Marketing is changing ... it is moving from a mass marketing dominance to one that we characterize as distributed, diffused, decentralized, direct, and database.When mass marketing was dominant and the mainframe computer was the norm in large marketing companies, the dominant computing model can be described as a Management Information System. Numerical data were stored on a central computer and accessed by marketing and sales professionals and managers via terminals and personal computers. The extracted data would be used to analyze past performance and to guide future plans. Because marketing was centralized, it was not necessary to provide access to the data to very many people or from multiple locations.
It is not hard to see that this central computing approach might not be optimal, or even adequate, in the new marketing environment. Marketing people are distributed throughout the organization, as well as outside the company in ad agencies, brokers, marketing research firms, data vendors, and consulting firms. Further, the field sales force plays an increasingly important marketing role, and they are located throughout the country, perhaps throughout the world.
It has long been realized by information system managers and professionals that a firm's computing model must match its business and strategy models. When change is slow, it is common for these two models to stay synchronized, but when change is rapid the business models get out in front of the computing models. That is what has happened in marketing.
Not only are these people distributed geographically, they undertake their own data collection, do analyses, write reports, prepare marketing plans, execute programs, and monitor the results. Some of them build databases. For instance, it is not uncommon for a promotion manager to build and maintain a PC-based database on past promotions. Marketing information is thus becoming distributed on the computing platforms of the people who are involved in marketing. Thus we need a distributed and decentralized approach to marketing computing.
Traditionally, marketing and sales systems contained data that were (a) generated internally within the firm via the order entry and shipment systems, or (b) purchased from marketing research firms that collected data on retail sales. The document Waylaid at the Information Waystation discusses some aspects of these data. A lot of marketing information is generated within the firm, and comes into the firm, in paper form ... perhaps because the firm's information system is not designed to capture it in digital format.
Marketing is a very communications intensive activity, and it
is rare that a marketing communication is composed entirely of
numbers. Thus the current computing model is not adequate to
handle marketing data. The new model must recognize the fact
that marketing information comes in multiple media.
For a firm to use their information system to understand markets,
the system must contain all the information about past programs
and events for influencing the markets. Thus, it should contain
past marketing events, and these events are in multiple media
such as video, audio, images, text, as well as traditional numbers.
Further, a lot of those numbers do not naturally reside in the
system unless steps are taken to include them as part of the marketing
But the source the most valuable marketing information may not exist in any media. It may be in the brains of the marketing managers and professionals who work in and who serve the firm.
For instance, a sales manager or broker who has called upon a
retailer for a number of years understands the policies and procedures
of his/her customer. In particular, s/he has an understanding
of how the customer is likely to respond to a promotion offer.
Looked at another way, s/he probably knows what type of offer
it would take to get the customer to deliver a desired level of
performance. We use the term insights to denote this type
of information. Managers have insights, and they use their information
to update them and to form new insights.
Insights are shared in conversations, but rarely put into a computer. That is not surprising, because computers are rarely deployed for aiding in conversations. At least, they are rarely deployed in marketing to aid in conservations. That is too bad, because recent research is pointing to the fact that conversations are the core element of work. People get work done by engaging in conversations because they find that their own information can be leveraged and multiplied by combining it with other people's information. Conversation is the means for such combining of information.
The challenge is thus to use computers to support conversations. That topic is discussed in Marketing's Shared Memory: Concepts and Implications for Computer Support.) Given that computers are used to support conversations, we need to capture that information so that it can be used by others.
The philosophy is to capture it all ... all forms of information, where ever it exists, and in what ever form it takes. This information is likely to exist in many computers, both inside and outside the firm. The old strategy of centralizing that information does not match the new marketing reality. Thus the philosophy should be a decentralized one in which the information remains in its natural "host" computer, be it a mainframe, a file server, a departmental computer, or a personal computer.
If the information is not centralized, means must be provided
for moving from one source to another. Thus, the information
must be linked together. Linking infomation is part of the
new model, as opposed to centralizing.
The linking of information has occurred for numerical and for textual data. Database models have served to link numerical data elements together, and hypertext systems have linked textual data files to one another. As the data formats expand to include video, audio, and other formats, a new term has been coined to denote the linking of all forms of information: hypermedia.
The new nature of marketing involves people working together, and these people are dispersed geographically ... they are separated in place and perhaps in time. Space separation refers to the fact that the people who work together are usually not together in the sense that they are not in the same physical location. In addition, they are likely to be separated in time in the sense that one person performs work at a time that is different from the schedule of other people in the group. This is obviously true when members of the group are dispersed around the world, as is the typical situation in international business projects. It is even common when people are within the same time zone because each individual has his/her own set of projects and activities, and schedules time for the group project when convenient.
The result is a need to support marketing and sales people who are separated in place and time. There has been significant research and development to meet these needs, and this work usually falls under the banner of collaborative computing, or computer-supported cooperative work.
The very essence of marketing is collaboration. It is rare, very rare, that a marketing decision is made without considerable discussion. This is true because marketing is not a science that has a set of principles that one follows in devising a marketing program. Rather, a skilled marketing person has long been considered to be similar to a master chef. Just as the chef mixes foods and spices to arrive at a meal that contains great dishes, the marketer mixes and matches elements of the marketing mix to devise a marketing program that contains great events.
Just as we cannot recognize a great dish until we taste it, we cannot recognize a great marketing program until it has had its day in front of the customers and consumers. Therefore, marketing decisions are difficult and thus involve multiple people. This has always been true, but the diffused and decentralized nature of marketing is highlighting the need for collaboration. The distributed nature of marketing is highlighting the need to use technology to support this collabation, and thus leading to a need for collabative computing.
Thus the new marketing computing and communications model must involve hypermedia and collaboration. It must support marketing and sales because they are inherently intertwined.
But there is one more element that should be central to this new C&C model ... it should contain intelligent applications.
The demassification of marketing has created a situation in which marketing has to be done many, many times.
The move from mass marketing is a move towards individual markets,
each of which must be understood and influenced. Consider the
plight of trade marketing managers who are in the throes of using
scanner data to understand and influence retail customers. Account
level data gives them the ability to understand hundreds of individual
We have coined the phrase what you can measure you can manage to denote the fact that information about the performance of an entitity (a machine, a person, a department, a customer, a market, etc.) allows you to manage it. So, as scanner data moved from market level data (with about 60 markets) to accounts within markets (to about 250 account/market pairs), it became possible for firms to manage at the account level.
We have also coined a related phrase "what you can measure someone will manage" to denote the fact that a competitor will surely step up to the challenge and opportunity of managing smaller and smaller customer aggregates. This truism leads to the conclusion that an individual firm may not have the luxury of ignoring the account level data because his/her competitors will find a way to take advantage of the data ... to understand and influence individual customers.
One interesting part of this trend towards micro-marketing is the evolution of the databases. First we had market-level scanner data (which followed SAMI warehouse withdrawal data, which in turn followed Nielsen store-audit data). Next we had account-level data within the market data. That is, the data vendors provided data about an account (e.g., Krogers) within a market (e.g., Raleigh/Durham). The next move is to store-level data.
The adoption of store-level data will explode the number of times
that a firm can do marketing. Understanding 60 markets has proven
to be a daunting task for most marketing organizations, and none
of them have mastered the use of data to understand and influence
250 accounts. So, how in the world can they market at the store
level? (These issues are explored in depth in The Co-Marketing Paradox.)
Most marketers have not considered this to be their problem. The store has not been the center of marketing thought. Stores are known to exist, and marketing managers do make store visits. But the notion of understanding and influencing individual stores has not been part of their thoughts. Further, it appears that the sales force has not been very successful in servicing individual stores.
An annual survey of retail executives in 1993 revealed that sales
representatives do not pass muster when it comes to servicing
"One area where no one gets rave reviews is service to stores. Trade executives give salesmen failing grades on everything from the usefulness of their visits, to the frequency of their visits, the caliber of their new item presentations and the quality of merchandising material offered." (Progressive Grocer Survey of Buyer/Seller Relations, April 1993)This is not surprising because most firms do not make store service a major component of the salesmen's job. Unless the firm uses store-door delivery, sales representatives are encouraged to call upon buyers, merchandisers, and category managers, not store managers ... much less department managers within the stores.
The arrival of store-level data creates a new competitive arena in which there will be increasing pressure to market at the store level. This is clearly the anti-thesis of mass marketing and thus is very foreign to most consumer packaged goods firms. But, it is a logical place to do marketing because the store is the place where products and consumers meet. The store aisle is where the action occurs, more so than the pantry where the items are stored, the kitchen or bathroom where they are used, or the TV room where they are advertised. Where better to influence a consumer than at the point where the consumer is making the purchase decision? In a lot of ways, the stores are the real markets. Accounts and geographic entities such as Raleigh/Durham are only data aggregates ... figments of the imagination that have been created and perserved because firms could not understand the real markets.
But how in the world can today's marketing and sales organzation even attempt to understand thousands of store-level markets, much less devise and execute marketing programs that influence those markets? It is not economically feasible to deploy the peole and resources that are necessary for such a task. The answer: leverage today's marketing and sales people via intelligent machines. Just as intelligence machines (e.g., robots) have been invented and deployed to leverage engineers and factory workers, intelligence can be placed in the marketing and sales support environment.