Valued customers

John Sidgmore is the man with the task of cleaning out the Stygian accounting stables left behind by WorldCom's flamboyant founder, Bernie Ebbers. Announcing the largest Chapter 11 filing in US history, the company's new CEO was quoted as saying: "Our value is not in the switches and the underground pipes, but in 20 million customers."

From a statutory reporting perspective this is clearly a non-starter. Customers can, and frequently do, walk away. From a management accounting perspective, however, his statement is worth examining. A company's customer base is often the most significant factor in deciding its value in mergers and acquisitions. The value put on a firm, based on the time-honoured "how much someone is prepared to pay for it" method, therefore forms a key part of that most controversial of balance sheet items: goodwill.

If an asset is valued on the basis of its potential for future income, then a realistic assessment of goodwill should be based on a heavily quantified forecast of future revenues from customers. Taking a conservative stance, you could argue that the only customers who should be be considered are those with whom contracts are in place. And, of course, it is only the profit contribution from these contracts that shouldbe assessed - never forget the small matter of fulfilling your own outstanding commitments under these contracts.

In practice, goodwill is always assessed on more than this. To arrive at a realistic price for the shares of a business in the event of an acquisition, there are many issues for prospective purchasers to consider. For example:

* What is the forecast cash flow (discounted as appropriate), and how much confidence do youhave in that forecast?

* What are the key forecast ratios, particularly profit/earnings and profit/turnover?

* Will the acquisition result in a dilution of earnings per share?

The model above offers a method of assessing the overall value of a business's customer relationships. You should be able to obtain the data required for the model from market research, sales forecasts and customer relationship management systems.

A number of questions need to be asked. Their significance will vary greatly, depending on the industrial sector and geodemographic factors, among others. They should include:

* What proportion of revenue derives from past and current customers compared with new business?

* How wide is the spread of customers in terms of current annual revenue, potential annual revenue, profitability etc?

* In so far as there is customer loyalty, what are the underlying reasons for that loyalty, and will it still exist during and after the implementation ofthe proposed changes?

* What expenditure is needed on continued marketing to optimise customer retention?

With answers to these questions, if only in crude terms, it should be possible to identify groups of customers to place a value on the relationship. A significant factor in defining these groups should be an estimated rate of decay to take account of variations in confidence with which annual revenue can be predicted for years into the future.

The annual decay factor is an estimated measure of customer loyalty, bearing in mind a possible optimisation of the trade-off between that decay factor and the budgeted marketing cost. The net present value (NPV) of the forecast annual profit contribution over all future years is derived from the annual decay factor and an appropriate discount factor. The sum of those NPVs should give an approximate value of goodwill.

It's clear that the model is sensitive to a number of highly subjective parameters. If the groups distinguish between "high value" customers, who make healthy profit contributions, and "loyal" customers who have a lower propensity to defect, it's possible to compare the NPVs of those groups using different assumptions for discount and decay factors.

With 4 per cent as the discount factor, the NPV would be the forecast annual profit contribution x 104/(decay rate + 4). With an annual decay rate of 4 per cent, for example, this gives a multiplier of 13; and for a 48 per cent decay rate it gives a multiplier of two. It's clearly crucial to back up any claimed decay factor with genuine supporting evidence, particularly if a low rate is claimed.