So you're committed to a data warehouse and need a "killer app" to justify your investment. Where do you look? Marketing applications are likely your single largest payback opportunity for your investment. By better targeting acquisitions, by improved cross-selling and retention of profitable customers and by intelligently repricing services, your firm could create more than $100 million in incremental market capitalization for every $1 billion in annual revenues. Technology-enabled marketing has already transformed merger and acquisition activity from focusing on "cost take out" opportunities to seeking "cross selling" opportunities. As an example, capitalizing on cross-selling opportunities is the overarching theme of the Citicorp/Travelers merger, of AT&T's acquisition of TCI and Teleport, and of AIG's purchase of SunAmerica, just to name a few. That's over $150 billion in just these mergers.
Two Historical Marketing Approaches
There's big opportunity in marketing applications. To find out how to unlock latent customer value, it is appropriate to examine historical approaches as well as emerging techniques. Let us first look at legacy approaches. Usually, these approaches are centered on one of two approaches. One focuses on improving marketing cycle time, with the objective of submitting offers prior to the customer making a decision. The other focuses on rigorous analysis, to predict what customers want and to tailor offerings to them. Unfortunately, in practice, these historical approaches have been mutually exclusive, polar opposites.
Simply put, these two approaches treat speed and precision as an either/or situation. To date, the choice has been to move quicker without incorporating advanced precision, or to emphasize advanced targeting without the means to implement results quickly. Fortunately, there are methods now available that combine both speed and precision. Before we examine these methods, let us look at the two major historical marketing approaches in more detail, referring to them as precision and speed.
Bringing precision to your marketing activities means answering the four big questions:
- Who do I target?
- What should I target them with?
- When should I do it?
- How should I bring the offer to market?
Effectively answering "who" means using data mining and/or OLAP techniques to model likely responders, repeat users, non-promiscuous acquisition targets, customers with profit upside, likely defectors and those customers likely to refer business your way.
Answering "what" means ranking your product offerings for each customer, based upon their preference and/or likely profitability. This ranking should also incorporate an understanding of your customer's price elasticity and responsiveness to the various ways you may present offers to them.
Answering "when" means using longitudinal and trigger analysis to predict when buying or defection activity is most likely to occur in sufficient time for you to do something about it.
Last, but not least, answering "how" focuses on the most appropriate channel to present the offer, considering individual customer preference, profitability and contact history.
Since who, what, when and how change with customer life events and competitive activity, these analyses must be refreshed on an ongoing basis. If this sounds like a lot of analysis, it is. Which is one of the reasons why marketing precision has been at odds with speed it simply has taken too much time and money to get these answers, so marketers have gone to market with coarse, broad-brush campaigns.
Focus On Speed
The other major strategy focuses upon speed by collapsing marketing cycle time. The emergence of "campaign management" tools has enabled marketers to better plan, execute and measure campaigns. A modern campaign management system will generate lists without requiring programmers, will automatically initiate marketing activity in response to known buying or attrition triggers, will measure the cost-effectiveness of marketing tactics, will better coordinate promotional activity so that customers are not bombarded with disjointed marketing activity and will customize a customer's experience at any of the enterprise's customer touchpoints. In other words, these tools seek to coordinate a continuous customized communications stream with customers.
Perhaps owing to the popularity of OLAP and report writing tools, most campaign management systems favor a "query-driven" method of coordinating customer communications. For instance, a marketer might seek a profile of profitable credit card customers and then solicit others that are similar to the profile. While profiling capabilities are part of most campaign management systems, robust data mining capabilities are usually not. In our experience, the absence of data mining in a campaign management system empirically costs 15-25 percent in lost marketing returns. In other words, a "query driven" approach might drive $100 million in attributable revenue, whereas seamlessly incorporating data mining would likely drive the figure to $115-$125 million.
We now have insight into the two major approaches that marketers have applied: precision and speed. And, we've seen that these approaches have been historically mutually exclusive. Is there hope on the horizon?
Fortunately, new technologies are emerging which enable marketers to actually combine speed with precision. These technologies are characterized by the seamless integration of analytical activity into campaign management systems. These approaches allow marketers to include behavioral models in their campaign management systems, to dynamically recalibrate such models as the underlying data changes and to implement advanced decisioning systems. By incorporating real-time behavioral modeling into your campaign management application, you dramatically improve your ability to deliver the right marketing proposition to the right customer at the right time in the right channel.
Let's explore the characteristics of these new technologies and identify how you can capitalize upon them.
First, enabling technologies that work with your data in a native relational format, such as campaign management systems and data mining, are now available in the market. The ability to natively connect to your relational database is crucial, as this eliminates data transfer and transformation activities between the enabling technologies. You should definitely avoid proprietary databases (common with some campaign management systems) or separate analytical databases (common with many data mining tools). There will be arguments about performance, but with computer chip price performance doubling every 18 months, it is far better to focus on the productivity of the user than your hardware costs. Focusing on the hardware may save you money now, but it may cost you millions in user productivity long term.
Second, the enabling technologies are beginning to present a seamless interface to the user, either through APIs or ideally through systems that are already integrated. The user ought not to care whether they are in campaign management, OLAP or behavioral modeling. Leading-edge software allows a user to model customer behavior, to generate a list, to track the results and to provide performance reports all from one interface, against one database. By choosing "seamless" systems, you will improve user productivity and reduce the need for "one-off" integration efforts and their subsequent maintenance. If an integrated system is not an option for you, insist that your components connect through a common API and select a systems integrator that will warrant these connections as the individual software producers upgrade their software.
Third, leading-edge marketing applications enable a continuous stream of customized communications to customers. As a result, they require scheduling mechanisms to enable the marketing activity to be initiated by a hard event (the first of every month) or a soft event (30 days prior to CD expiration, terminated EFT deposits into a mutual fund, etc.). The problem is that campaign management, data mining and OLAP tools have independent scheduling mechanisms, so that coordination has historically been a manual, error-prone process. There are two choices to address this. There are software products that combine behavioral modeling, campaign management, measurement systems and an integrated scheduler. This is probably the best choice. The second choice is to select an independent scheduling tool that works with your data mining, OLAP and campaign management systems. Again, it is wise to select a systems integrator that will warrant the connection between all of these products over the life of the project. With either approach, integrated scheduling ensures that the execution of your marketing applications is properly ordered and coordinated.
Fourth, combining speed and precision has a significant impact upon your database design. In short, it fundamentally drives a shift from a traditional analytical focus to an event-driven focus. While the database will continue to contain traditional analytic information such as lifestyle and demographic information and past customer behavioral information, it will also contain detailed information on outbound marketing "events," inbound customer contacts and competitive activity. Moreover, databases designed with event-driven marketing in mind imply more frequent loading of data into your marketing data mart or data warehouse. It spells "foregone opportunity" for your operational systems to collect data on a potential attrition risk without turning this data into actionable marketing activity. The crucial point for you is to select pre-designed data models that are event-oriented, or to select a consulting firm that has experience designing a custom event-driven data model.
In summary, we have explored the two major historical approaches for marketing applications: those focused on speed and those focused on precision. The lack of enabling technologies to cohesively link these disparate worlds has presented a significant impediment to maximized marketing performance. Fortunately, new integrated enabling technologies are now available. The characteristics of these technologies are their ability to work with your data in its native relational format, to present a seamless interface to the user, to coordinate analytical and campaign management activity with a single scheduler, and to base it all on an event-driven designed database. The incorporation of these four characteristics into your plans for marketing applications will be handsomely reflected in your marketing performance and ultimately your share price.
See also: The Do's and Don'ts in Building a Customer-Centric Data Warehouse by Travis Richardson
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access