ON MEASURING THE COST OF CUSTOMER ACQUISITION.

What does it cost to acquire a new customer? Most software marketers have only fuzzy answers about their own company's costs, and they're even more in the dark about whether their customer acquisition costs are changing and how those costs compare to the competition's numbers.

Ironically, it's the dot-com companies--with their notoriously odd accounting practices--who are beginning to give this little-used metric more visibility. Recently, for instance, an Intermarket Group study revealed that Barnesandnoble.com spends $42 to sign up a new customer, compared to Amazon.com's $27.60, Priceline's $32.30, and Beyond.com's $29.30. Especially in low-margin businesses, numbers like this instantly show who's running the most efficient marketing operations (and who might be dangerously at risk).

Arguably, the cost of new customer acquisition is one of the best ways to measure sales and marketing performance. Traditional metrics tend to focus on budget items ("We'll spend 10% of sales on advertising, $20,000 a month on public relations, two percent of each deal for commissions"), rather than on the customer relationship ("We'll spend $50 to acquire a $1,000 revenue stream"). Most marketers would agree that signing up new customers is more important than fine-tuning the corporate budget, yet the company's metrics impose a perverse standard of performance--hitting the budget numbers.

Moreover, we've noticed that budget-based metrics often encourage a throw-away attitude about customers. It's a truism that long-time customers are vastly more profitable than newcomers, yet most companies keep better track of office supplies and magazine subscriptions than they do of their customer relationships. In the end, a customer is a revenue- producing asset, and it makes sense to account for how that asset is bought, maintained, and put to productive use.

The trouble is, of course, that traditional accounting systems don't really know how to treat customers as acquired assets. As a result, most tracking systems are homegrown and fairly primitive. Some costs get lost in the shuffle, and large overhead items may end up allocated in arbitrary ways. Still, the tracking system should be able to supply data about a few key acquisition metrics:

* The acquisition cost trendline: Over time, the cost of acquiring a customer ought to decline steadily, driven by higher sales volumes, enhanced brand reputation, a more experienced sales force, and product improvements. If costs aren't improving, that's a red flag: Either the market is getting tougher to penetrate, or the marketing organization hasn't figured out how to operate more efficiently.

* The long-term value of a customer: Obviously, it makes sense to spend money on customers who are most likely to keep buying products and services. If new customers are just one-shot buyers or if there aren't follow-on products to sell, upfront acquisition costs should be kept as low as possible. When there's a big downstream revenue opportunity, however, it's often worthwhile to lose money on the first sale (for instance, by giving product away) to lock in a new customer.

* The impact of special promotions: A lot of non-budget marketing tactics--deeper discounts, stronger guarantees, Web site makeovers--can reduce customer acquisition costs dramatically. Inevitably, some of these tactics will be home runs, others may be base hits. In early-stage markets, there's usually a good deal of guesswork about what kind of tactics work best, so companies that test aggressively and creatively almost always end up with a huge advantage in customer acquisition. It's not unusual for the smartest players in a market to attract two or three times more new customers than the competition, with no higher level of spending.

* Acquisition cost by channel and segment: Looking at customer acquisition costs by channel can be an eye-opening experience. (In fairness, it's also important to look at long-term revenue potential for different customer segments and channels.) The differences usually suggest major shifts in how a company spends it marketing dollars--and may trigger a long-overdue decision to walk away from a few high-cost, low-profit customers.