By Ruben Overkemping, CRM Solution Manager, Avanade

In the first article of this series, we introduced the phrase "data is the new oil". While the analogies may seem obvious in some ways, there is also a big difference: oil is something we are running out of, while the amount of stored data continues to explode. So while oil companies are endeavoring on the greatest adventures in the most inhospitable places to find more of this precious resource, the challenge in the world of data seems to be how to control the flow.

We are drowning in an ocean of data, looking for ways to escape the full force of the currents while still harnessing the right insights. In this article, we will focus on how we can control the flow of data that is coming out of the (inhospitable) world of data centers by using an actionable insights platform built on Microsoft technology.

The components that make up this Microsoft platform range across multiple products: Office, SharePoint, SQL Server, Azure, and Dynamics CRM. To explain how these products work together, this article will explore the role these products play in the platform and dive deeper into the products that help a company to get the know its customer.

Getting to know the customer is the first step in a continuous process to make actionable insights work for your organization and will be the focus area for this article. Other steps we see in this process of actionable insights are reach the customer and deliver the experience. Upcoming articles will describe today's role and vision of the platform related to reaching the customer and delivering the experience.

CRM Customer Experience Cycle

By exploring the theme of knowing your customer, we are entering the world of business intelligence, including issues related to data quality and enriching the customer profile.

Knowing your customer in today's world means refining the organization's data and algorithms to build a high quality customer profile. The history of the analytic algorithms used today to build these customer profiles is very interesting. You could say it started in eighteenth century when the parish priest Bayes thought out his naïve Bayes theorem[1]. The "naïve" part of the theorem name isn't referring to the mindset of its inventor but it might as well have as we will see.  

Bayes devised a formula that is used to calculate probabilities which are not only based on the aspects that are known, but even more on the unknowns (inverse probability). In his own time there was no actual use for this theorem as there was no way to calculate outcomes for this algorithm (because of the lack of computing power) and the use cases for using this inverse probability also were non-existing.

Let's now move from the world of a new theorem that had no customers (as we use the term today) nor related use cases, and into a world where Bayes' theorem became like the "discovery of fire", in the world of marketing in the nineteen eighties and nineties. Bayes and other theories behind algorithms such as time regression, sequence clustering, etc. became part of the domain of specialized marketing departments and data specialists and it usually remains there still. But the world is changing and today these specialized domains are becoming more and more the domain of empowered power users in the marketing and customer service departments.

Technology advances and experience with data mining algorithms provide us today with really empowered self-service analytics and the Microsoft platform proves this today. Special role within this world is the Microsoft SQL stack which provides affordable, empowered, high-performance self-service analytics.

Self-Service Analytics

Most people think of data mining as drilling down into the data to try to understand the data based on what is available. But data-mining is providing more thanks to the Bayes theorem as it also provides insights on what is not in the data itself but can be "predicted" from that data.

Microsoft Excel can become an interface for reaching the self-service aspects of enterprise analytics, powered by server-technologies like SharePoint Excel-services, SQL Server Analytics Service, and Microsoft HDInsight (Hadoop Big Data in the cloud). This empowerment of end users does not mean that you take a step back in enterprise manageability. Instead, you take a leap towards consumerization of IT in the field of analytics without losing the managerial grip.

To get an idea on relevant business questions that could drive the use of analytics in the world of CRM, consider this non-exhaustive list:

  • What is the profile of our high value customers?
  • Where is our churn increasing?
  • What is a common theme to customer complaints?
  • What customers are responsible for most of our support costs?
  • What is the next best offer I could make to my customer?
  • Through which channel could I reach my customer in the most profitable way?
  • In which stage is the customer in the buying experience?

Most of these questions can be answered today by non-technical users doing self-service analysis through free-form reporting that allows them to integrate data from disparate sources and drill-down and understand the root cause of data anomalies.  For other questions, it is still important to look at a broader spectrum of analysis possibilities and also to use the data integration possibilities to actually enrich the customer record in the CRM system.

For questions that can be answered by non-technical users performing their own reporting and analysis certain qualifications of the users and use cases will include:

  • Users who are very familiar with the business data and have strong Excel skills.
  • Users who want to easily drill down, pivot, filter, and format the data.
  • Users who are often integrating information from a variety of sources.
  • Users who are usually working with small to medium sized data sets.
  • Users who have minimal specialized technical skills such as SQL, MDX, or other query languages.
  • Although the analysis might be shared with others, distributing the information on a regular basis is not typically the primary purpose of this style.

We will see more on the use of analytics as part of the upcoming article around Reaching your customer where we will show how creating micro-clusters with predictive analytics coming out of the Microsoft platform and enhancing the CRM campaign-records.

Data Management: Enriching CRM

The first part of this getting to know your customer article shows how you can use analytics to answer business questions around your customers. The other challenge of knowing your customers is the verification that your customer record actually represents your customer. One anecdote from my own experience - somehow an online retailer thinks I am interested in offers for skirts because my girlfriend once used my account (and credit card) to order herself a skirt. They could have deduced with all of my other (of course very masculine) gadget orders that the skirt should have somehow been an anomaly to my customer record.

Microsoft provides the tools in its platform to deduce the quality of your customer (lead, prospect) data attributes and a way to increase the data quality around these. Data quality can be managed by using the broader spectrum of Enterprise Information Management processes to help produce accurate and trustworthy data. The goal is to deliver credible, consistent data to the right users with end-to-end data integration, cleansing, and management. Within the Microsoft SQL stack: Data Quality Services, Master Data Services, and Integration Services play a role in achieving this.

Data Quality Services (DQS) provides the functionality to cleanse and de-duplicate data. This functionality can be used in conjunction with Dynamics CRM to manage data when it comes into CRM but also to keep the data quality high as part of a continuous process. From an architectural perspective, think of a job (a-synchronous worker process) that checks which data records in your system need a data quality validation and to offer these records to the data quality services. Use the advanced functionality of the DQS to determine the quality of the data attributes and correct or enhance the record where needed and update it in CRM accordingly. As for advanced functionality, think of fuzziness rules around possible duplicate records and exception highlighting based on inverse probability arithmetic.

Master Data Services (MDS) delivers the features to govern the data available in your systems and to manage the data governance rules to, for example, determine your Golden Customer Record and hierarchy. In enterprise scenarios, MDS can apply data security regulations to the different data domains that are applicable. CRM-integrated MDS scenarios are receiving increasing demand within the Financial Services Industry where the regulations coming from for example Basel III and the Dodd Frank Act require the organizations to unconditionally identify unique customer records and their possible impact on the related outstanding risks.

Integration Services is an essential component in many of the cases that are seen today where CRM has to provide an integrated view of the customer. With the increasing number of data sources available within and outside of the organizations the data silo challenges of the past are only going to explode in the future. Managing the processes of integration through a manageable platform is essential to prepare for the future integrations coming from and flowing into internal and external (cloud-based) sources.

Microsoft Enterprise Information ManagementFrom <>

Before we go into the other aspects of the customer-experience-engagement-platform our next article will go into getting to know your customer by using the Microsoft platform possibilities around Big Data in conjunction with CRM.

[1] A recommended read on the history of many common artifacts of our private life is the book from Bill Bryson, At Home: A Short History of Private Life.