Design CRM Analytics to Deliver the Most Value with this 4 Tier Model

Highlights

  • Customer Relationship Management analytics is less about software tools and technologies and more about defining content and answers that most drive customer experiences and business performance.
  • The Analytics Continuum defines analytics value and impact in a progressive four tier model. Higher value analytics deliver insights that yield more powerful business decisions.
  • The information continuum identifies information types and value and aids the progressive journey from descriptive data to insights, and from backward to forward looking information.
Johnny Grow Revenue Growth Consulting

The CRM Analytics Value Continuum

Not all information is equal in value and business impact. Focusing on the analytics that deliver the most impactful results will accelerate both business intelligence strategies and revenue growth.

The information continuum, which we illustrate in a pyramid model, shows how different types of information and insights deliver progressively greater business impact.

Analytics Continuum

See the CRM Analytics Value Continuum and how it defines the information content that most improves customer interactions, staff decision making and company growth.

Click to Tweet

The Descriptive tier is your standard historical reporting. It accumulates past transactions and displays them in simple lists or reports. CRM software offers out of the box lists, views and reports to produce descriptive reporting.

The Diagnostic tier builds on the prior descriptive reports by permitting drill-down analysis and possibly some causation correlation. It's a clear step up in information understanding and value. However, it only works if you have configured and exposed the information variables so they can be analyzed, modeled or interrogated. Most CRM systems offer some diagnostic reporting, but it's a minority compared to the large volume of descriptive reports.

Advancing from the diagnostic to the predictive tier shifts information from hindsight to foresight. Predictive analytics sift through historical data to show patterns, relationships, trends and anomalies and create simulation, propensity and predictive models. These analytics essentially extend the trajectory of data to deliver data-driven forecasts. CRM systems such as Microsoft Dynamics 365 and Salesforce offer some basic predictive analytics in the application and more so, offer artificial intelligence (AI) tools (Azure machine learning and Salesforce Einstein) so that companies can make their own.

Prescriptive analytics use AI and machine learning to predict and recommend specific actions to achieve forecasted outcomes. The actions are normally accompanied with confidence levels that show the likelihood of success.

Many times, AI algorithms sift through very large volumes of data to make decisions more quickly and accurately than people. Other times AI algorithms are designed to engage people for input and aid better human driven decision making. These decision support capabilities are extremely powerful and have enabled industry adopters to achieve competitive advantage. However, despite some of the hype, no CRM software really delivers these in a packaged fashion. They must be created based on the objectives of each company.

Advancing the value of information and making your way up the analytics pyramid does not start by evaluating tools or technologies. It starts by defining what information will most drive better decision making.

Start with Analytics Use Cases

CRM analytics trends show that most companies are data rich and information poor. To remedy this situation, they must start by identifying how data can best be used to achieve business growth.

As shared in the CRM analytics strategy framework, the first step of any business intelligence program is to define the questions that will reveal the most impactful answers. Below are some types and examples of questions that when correctly answered can boost company performance and revenue growth.

Product questions:

  • What are the optimal prices I should be charging for my products and services?
  • Which of my products or services are declining in customer value, or what products and services would my customers buy if I offered them? Or put another way;
  • What new products and services should be created, and which should be retired?
  • What are my capture products, and how can I grow revenues by lowering the margin on those products, and bundling them with higher margin products?

Customer questions:

  • What 20 percent of customers generate the majority (often around 66-75 percent) of my margins and profits? And why? And how do I elevate other customers to this segment?
  • How do I increase share and keep this most profitable customer segment longer?
  • How can we reduce top customer segment concentration? For example, how can we use look-alike modeling to find new customers matching the attributes of the top 20 percent?
  • How do we find customers with characteristics or patterns of our most profitable customer segment, but not yet in this top margin segment? And how do we get them there?
  • What 5 percent of customers contribute negative profits - and how do we reduce costs to serve this segment?
  • What types of customers put the most downward drag on margins, and how do we restructure activities and processes to lessen that drag?
  • For customers that deliver the lowest margins or financial losses, how are those customers different, what activities or processes drive the biggest losses and how can they be restructured or retired?
  • What 5% of customers deliver the majority (generally more than 50 percent) of the referrals that generate new sales? How do I elevate other customers to this segment?
  • What's a customer worth? What's the customer lifetime value (CLV)? What are the primary variables that impact CLV? What is my CLV by customer segment?
  • Will you earn more return from marketing to fewer high value customers, or more mid-value customers?

The data you already have can probably answer all or most of these questions. While the prior questions may seem like a lot, it's really a small sampling that when answered will modify your business development strategy and operational plans.

When we perform analytics workshops with clients, we generally spend about a day defining the questions that when answered will lead to the biggest revenue gains. It's a tough exercise because clients will ultimately be left with dozens of high impact questions and are forced to prioritize. Fortunately, prioritization can be automated if you design your questions to be answered with information assets and categorize each asset by role, type, value, cost and payback.