X + OMost companies have boatloads of “O” or Operational data on their customers. Depending on the nature of the firm, this data will include what the customer bought, when, how, for how much, how often and an array of details that vary by firm, sector and customer type. This Operational data often is further enriched with household details – anything from income and home ownership to household composition and the types of cars driven – obtained from either third-party databases, loyalty card programs or direct customer inquiry.
On the other side of the farm in a separate silo (or multiple other silos) companies typically collect and store their “X” or Experiential data from customer surveys, as well as other sources, from social media and emails to contact center info and comment cards.
X + O = the motherlode, the Holy Grail of data. Most firms, however, simply fail to connect the dots across data sources. Whether it’s because of turf issues, resource constraints, lack of knowledge and tools or some other reason, the result is a failure to realize the benefits of data integration, which is the heart of what is popularly referred to as “big data.”
Why Data Integration?
I’m a believer in the economic value of CX: CX strengthens customer loyalty, which, in turn, drives economic value to the firm. But I’m a believer because I have run the numbers with numerous companies using their X and O data. This is not a theoretical belief; it is a belief derived from validation.
You can read all sorts of articles (including some of mine) to make the general case for the economics of CX; but how can you possibly establish the value proposition for YOUR company without using your own data? Is the economic gain trivial or significant? Is it worth the investment? (after all, delivering a great experience isn’t free). Where is the “point of diminishing returns?” (economist-speak for the point where the incremental cost of improved performance equals or exceeds the incremental value created).
X + OBy the way: how do you know which metrics to use? You can, of course, simply assert that (fill in the blank) is the “best” metric. The truly best metric, however, is that which best explains what drives value for your firm. Despite the claims of some companies/experts with branded metrics they push, the truth is that all CX and loyalty metrics are highly intercorrelated – as they should be: after all, they are trying to measure related things. But if you are wiring a metric into your KPIs, strategy or bonus plan, shouldn’t you be using that metric which best explains customer economics for your business model? What’s the point of trying to optimize your performance on a metric that doesn’t optimize your business outcomes? And the only way to determine which metric best explains economic gains for your company is by integrating X and O data.
Few companies manage their customers as an undifferentiated mass. Most segment their customers based on some criteria defined by O data. This might include information on channel usage, tenure as a customer, product usage, “capacity” (that is, what the household’s total current/future spend on the category may be) and any other data that is relevant to the business. If part of the segmentation strategy includes the delivery of different experiences to different segments, then, once again, the X and O data need to be integrated.
There are any number of business questions that can be answered only by integrating X and O data. For example,
• Do delighted customers stay customers longer than other customers?
• Do they give you a larger share of their category spend, buy more/more expensive products and services and recommend others?
• Given resource constraints, how should you allocate efforts at customer retention and the recovery of lost customers?
• Where should you invest (or disinvest) resources around customer experience to get the best return on your investment?
• How do differences and changes in experiences across business units/divisions/regions/stores translate into relative differences and changes in business performance?
In the final analysis, moreover, it is only by integrating X and O that you can validate the results of your CX efforts. Have customers that you delighted rewarded you with loyalty behaviors, that is, behaviors that create value for your company? By contrast, have customers you left disappointed punished you with behaviors that reduced value for your firm by their defection, reduced spend or negative word of mouth? The proof or lack of proof of these behaviors will be embedded in O data some months down the road. (The issue of timing is an important consideration that adds some complexity to analyzing the integrated X and O data.)
Any dataset viewed in isolation is an underdeveloped resource, like an unrefined raw material. It is only by connecting these otherwise isolated datasets that you can realize the power behind the adage that “the whole is greater than the sum of its parts.” While this applies to any data set, it is especially true for X and O data, as the fundamental premise of CX is that changes in X, the customer experience, will lead to changes in O, operational outcomes. You can assume or theorize that this is the case. My recommendation, however, would be to integrate your X and O data to analyze this and other outcomes.