March 2009

CRM: Is it really a data democracy?
Sharing customer information is a risky business. In their quest for relevancy, companies must tread carefully. By Emma Warrillow

The 360 degree view of the customer is a fundamental pillar of the CRM promise; data about all our customers in one central location – accessible to everyone, right? But should CRM really be about democratic access to information?

By now, most of us have bought into the fact that having a more complete picture of the customer enables better recommendations, more consistent servicing, and optimized marketing spend. It enables us to avoid the embarrassment and expense of mailing the customer an offer for something they already have purchased from us; to make recommendations before the customer even realizes they need a product; and to make sure that the CEO of our largest client is recognized as a VIP when she comes in with her kids on the weekend.

However, sharing customer information does come with risks.

Let’s start with the obvious; not all information about customers should be shared across your firm. If a salesperson develops a relationship with a client and is given their children’s names and interests, that knowledge does not need to be extended across the organization; can you imagine the discomfort if an unknown call centre rep asked about “Johnnie’s hockey season” or worse still, “How is your cancer treatment?”

Some CRM systems have technologically enabled walls to be erected to data. Users can create private contacts (a girlfriend or brother, perhaps), keep private notes, and in some cases, keep private certain pieces of information such as client home phone numbers or cell phone numbers.

Emphasis on training
When information is widely available, training becomes critical. Customer-facing personnel need training to help them understand appropriate uses of sensitive information. For example, knowing about the children mentioned above might help the call centre rep suggest appropriate products or be sensitive as to when to call, but the private details do not need to be repeated to the customer.
The challenges with sharing apply not only to the raw information collected from customers but also the insights you glean from it. Leveraging customer data through analytics is another fundamental underpinning of CRM success; surely these insights should be shared?

Thomas Davenport has written extensively about firms that have “built their very business on their ability to collect, analyze and act on data.1” He has argued that the most successful organizations will be those that compete through their use of analytics and data. In fact, he argues that “to compete on that information [gathered from ERP, POS, CRM and other systems] companies must present it in standard formats, integrate it, store it in a data warehouse, and make it easily accessible to anyone and everyone.”

So, if Davenport is right, providing democratic access to information to everyone in the organization is critical. However, the rest of the sentence is critical too. If information is to be shared and accessed by many, that information must be presented in a standard, integrated way. This helps ensure that users understand what they are looking at and it is not open to misinterpretation. For example, storing all dates in mm/dd/yy will ensure that a client birthday of April 1 (04/01) will not be confused with one on January 4 (01/04).

Customer analysis and data mining
Customer analysis can yield considerable insights into your customer base and inform decision making. However, customer analysis and data mining have typically been seen as something done by people with pocket protectors in dark rooms - the realm of statisticians. Results are couched in PowerPoint presentations and words like “statistical significance” and “regression” are bandied about! Decision makers are provided only what it is deemed they “need” to see.

I recently heard a speaker at a Web analytics session compare traditional marketing analysts and data miners to “librarians” who restrict access and control what is shown. He was arguing that Web analysts (those who analyze online behaviour) view this differently; tools to support Web analytics have typically been more user-friendly than their statistical counterparts. In many organizations, these tools or the raw reports generated are readily provided to decision makers at all levels.
Putting information in everyone’s hands, however, can lead to misinterpretation and flawed decisions. I remember well a case where a response model was misinterpreted and a decision was taken to mail to all the poorest responders – a costly mistake. In the online world, “time spent on a site” does not typically include the last page visited and thus, a one-page visit would be considered as “no time on the site;” a lack of understanding of subtleties such as this could cause inappropriate decisions to be taken.

True analysis is not about data but rather the interpretation of that information in the context of business issues. This kind of analysis can be an iterative process, where one piece of information leads to more questions. Analysts cannot be locked away in dark rooms and treated like revered rocket scientists but neither is every business person an analyst.

Analysis is key
I believe that relevant information must be shared across organizations. Like Davenport, I believe that analytics can enable companies who make great decisions and achieve success. However, I also believe that true analysis—not just data—is necessary to create insights and make good decisions. Analysts and statisticians are necessary to ensure data is correctly used and interpreted; and analysts must also work closely with the business and understand how it works.

So, in short, I argue that data sharing in CRM is democratic but it is not an anarchic free-for-all. Sharing of information and insights is critical but opening up all data stores to be plumbed by anyone is fraught with risks.

Emma Warrillow runs Emma Warrillow & Associates Inc. (www.emmawarrillow.com) and helps companies articulate their analytic strategies and make the most of their customer data. She can be reached at emma@emmawarrillow.com 1. ”Competing on Analytics” Thomas H. Davenport, Harvard Business Review January 2006