Let's start right at the beginning with the objectives.  The marketing objectives.  We have already heard about the benefits of building relationships with customers and prospects, moving them along that staircase of involvement from suspect to advocate.  But you need to ask yourself some questions:

How is that appropriate to my business?

Do I want to establish a relationship between customers or prospects and my company?

Do I want the relationship to be built with my sales force, distributors, franchisees or branches?

Where will the information come from?  How will it be kept up to date?  How do I get the commitment to the programme from those who will use it?

The term `database marketing' does quite naturally conjure up visions of systems, computers and software and will present an often-daunting image to the marketer.  This is why the initial steps must remain marketing-led:  after all, the database to be created is to become a key marketing tool, and at this stage it is the marketer's prerogative to expect `Savile Row' treatment rather than `off-the-peg. 

Begin with a check list of factors that must be taken into consideration at the earliest planning phase:

Your organisation.  How do you sell?  Whom do you rely on?  What levels of control can you exert or would you like to exert?  What responsibility do you have for assisting their efforts?

What is your market?  What business are you in?  Where are your lines of communication?  Where does your business come from? Where are the opportunities to be exploited?

What, therefore, are the aims of the database project? 

It is vital that, at this initial planning stage, the aims are agreed and set, since this will determine the functionality of the system yet to be devised. 

Do you intend to develop a communication programme with your customers and prospects?  Will this be by direct mail, telephone, sales call?  A mixture of all three?  Or do you merely intend your marketing database to be used for market analysis and management reporting?

What data is there already in the company and how is it being managed? It is important for the marketer to understand the basic functions of data management that are currently undertaken within the business, since some useful information may already be in existence.

Feedback to the database.  What are the opportunities from the operational activities of the business for feeding the database with new and updated information?

Do you intend to qualify contacts by questionnaires or telephone marketing activities?  Do you expect salesmen or distributors to submit information through regular reporting procedures?  Do you expect the database to benefit from information gathered through direct response advertising, magazine inserts and exhibition visitors, for example?

How do you intend to develop your communication strategy. 

What constraints affect your operation: legal, operational, resources. 

The objective at this stage is to prepare a business specification for the systems people.  This is the beginning of the bridge across that mighty understanding gap between marketing and systems.  The business specification sets out in simple terms what the marketer expects from the system.

While the project should stay within the remit of the marketing and sales division, there is utmost need to involve the systems colleagues from the outset. Only they will have the complete overview of other developments within the organisation that may benefit the marketing database project from something as basic as where to find a particular piece of information held on another system to fundamental directions concerning data sharing, data transfers and other operational aspects.

I recall attending a conference in New York where a senior marketing V-P from a major corporation was explaining how she established a marketing database within her division. A questioner from the floor asked at what stage did she involve her MIS department, to which she replied that they were invited to the launch of the marketing system!

While there may be differences in approach or understanding between systems and marketing people - some even refer to it as a cultural gulf - you guys need each other, if this project is to get off the ground and be implemented in the most cost-effective way. This cultural divide also manifests itself in both the development and operating stages of the project.

Marketing systems, in their own right, tend not to have very high priority when compared to those systems seen as supporting the lifeblood activities of the company (production, financial, etc) and often, if placed with the internal MIS division a marketing database project will often slip down the job diary; similarly, the often ad hoc processing requirements of a marketing system can play havoc with the regular orderliness of systems for other disciplines. This is where an external resource should be considered.

Let's revisit the steps necessary to get to this stage:

First:        Define your marketing plan for as far into the future as you can.

Second:  Define the breadth of operation - include specifically the relationship with other areas of activity ... Sales force, distributors, etc.

Third:     Determine the information you need to achieve your marketing plan.  You must have access to the information you need, but similarly you do not want to encumber the system with information that you will never use.  A telephone number may be important if your plan includes telemarketing; the customer's secretary's name may be important for the salesman.  'Take all you want but use all you take!'.

Fourth:   Review your sources of new and updated information. Decide how you intend to obtain additional information and take this opportunity to check, yet again, what information could be available but not required.  At this stage you must also decide upon the method and frequency of updating the database, who is to have access and how they can effect their requirements.

Last but not least you must consider the future.  There is a need to consider marketing objectives as far into the future as possible:  however, before the overall functions of the system can be determined, everyone must understand and agree the on-going objectives.  These, in generic terms, hardly ever change, since the ultimate objective of a data based marketing project is the continual striving for increased efficiency in marketing activity.

You can define this efficiency as:

    - Increased response rates

- Higher incremental sales

- Reduced cost of sales

- Increased customer retention levels

The factors that affect the achievement of this efficiency must form part of the specification of functions upon which the systems specification will be based. You will need to consider the management of the database, how you will identify the market sectors, how communications and propositions will reach their audience and how the channels of distribution will contribute and benefit.

The business specification provides the vital link between what the marketers expect to be able to do and what they end up getting.  It also provides the basis for discussions around these aims and objectives, so that the systems solution recommended will fit all eventualities.

However, marketers' plain English is sometimes a different plain English from their systems colleagues.  This is where the services of an external consultant can be invaluable.

The systems man will emphasise that the best combination of talents is to find a marketing consultant who understands the applications and capabilities of computers in marketing and can relate to the computer people.  Having assisted in establishing the data based strategy and defining the needs, he will be able to translate them into logical steps in terms that can be understood by the technical team, while permitting the marketer to keep control of the project.  

having considered the marketing strategy and how this really is the corner stone of our solution the second step is consideration of the data in terms of structures and volume and sources and applications and, last of all, the software and hardware requirements.

When considering software you must determine its portability. Can the software be transported between different hardware platforms? For instance, can the same software run on a pc, on a mini computer or indeed on a mainframe? One needs also to consider the integration of the software with other solutions currently operating or proposed for operation within the business.

A key aspect, of course, is the initial cost. This can usually be broken down into a number of headings, starting with the base software cost, added to which are any additional developments or changes needed to meet your business requirements more accurately, and any licence fees payable to use the software.

Complementary to the initial costs is of course the cost of change. Hence the flexibility of the software is a key consideration and this is best illustrated by considering that marketing strategies will change. Therefore, the software must be adaptable at a reasonable cost, not forgetting the support to be given by your software provider in terms of Help lines, New Developments and assistance with software failures.

One question that is asked practically in all cases is should one go bespoke or turnkey?  Let's briefly examine some of the attributes of each: The bespoke solution is software that is written specifically to meet your precise needs.  On the other side of the equation we have the turnkey solution, software very much like the spreadsheet software you might have bought to run on your office pc.  Fill in the order form and it arrives next post.

Looking at some of the points relative to this debate, it is really very much one of balance, and in coming to a decision you have to weigh the advantages of one against the other.  Bespoke software for instance should precisely suit your business needs.  It is written specifically for you.  It is likely to be proprietary, which means it may be difficult to integrate this particular software with the rest of your computing solutions, and it's unlikely that it will be able to be moved across different hardware platforms - i.e. from PC to mainframe. 

The turnkey solution, on the other hand, is more likely to be what is called an open system - which allows it to be operated in many environments.  But its functionality may compromise the business requirement.  It probably won't meet the marketing requirements precisely. 

The cost comparison is often interesting between the two: clearly a one-off development down a bespoke route is likely to be much more costly, not only in terms of the initial payments for the base software, but also in terms of ongoing maintenance and support, whereas the turnkey solution is likely to have numbers of users and therefore the support costs can be spread across a much wider base - consequently they are likely to be lower. 

The same problem also relates in terms of risk.  Turnkey solutions often have large user groups who between them can offer a level of support that is very cost effective.  The bespoke solution, on the other hand, relies on one sole supplier. A difficult problem, but not insoluble. 

There is a solution, and a solution that is now being actively addressed by the software industry. This is given the term tailored software and really is a combination of the best elements of both bespoke and turnkey solutions.  From the software supplier's point of view, a number of the marketing requirements are likely to be similar from one application to the next.  For example, a business to business database is likely to hold certain types of information:  title, initial, surname, and address of the individual contact; something about his purchasing pattern; probably something about the existing products he owns; and maybe something about his repurchase intentions.

Therefore, if a number of things are similar, a skeleton database can be developed.  Similarly in terms of communication programmes:  we are likely to want to talk to customers at different points of their buying cycle, and also to communicate with prospects either through a marketing strategy or on a tactical basis when we have particular market conditions that relate to a specific product or product line.  You will almost certainly need to add new prospects to the database and also to identify when a prospect has moved through his buying cycle and become a customer.  In computing terms, these functions will be similar from one solution to the next, therefore we can develop alongside our skeleton database on a number of functional modules.

Now clearly one can see that with this kind of approach, much of the initial development work has already been done.  What remains for the software supplier is to take your precise business needs as developed by your consultant and tailor these pre-built modules to meet the business requirements; so, the major advantage of bespoke software is acquired in a much more cost effective way. 

The solution is still likely to be proprietary but having been conceived and structured in the way described, it is likely to be much more compatible with existing systems and, because the basic functions are similar from one solution to another, there will be many more users of the base software.

 

This clearly has advantages in ongoing support and maintenance, and effectively gives you an optimised route - best performance, moderate levels of cost and risk, compatibility with your existing systems, etc.

However, the most important element is, that it meets your business needs specifically.

When considering hardware there are three golden rules ("Oh no" you cry "not another set of golden rules!") which will help to dictate your choice of hardware.

Rule 1 - How large is the database?

The conclusion we can obviously draw from this model is that until you understand the number of communications you wish to make, the rate at which those communications might be made and the number of records on the database, one cannot accurately size the machine requirements to provide optimum performance.

In summary, the key to getting started on your marketing database project is knowing totally what functions you will expect to have at your disposal, how these functions will provide the information you require, and how this information addresses your marketing objectives. 

This may sound a little upside down - in fact it is.  

You have to work backwards from any implementation point, to a stage where all that is being considered is the marketing activity.  You cannot just create a database system, load whatever data you have onto it and expect it to be the answer to your problems.

ã Michael Collins 1993-2001