The Analytics Maturity Model Spotlight: Your Guide to Success
08/16/2019 by Chris St. Jeor Modernization - Analytics
08/16/2019 by Chris St. Jeor Modernization - Analytics
Survey after survey shows that companies are struggling to utilize data and analytics to transform their organization. Notably, a 2018 Gartner survey revealed that almost 87% of companies believe their analytics strategy did not provide optimized business outcomes, and that 91% have not reached “transformational” level of analytics maturity despite it being a top priority for CIOs.
McKinsey found in a 2019 survey that “constructing a strategy to pursue data and analytics” was the top reason, chosen by 21% of responders, why organizations were effective in analytics, and also the top challenge, chosen by 24%, for organizations that were ineffective in analytics.
Even when embracing data and analytics, a large portion of senior marketers reported to Gartner in 2020 that analytics have not had the expected influence, largely because “analysis does not present a clear recommendation.”
The hype is not matching reality.
As with any endeavor, companies must have a planned strategy to achieve analytical goals. Once the plan is in place, an organization will need support from leadership and follow through across the organization to optimize the impact of their new data-driven strategy. The analytics maturity model (AMM) can help with building a plan and also identify actionable steps for improving analytics outcomes.
The analytics maturity model (AMM) has its roots in the software capability maturity model (CMM). Both models describe the stages a company travels through to reach process maturity in analytics and software, respectively.
The following model shows how businesses evolve from chaos to an optimized data-driven approach.
To discuss how businesses can move up the maturity model, we will present the 5 levels of the analytics maturity model through the lens of a sales organization. As needs become more analytically driven, businesses have the opportunity to evolve their understanding and analysis of the sales process. Each level – from chaos to optimized outcomes – will be presented below.
Let’s talk about each step in detail:
In Level 1, the typical company is dependent on spreadsheets and ad-hoc analysis. Everything is a one-off, so rarely do tasks go beyond one person doing work in a single spreadsheet.
The best view the company can expect to have of their sales figures is a rolled-up monthly view. When things get really fancy, the analyst might even add a time series chart showing the historical view of monthly sales.
Typically, reports are produced on the individual level and exist on individual machines. There is no central data mart or sharing of code. Equations exist within the cells of excel spreadsheets with no repeatable process or peer evaluation. Usually, this information is more of a process artifact than something someone planned. There is undoubtedly no analytical strategy at this point.
As the business reaches new success, the process knowledge becomes unwieldy. Managers and other leaders fail to see this situation as a real or solvable issue. They are focused on survival.
Because of the siloed equations and lack of coordination, this situation often leads to spreading incorrect information, extensive comparisons of differences in reports, and lengthy inquiries seeking clarity.
The company cannot function for long without better planning and preparation. They realize they need to develop a broader view of the business to survive. They must begin a transition away from this kind of myopic task management for reporting and begin putting processes in place.
In Level 2, a company’s actions move from ad hoc spreadsheet reporting to descriptive analyses of the past. Managers capture a broader view of their processes and better understand the big picture. A more data-driven approach allows managers to streamline the business and cut out unnecessary costs from the sales process.
To simplify work, departments adopt applications to record daily work activities and aggregate views of cost centers.
Each department selects the tool that best meets its needs. Most folks likely continue to use low-level analytics tools such as Microsoft Excel. In most cases, these applications produce reports for daily activities. As a result, the decision-makers understand the value of record keeping, but that is all.
Few managers are giving any thought to how the underlying data is stored or accessed because they do not yet understand the full value that the data offers and are confined within their own point in time statistics. While the expanded record keeping has helped drive some cost-cutting decisions, they forget to ask the critical business questions that data and analytics can quickly answer.
Since leaders do not yet view the data as anything except a process by-product, the company is not using the data strategically. If analytics are present, the results are confined to one or two managers and are rarely shared across the company.
This stage is where most leaders realize they need a strategy but may not understand what to do. To enter the next evolutionary phase of the analytics process, the company needs to break down its siloed views and metrics, and align its efforts and goals company-wide.
As they are growing, the business wants to take on the behaviors of larger, more successful companies. They move away from spreadsheet reporting and move to create cross-departmental goals and collaboration. They realize that a different approach to analytics maturity is needed.
A natural transition of this company-wide approach is to unite the sales and marketing team. By combining resources and data, the two departments can perform A/B testing to measure the success rates of competing marketing campaigns. With the first step into analytics, the teams can fine-tune their campaigns to have the greatest impact on sales. However, this evolution doesn’t occur overnight.
Up until this point the data has existed in flat spreadsheets and little governance has formed around a unified data model. All of the data sources make it confusing. The management team starts to realize the value and the organization’s need for proper data governance.
At this point, IT may begin moving applications and data into the cloud. They must ensure they do their cloud homework.
Companies also begin to form business intelligence competency centers (BICC) to create a standardized approach toward the adoption of business intelligence (BI) tools. The management team starts using dashboards and scorecards to measure performance and set alerts.
Analytics advocates often appear within the enterprise during this level of maturation. They may be from the IT department, such as the CIO. It is also possible that an aspiring data scientist or seasoned analytics professional will lead the charge.
Either way, the advocate stands out by understanding the value of data. They also recognize that the management team could be persuaded to realize the full value of the data. With proper influence, statistics training, and support, this person can move past obstacles and ignite the business.
However the skill set arrives, the business must welcome it.
In Level 3, the company aligned to standards to be more strategic. In Level 4, the chief executives and the other leadership teams want to use data across the enterprise. One of the chief executives spearheads the effort to move the organization forward using data as the key to understanding the sales pipeline and serving customers better. They begin to build out a dedicated analytics team.
The analytics team must be able to communicate with data. Leadership sets the expectation that reports should have actionable conclusions. The analysts have data storytelling skills and use a functional variety of ways to communicate their data findings.
An example of this next level of analytics is for the analytics team to build “what if” scenarios for the sales forecasts. These scenarios allow the executives to envision the impacts that various campaigns could have on sales. Data becomes the driving factor in the decision-making process.
Integrating analytics across the enterprise is often the most challenging level for companies to achieve. For traditional companies, this is often a radical change in their business model. They want to be data-driven but are unsure of where to begin.
Initially, this means more investment in sophisticated data platforms that can handle large volumes of data with ease and speed.
With success, this analytics team matures into an Analytics Center of Excellence (ACOE) and is ready to take on machine learning and AI projects.
In Level 5, the company has fully implemented a data-driven approach to the business decision process. The executive team includes firmly defined and established roles for a chief analytics officer (CAO) and/or chief data officer (CDO).
The enterprise has embraced a data-driven culture across the business. They seek occasions to incorporate analytics and encourage that mindset throughout each department. The organization uses those analytics to optimize how things are done – may the best idea win!
At this point, the ACOE has adopted mature analytics platforms. These platforms allow them to share models across the enterprise. The company also encourages its vendors to use analytics for improvements, thus shaping new industry standards for analytics.
The analytics team is fully developed. The team is capable of optimizing existing models. The business begins to deploy real-time machine learning to every aspect of the business. A common example is the deployment of real-time customer recommendation engines to drive product bundling and upselling opportunities.
The team has moved to full machine learning solutions capable of providing prescriptive analytics alongside descriptive and predictive analytic solutions. The iteration required for model improvement become integrated into business processes. Additionally, there aren’t separate tasks for updating models with new data because they are set up to ingest new data as it comes available and update model parameters and structure as needed.
Even if the models are left alone, they stay up to date with the latest information, continually maintaining their performance and predictive power.
In order to progress through the levels of the AMM, companies must take the steps to get their strategy right. They must implement the right tools. They must adequately train workers and managers. And the managers must provide the needed support to generate analytics.
Our team has assisted multiple companies as they transform from simple reporting to analytic maturity. Using the levels above, you can lead your business in creating a successful road map.