Technology for Optimizing Customer Relationships
Introduction
CRM presupposes the existence of an effective CRM strategy — a solid way
to manage customer relationships in the first place. Once a CRM strategy is
in place, a customer-centric database provides the foundation. Next, enabling
technologies such as real-time event management allow companies to customize
customer interactions based on knowledge obtained from the customer. Then the
entire organization and the marketing and customer service process must be aligned
to support CRM. Finally, companies must apply this information in innovative
ways to optimize customer relationships.
Taxonomy for CRM
CRM is in its infancy, promising a new era in marketing and customer service.
Yet many companies find themselves dangling in limbo as they wait for capabilities
and technologies or practical CRM applications. While there is no one-size-fits-all
solution to this dilemma, those companies able and willing to innovate will leapfrog
the competition.
Let's examine the taxonomy of CRM. The starting point for an effective CRM strategy
is a solid customer database. There is simply nothing to manage if your database
is nonexistent or in disarray. And let's face it, a product-centric or account-centric
design just doesn't cut it. Companies must capture customer data across the enterprise
and consolidate this data into a central customer-centric database. The heart
of CRM relies on solid database design, construction, and ongoing maintenance
and enhancements.
CRM requires knowledge — just having a database won't suffice. Companies
must be intimate with their database and all of its gory details. This means that
companies must know (not just be familiar with) the data and values, and how they
are related. And since a database is not static, companies must commit to continuous
improvements.
Companies should get the absolute most out of customer relationships without settling
for the status quo. E-commerce can provide optimization across all enterprise
touchpoints, but this, too, requires the existence of a solid way to manage the
relationship in the first place.
CRM demands the ability to act on the knowledge gleaned from your customer touchpoints.
What good is knowledge if you can't apply it to optimize your customer relationships?
More about that later.
First, let's look at the capabilities and technologies required to both manage
and optimize customer relationships. These technologies will allow you to tap
CRM's ultimate potential.
CRM Technical Infrastructure
Figure 1.0 represents four major technology capabilities needed to implement
a CRM infrastructure. Database construction is a foundational technology. The
other technologies are enabling technologies.
The CRM Foundation — Marketing Database Construction
Database construction is really at the heart of CRM. Most of your source data
will come from internal operational systems and files, including customer files,
transaction history, billing and accounting, fulfillment and shipping, promotions
and responses, customer service, Web page visits, and leads. Sometimes called
operational data silos, these source systems and files were initially built for
specific applications, and the data in them is typically fragmented, inconsistent,
and unsuited for your marketing purposes. Not only that, but you are constantly
bombarded with significant changes to customer contact information, accounts,
titles, and activities.
What does this mean technically? You'll need to apply software tools to handle
the complexities of transforming your data and improving its quality. But there's
another key area where the right application can assist with your major goal in
building your marketing database: pulling together a unified marketing picture
of your customer relationships. This is done by creating and maintaining multiple,
hierarchical views of your customers, linked to their histories with your organization,
and enhanced with external data on credit, behavior, lifestyle, and preferences.
This cyclical, business-rule-based process — called Relationship Continuity
Management — can be automated.
CRM Enabling Technology — Data Analysis and
Mining
Let's look at the kinds of questions that marketers need to answer. For example,
"How many business prospects are located in the Western region, achieved more
than $1 million in annual sales in both 1998 and 1999, and have NAICS codes beginning
with 23, 31-33, or 42?" Quickly counting customers or prospects who meet a specific
profile, and iteratively refining the segment criteria based on these counts,
requires high-speed access to the marketing database. Marketers need a data analysis
tool that bypasses the arcane complexity of SQL and gives quick answers without
waiting.
"How do my top three customer segments contribute to total profitability, across
product lines, channels, and months of 1999?" Answering this question and then
drilling down into the details can't be done easily in a basic query and reporting
tool. It requires interactive navigation through a multi-dimensional view of the
marketing database, optimized for speedy calculations. So querying capabilities
need to be augmented by online analytic processing (OLAP) tool functionality,
which facilitates asking complex marketing questions and delivers quick responses
to them.
"What is the best product to cross-sell my profitable households with a high risk
of attrition?" Statistical analysis and other modeling tools help marketers build
models of customer and prospect behavior. These models predict responses to promotions
and lift from campaigns and marketing measures such as lifetime customer value.
Depending on the experience of the user, some tools can automate part or all of
the optimization process. Recently, marketers have been exploring new types of
data mining tools that search for unsuspected — but possibly significant
— relationships among customer attributes.
How does the proportion of affluent households within a six mile radius differ
between my top- and lower-performing branches?" A tool for geographic visualization
adds a new dimension to answering questions like this one. Through on-screen
maps and graphical displays, your customer demographics and geography can come
alive.
Enabling Technology — Campaign Management
To act on the insights gained through analysis, modeling, and data mining tools,
marketers need to continuously define and orchestrate campaigns.
In the past, this process was labor-intensive, especially in coordinating the
many steps of a real-life, multi-cell campaign. But a modern campaign management
tool frees its users from low-level production details, allowing them to concentrate
on the marketing issues that go into successful campaigns.
A campaign management tool should connect to the same marketing database as your
analysis tools. This lets marketers use any criteria to define cells, which are
groups of customers to be targeted for particular promotions. It also means that
model scores are immediately available in the shared database. For rapid campaign
development, fast counts are important, as is the ability to easily refine cell
selection queries.
After defining output formats (from mail tapes to e-mail), marketers then need
support for production scheduling, updating marketing databases with promotional
history, and reporting on the effectiveness of promotions.
Determining each promotion's return is critical. The marketer should be able to
define what constitutes success, such as a specific product purchase within 30
days of the promotion. This allows for an integrated look at response behavior
and the profit associated with it, balanced against promotion costs. Cell-level,
campaign-level, and universal random control groups are used to distinguish real
promotional results from overall business trends. Then the scene is set for further
data mining and analysis, to optimize next steps for individual customers and
new planned promotions.
Enabling Technology — Interactive Relationship
Management
The fourth major CRM process allows you to coordinate sales and service messages
across customer touchpoints, based on customer interactions. For example, a bank
might register an unusually large deposit in a customer's checking account. This
represents a marketing opportunity for the bank's investment division, prompting
a phone call to the customer. Or a customer of a recreational equipment company
might be browsing camping product information at the corporate Web site. In addition
to showing pictures of tents online, the company can follow up with a personalized
email — based on both the customer's history and the recent inquiry —
offering a "good-customer" discount on a new line of family-sized tents.
Campaign management uses historical business intelligence about your customers
to craft contact strategies. Real-time relationship management complements this
approach through the use of low-latency tactical intelligence from customer interactions.
A rules-based engine is needed to decide the significance of events and drive
real-time relationship management. It responds to real-time interactions using
pre-modeled customer predictions coupled with a recent snapshot of customer history.
For example, you might want to predict the most likely product for the customer
to buy next, put together a customized message and price for that product, and
factor in the customer's recent contacts with your organization.
Depending on the recommended channel and action, this central tool may link to
another, channel-specific tool to best communicate the actual message. On the
Web, for example, there are a number of software tools that handle personalization,
dynamic Web page construction, and interactive customer dialogs.
Optimizing Customer Relationships
In the brave new world of CRM the possibilities are endless — and the source
of these possibilities is our ideas. Ideas are driven by information gathered
from all customer touchpoints, and they are constantly reshaping marketing and
customer service possibilities. Yet many companies remain glued to a product-centric
or mass-marketing model. A paradigm shift is required to move to a CRM model,
especially one that not only manages but also optimizes customer relationships.
Let's examine some CRM possibilities.
For most retailers, markdowns represent a paradox. For companies that embrace
CRM, markdowns represent the ability to maximize profitable sales by maximizing
initial price business. Typically, clearance sales come at the expense of regular
price business. When the rate of sales on clearance merchandise goes up, the rate
of sales on initial priced business declines. The challenge to the marketer is
to minimize the amount of time clearance merchandise is on the floor.
There are segments of customers on any database that are motivated solely by price
reductions. By isolating these customers, the marketer can target markdowns to
the group most likely to respond and minimize the exposure of full-price shoppers
to the discounted merchandise.
Assuming you have a solid CRM foundation, you can identify customers who do the majority of their shopping (or indeed, all of their shopping) at markdown. You can tailor a message to this segment. The result will be that markdowns will move more quickly, protecting your regular price business and ensuring the needs of each shopper are met.
Let's move to the world of e-commerce and the brokerage industry. Some brokers
may be using online feeds from the trading floor to communicate buying opportunities
to their clients via e-mail. Is this CRM? Perhaps it is. Is this customer optimization?
I don't think so.
Customer Optimization
Customer Optimization presupposes that you have the foundation to determine the
significance of individual customer interactions such as purchases, transactions,
service inquiries, and Web requests and that you can filter and analyze high-volume
data from all customer touchpoints. With this enabling technology in place you
can isolate customers such as those who only purchase fixed-income securities
and you can identify that they purchase them in a regular pattern — perhaps
on the same day every month.
Now instead of simply sending e-mails based on information from the market (which
is, unfortunately, old news the moment you send it), you send an e-mail once a
month promoting fixed-income opportunities. You also promote post-retirement investment
planning to these individuals because you have determined that these individuals
are over the age of 65. This is what customer optimization is all about.
The credit card industry must constantly acquire new cardholders. In the past,
one of the only weapons in their marketing arsenal was to flood mailboxes with
solicitations and clog the phone lines with telephone calls. Don't get me wrong:
this works, but it simply is not the optimal solution. So how can credit card
issuers optimize their customer acquisition efforts and maximize the profitability
of direct mail?
First, your CRM campaign management arsenal coupled with some statistical know-how
will help you determine the optimal number of pieces to mail to generate your
desired response rate (see Figure 3.0) . The right tool will also determine the
optimal number of times to promote a given specific offer. Now, instead of spending
your entire budget flooding mailboxes and clogging phone lines, you can redirect
your marketing dollars to pull some optimal prospects to your Web site.
Many companies are losing money in e-commerce because the fulfillment process
and the customer ordering process are out of synch. Many consumers are comfortable
doing business online, but they may have some questions. Yet companies often welcome
customers with the following greeting: Press the number one to speak to a representative
or visit our Web site for special Internet promotions." And of course, if you
want to speak to a representative you have to wait! Then if you check out the
Web site and call back to speak to a representative, the individual on the phone
knows nothing about the promotions on the Web.
Such integration challenges remain critical for
most companies conducting business on the Web. Since aggregating and managing
data from different systems is the single largest expense of e-business, it's
important to spend your money wisely.
Your CRM campaign management arsenal coupled with some statistical know-how will help you determine the optimal number of pieces to mail to generate your desired response rate.
Technology for Optimizing Customer Relationships
Companies who can capture, analyze, synthesize, and integrate data from all customer
touchpoints can transform this knowledge into a powerful CRM arsenal. Without
it, they are stuck in CRM limbo.
The following is a framework that will move you out of limbo and help you optimize
customer relationships:
- Don't wait!
- Lay down a solid foundation. Remember that a customer-centric database provides a fundamental value — a total customer view. The view should be vertical (multiple sources) and horizontal (longitudinal history).
- Once you have the foundation, add the enabling technologies, including interactive relationship management. These technologies will provide the capability to optimize customer relationships.
- Ensure that your fulfillment process and your customer ordering process are in synch. This means extending your e-business into your enterprise's core business processes.
- Commit to continuous improvement. Remember — undertaking a relationship-centric solution is a journey, not an event. CRM is the beginning, not the end, of successful marketing automation. Plan on continued optimization — across all dimensions.
- Align your culture, your infrastructure, and your entire organization to support your CRM initiatives. This is easier said than done. But if you don't do this you will be stuck in CRM limbo.
- Finally, let your ideas and imagination open up the world of CRM possibilities. Remember — the sky's the limit when you have a solid foundation and enabling technologies.
Conclusion
CRM is a journey and all journeys face perils. One of the biggest perils that
companies can avoid is being stuck in limbo as they sit there and wait for tools,
technologies, or applications. The time is now — get your CRM journey started
or back on track today.
