How Customer-Centric Marketing Impacts Databases

By Jeff Fowler

Customer-centricity is a cornerstone of CRM. While it has the potential to reap tremendous benefits, it carries tremendous challenges. Today we’ll take a look at the influence customer-centricity has on database design, the technical hurdles to overcome, and the solutions being adopted.

Before we get started, let’s establish just what “customer centric” means. Keying the words “customer centric definition” into Google, I found the following definition at the very top of the list: Placing the wants and needs of the customer as the central focus of all business practices within the firm. Seeing your business through the “eyes of the customer.”  (I’m sure we all can rattle off a dozen companies who do this without even breaking a sweat.)

Regardless of who’s actually doing it, companies that want to become customer-centric start by centralizing everything they know about their customers into a single repository called a “data warehouse,” which we experts refer to as a GBDB (Great Big Data Base). Depending on your industry, creating the GBDB may require data from a billing system, website, operations, fulfillment, inbound call center, and of course marketing. For now we’ll wave a magic wand and assume we’ve gathered every scrap of data pertaining to our customers; now we’re ready to start mashing it all together into something we can use for CRM. Again employing our wand, let’s magically synchronize this data, thus aligning “Jon Doe, 123 Main Street #123, Washington, DC” with “Jonathan Q. Doe, 123 Main St, Apartment 123, Washington, DC” and all Jon Doe’s other permutations.

That was easy! We’ve gotten data from disparate systems, cleansed it, and made everything line up. Now there are two big problems to solve. The first is storage – we have a heckuvalotta data needing a heckuvalotta space. Fortunately this is the easiest problem to solve, wand or no wand, as today disk space is dirt cheap. A quick search on disk drive prices shows that in 1960, a 10 megabyte IBM drive cost around $36,000, while in 2007 a 4 gig drive (400 times larger) cost $105.93 (360 times less)! Fortunately for us all, benevolent petroleum companies have inverted this ratio over the same time period to keep things on an even keel.

The second problem is more daunting: how to make all this data “fit” into a single, integrated database. Our magic wand cannot help us here – modern relational systems simply cannot blend customer data which includes name & address, multiple telephone numbers (home, work, fax, cell), multiple email addresses, demographics, purchase (and payment) information, product information, solicitation history, and inbound contact history without introducing a significant degree of complexity, which in turn renders our GBDB difficult to use by mortals.

These two issues lead us to problem #3, which is expressed in a simple mathematical formula and can be applied to any relational database:

There’s a story about a frazzled scientist who rushes into his boss’s office and proclaims: “Boss, I’ve got some good news and some bad news. The good news is that we’ve managed to invent a device with the capacity to store all information known to mankind!” “That’s incredible,” says the excited boss, “but what could possibly be bad about that?” Says the scientist: “it takes all of eternity to process.” Complexity and performance are the Achilles heels of data warehouses. After going through so much work and building our warehouse, now we find we can’t use it. So where do we go from here?

Having assembled our GBDB, we arrive at the conclusion adopted by companies who must quickly and easily reach their customers and also need fast and convenient reporting: we’ll spin off a marketing “datamart,” in which information from the warehouse is aggregated and stored in a much more simplified and streamlined fashion to make queries easy and fast. This datamart is typically read/only except during periodic refreshes, which occur after hours.

So the way many companies achieve customer-centricity is to create a data warehouse, but once built it’s not directly useful for marketing (unless you’re willing to wait an eternity for an answer). So just wave that magic wand, create your marketing datamart, and you’re ready to play trumpet on the CRM bandwagon!