Just as companies should look to form strategic relationships, they should try to sell to strategic customers. An impressive customer list can give a company a reputation as an innovator or a technological leader. Tandem, the pioneer in “nonstop” computers that never fail, sold its first system to Citibank in New York. To outsiders, the message was clear. If Citibank trusted Tandem, then Tandem must be a winner. Business Week quickly ran an article on Tandem, based in part on the Citibank reference. If Tandem had gone to Southwest Mutual as its first customer, it probably would have taken much longer to develop its reputation.

Key customers can help in other ways too. They can give valuable feedback, providing a company with new ideas on how to improve a product. What is more, key customers feed information about the manufacturer into the word-of-mouth network. If every key customer tells two others, and each of them tells two others…. You get the picture.

Companies should pay attention to choosing the right customers even before they introduce their product. Picking the right beta sites to test early versions of a product can be critical to the product’s ultimate success. Valid Logic, a new manufacturer of computer-aided engineering systems for the electronics industry, found that one of its beta sites, Convex Computer Corporation, provided important suggestions for improving its product. Convex, itself a high-powered startup, was developing a highly sophisticated supercomputer, and thus was able to push the Valid Logic system to its limits. At first, the Convex engineers complained about the Valid system. But as Valid responded to the criticisms and modified its prototype, Convex became an important Valid supporter.

Convex itself faced an important decision in choosing beta sites for its supercomputer. We had long discussions with Convex about the decision. Should the beta site be at a university? How about a government agency? Or a military contractor? Each possible site had its own unique characteristics. Convex would learn different things from different sites. Equally important, the choice of a site would start to position the company. By choosing a military contractor, Convex might be positioned as a military supplier, and then find it difficult to sell into the commercial market. Ultimately Convex chose two beta sites: a semiconductor company which is using the computer for chip design and a petroleum company which is using the computer for geophysical research.

Deciding which customers to sell to requires creative segmentation of the market. Many companies selling industrial products give segmentation little thought. They segment the market geographically, or into big and small companies. This is especially true of startups. Ask a startup: What’s your market? The answer is predictable: The Fortune 500. They all think the Fortune 500 is some type of magic formula. Sell to the Fortune 500 and you’re bound to succeed.