CRM的概括介绍--CRM: The Power of Prediction

shiyiming

By making smart use of data and information, predictive modeling and analytics can lead to vastly improved customer relationships ― and for your organization, intelligent, cost-efficient sales and marketing
By Hussain Sabri

Empowering the sales force with information tools continues to be a prime concern for commercial enterprises across the products and services spectrum. Organizations have transformed the term "customer acquisition and retention" from an internal performance assessment metric for sales associates into a critical bottom line of competitiveness for firms of all disciplines.

However, this transformation begs an important question: How do you strike a perfect balance between customer incentives and corporate profitability? A lesson many corporations now know is that "one size fits all" doesn't work as a model for customer care. With service personalization becoming the means of responding to this lesson ― and defining competitive success in many industries ― corporations are struggling with how to make personalization a reality.

The answer is predictive modeling. Using the wealth of information the organization already possesses on its clients and customers, which may be augmented with purchased data to help better understand lead potential and prospective clients, businesses can predict the range of products and services that best suit particular customers. Another benefit of predictive modeling is that the techniques only get better with time. The margin of error (that is, deviation between expected and actual results) gets fed back into the predictive modeling system as a secondary input and works as a calibration factor. The result is a predictive CRM system that continues to improve, especially as customers take advantage of the package of products and services that are personalized based on demographic and characteristic attributes and refined over time (see Figure 1).

Value Metrics
A customer lifetime value (CLTV) metric is a net present value calculation that illustrates the relationship between a customer's revenues, expenses, and expected life of the relationship between the customer and the company. CLTV focuses on customer behavior and often incorporates the following factors:

Initial services bought by the customer
Future sales of products and services
Customer service costs
Relationship marketing expenses
Cross-sell revenues
Probability of future purchases and customer retention
Credits, discounts, or other incentives used to keep an account.

You can calculate the CLTV value as the difference between revenues and expenses minus the cost of promotional marketing used to retain an account; all values are discounted back to the present. The CLTV model in its most basic form is shown in Table 1.

You can see that some of these parameters, such as PE, are invaluable to the design process of products and services. Such a parameter "fact" plays a key role in validating promotion expenditures based on projected-to-actual deviation of results.

So, the road is paved to the sales force ― or is it? The fact is that most list brokers don't offer customer-level information. Instead, they offer information (such as income or family size) amalgamated for a group of people (such as 9-digit or even 5-digit ZIP codes) along with their customer attributes.

Statistical Modeling To The Rescue
A clustering algorithm can split the demographics of customers compared to prospects into lightly overlapping groups. Clustering is the development of a predictive model that labels a new instance (in our case, a finite set of demographic attributes) as a member of a group of similar records (a cluster). The number of clusters can either be specified or detected. For marketing purposes, the practice is often to split the customer spectrum into a finite number of clusters to simplify the product packaging process.

Once we have concentrations of demographic attributes, we can apply a second technique, affinity modeling, to predict which products and services sell best together. In its simplest form, we perform affinity modeling by designing a "correlation coefficient" calculation formula. The correlation coefficient usually ranges between -1 and 1. It measures the degree to which we can relate two continuous columns. Usually, the coefficient is denoted by r, which measures the linear association between two variables. If a perfect linear relationship with a positive slope exists between the two variables (for example, between a bank account balance and accrued interest), we have a correlation coefficient of 1. Then, if this positive correlation exists, whenever one variable has a high value, so does the other. If a perfect linear relationship with a negative slope exists between the two variables (such as that between stocks and cash balance in a brokerage account), we have a correlation coefficient of -1. In general, if a negative correlation exists, whenever one variable has a high value, the other has a low value.

Facing Reality
By now, we have estimated what to sell to whom (and for that matter, when to sell it to them). Should we commence action and contact all of our clients and prospects?

Not yet. The world isn't a perfect place ― and neither is any given products and services package. We need to estimate the probability of a particular customer acquiring a particular package. Therefore, the burden is on the data analyst to provide a prediction about what kind of customer fits the bill for a particular products and services package. Regression is the appropriate technique to achieve this pairing.

Unlike estimated results, the data that all regression algorithms use as input is real; it's data of past choices of packages of products and services by our segmented customers. The data tells us the accumulated, actual acquisitions (by customers and clients) of products and services offered by the company.

Understandably, we will discover some diversity in what we could interpret as a perfect match between customers and products. In Figure 2, a generic demographic attribute of the customer (represented in the horizontal axis, using income for example) produces a representation of the likelihood (represented in a vertical axis) that this customer will acquire the product. Obviously, the customer's preference varies: but for marketing, we need a baseline. This is what the red line in Figure 2 gives us. The baseline comes from a calculation based on data deemed acceptable (that is, the data inside the rectangle in Figure 2).

Producing that central regression line is pivotal to the whole process. Once you have it, it becomes the guiding basis for future product and service marketing to customers based on finite ranges in segment-defining attributes. (For example, number of children is an important demographic factor for life insurance packaging.) This baseline is most valuable when used outside the range of data you have ― in other words, for predicting matches between customers and products and services. You can expect that the further you stray from your "confidence interval" (that is, the regression line range for which you have actual data), the wider the margin of error. You can see this effect in Figure 3. There, the portion of the regression line adjacent to the bottom-left corner of the acceptable data zone (outside the inner rectangle) isn't deemed to be of applicable significance: and therefore, you wouldn't expand the regression line for it.

Event Prioritization
Now the circle is complete. You have:

Pulled representative demographic customer data
Segmented customers based on demographics
Packaged products and services based on correlation analysis
Created a predictive baseline model based on past sales history to match customers with products and services.

The next step is to put these predictions to work. You achieve the relationship between customer groups (CGs) and products/services groups (PSGs) through an associative entity whose primary key is the union of CG and PSG entities. Nonkey attributes for this associative entity are pairs of:

Values of suitability measures that associate the CG with PSG
Degree of confidence associated to values stored in 2.

In addition, each of CG and PSG contains, in turn, their own nonkey attributes, which are instances of value functions (for example, income bracket for CG and cost of supplementary material, such as tapes, for a videocam PSG).

All three entities (CG, PSG, and their associative entities) are populated from derived data. Associative entities give us the all-important connection point between first-degree CRM data (that is, nonderived or operational data, such as the customer's name and the product's price) and the CG/PSG entities. Associative entities enable you to recognize that a particular customer might have various membership types in various CGs. For example, a homeowner can also be a renovation contractor ― both identities are possible customers of a building materials store. Ultimately, this association allows you to target the same customer with different campaigns according to the route you take from the product to the customer (via the PSG and CG entities).

Forward Potential
By predicting a relationship between products and customers, then "training" this relationship (calibrating its accuracy rate based on field-testing), you improve the efficiency of your sales force. The sales force can now prioritize customer contact scheduling according to the relative value of the customer against the rest of the list. Second, your predictive intelligence allows decision makers to refocus new product development efforts to concentrate on time-tested themes. Finally, the business can approach individual customers with the right combination of products and services based on their operational and derived profile.

An important last point: The integration point you've created between derived and operational data via the associative entities allows you to accommodate conventional CRM systems. These can continue to function in their traditional fashion should you decide not to use derived data. The integration point can facilitate a smooth transition of the CRM processing base from only operational data to the combination with derived data.



Hussain Sabri [Hussain.sabri@morganstanley.com] is director of Data Architecture and BI Solutions at Morgan Stanley