?
- Which channel caused a spike in my web traffic?
- Where did the increase/decrease in revenue come from?
- Where are there growing opportunities/sales in my business?
- How did my campaigns perform?
- Where do we need to focus attention next week/month?
- Who are my most valuable customers?
Phase 3 - Predictive Phase
The focus is no longer on looking at the past, but rather understanding how the future looks and what you can do to affect it.
Your technology stack is growing. You've got the infrastructure to process data at scale and reliable storage across a data lake or enterprise-level data warehouse. You have a business-wide solution for visualising the data and each team has 90% of what they need to get the job done. Now it's time to tackle the more challenging questions.
Your internal team has expanded. You're now an in-house analytics centre-of-excellence, spanning multiple disciplines across data engineering, visualisation, analytics, and data science. By working together to solve challenges across the business, most departments have become self-sufficient from the dashboards, analysis, and recommendations you provide.
The next significant opportunity is to work out how to optimise further with predictive analytics. Can you identify the best segment to target in your marketing for a specific product? Or work out the best new subscription for an existing customer as they're about to renew? These are the components that will give you the edge over your competitors.
Senior leadership are now seeing enormous returns on investments made into the analytical engine of the business and are working from the top to have the data-driven culture embedded company-wide.
With an enterprise-level data ecosystem in place, the governance around every data source is managed with a high level of scrutiny before it even makes its way into the system. This ensures high-quality data that everyone trusts.
Questions you can answer in this phase:
- How likely is this person to convert?
- How much will I sell next month?
- What will be my best-performing products?
- Who should my future customers be?
- Why are my customers churning?
Phase 4 - Embedded Intelligence Phase
The final phase of data maturity is transitioning from being predictive on an ad-hoc, project-by-project basis, to embedding the intelligence within your business systems or processes.
Examples of projects you might undertake at this phase are:
- Deploying a churn model within your customer relationship management (CRM) platform so you can quickly identify which customers need attention.
- Enhancing your website's product search functionality to identify the specific customer that's browsing and having recommended products for individuals targeted in email campaigns.
At this stage, you have the nuts and bolts in place from the previous phase. The main addition to your tech stack comes from the technology you use to build and (most importantly) deploy machine learning models.
With the entire business now being entirely data-driven, the need for more data (and insights) to be available faster to more people means that your team has scaled. You now have teams for each data discipline. You've also brought onboard machine learning engineers to build and deploy production-ready models to use at scale in your systems or products, and day-to-day operations.
The skills around data have improved significantly across the business as other functions are recruiting data experts directly within their teams. For example, marketing may have a digital marketing analyst who's an expert on how to use data to make strategic marketing decisions.
Data is now paramount in how the company plans to evolve and is at the forefront of the company vision and project roadmaps.
In the final phase, governance around data is controlled heavily by being integrated into business processes. A company manual is available that outlines all data in the business, along with relevant data owners who ensure that this is kept up-to-date and in line with legislation.
If you know how to answer these questions, then you can consider yourself at the summit of data maturity:
- How can we make sure we optimise product recommendations for individuals in our app?
- Can we understand, link, and react automatically to each customer's individual needs?
- Can we embed our churn model into our chatbot so that it can respond accordingly?
- Can we include our propensity model in our marketing so that we have the best ROI (targeting those most likely to buy a particular product)?
Keep your eyes on the prize
Each phase within the curve is wide. There's a lot of business change needed to move through them and it can take substantial investment and time to do so.
The key point to mention is that the technology and tools you're using are only weapons in your arsenal towards delivering the most value to your business. They don't create business value alone. You need the blend of people, knowledge, and process to extract value from your data. And oftentimes, this requires the guidance of powerful data visualisation tools and specialist data analytics experts. 
