6	Managing Lock-In

The great fortunes of the information age lie in the hands of companies that have successfully established proprietary architectures that are used by a large installed base of locked-in customers. And many of the biggest headaches of the information age are visited upon companies that are locked into information systems that are inferior, orphaned, or monopolistically supplied.

In the last chapter we saw how to identify, quantify, and classify the basic sources of switching costs and lock-in. Armed with an improved understanding of switching costs, we are now poised to explore strategies for managing lock-in.

The first portion of this chapter is directed at buyers of information technology, which includes virtually everyone in today’s economy. We all experience some degree of lock-in, yet we all make mistakes dealing with it. To help prevent those mistakes, we provide you with a catalog of strategies to minimize lock-in and avoid monopoly exploitation. We’ll see that you can even make your own switching costs work for you, if you get the timing right.

The remainder of the chapter delves more deeply into competitive strategies for companies that sell their products and services in markets where customers face significant switching costs. As a vendor, you might think that your customers’ switching costs are their problem, not yours. Not so. If you’re trying to break into the market with a new technology, you can ill afford to ignore the costs that your target customers must bear to switch to your products. By the same token, if you are an established player, the extent of the threat you face from upstarts is driven in large part by your customers’ total costs of switching from you to your competitors. Understanding and valuing customer lock-in is a key component to competitive strategy in the network economy.

LOCK-IN STRATEGY FOR BUYERS

Every user of information technology faces switching costs. Before you can craft a strategy for dealing with your own switching costs, you need to know how to identify and measure them. We covered that ground in the previous chapter.

Suppose you are about to select a brand of software to build a mission-critical database. A major consideration in this decision should be how difficult it may be to convert your data to other formats in a few years. You would also be well advised to learn whether you will be dependent on a single vendor to make improvements to the database in the future. Properly measuring these switching costs before lock-in occurs could be worth millions of dollars to your organization down the road.

Once you recognize these future switching costs, what can you do about them? Basic strategy for buyers of information technology who are anticipating lock-in consists of two key elements:

• Bargain hard at the outset of the lock-in cycle for a sweetener or some form of long-term protection before you become locked in.
• Take steps to minimize your switching costs throughout the lock-in cycle.

We are not advocating using either of these tactics exclusively; they should be employed in concert. Let us examine them more closely.

Bargaining before You Become Locked In

As a buyer negotiating for the purchase of a new information system, the best time to bargain for all manner of goodies is before you get locked in. Some sweeteners put money in your pocket right away: initial discounts on hardware, an extended warranty rather than a service contract, or support in switching from your previous information system. But don’t think only in terms of today’s savings. Think ahead to the entire lock-in cycle as you negotiate, and be creative in what you seek: service and support guarantees, free upgrades for some period of time, or most-favored customer treatment. Whatever concessions you seek, your bargaining position will be weaker once you make sunk, supplier-specific investments. An excellent and current example of a savvy buyer exerting his or her influence early in the lock-in cycle can be found in TCI’s choice of suppliers of operating systems for digital set-top boxes for its cable subscribers. As the Wall Street Journal reported, “For eight months, the cable-television industry, led by Tele-Communications Inc., has approached negotiations with Microsoft Corp. for the next generation of television set-top boxes as if it were about to mate with a black widow spider.”¹ Fearful of becoming locked-in to Microsoft, or anyone else, TCI has carefully kept its options open. While ordering millions of copies of Windows CE from Microsoft, TCI retained the right to use Sun’s Java operating system as well.

To extract the best possible deal, you should emphasize the switching costs you will need to incur in selecting a new vendor, such as retraining costs and disruption costs. This tactic is especially effective if you can credibly threaten to continue using your existing system for a while and thereby avoid bearing any switching costs at all. If you can convince a would-be new supplier that your current system still works for you, or that your costs of switching to his new system are large, you stand to get a better deal. Pointing to companies like your own that are doing fine without investing in state-of-the-art technology will strengthen your hand. In relatively mature markets where most buyers already have incumbent suppliers, delay can be a very valuable negotiating ploy.

Another tactic to extract favorable terms up front is to convince your supplier that you are the type of customer most worthy of a very attractive initial package. Perhaps you can establish that you are likely to make substantial follow-on purchases. Honeywell will discount its factory automation system to establish a beachhead at one site of a multisite customer.

A third approach is to convince suppliers that you are capable of influencing the purchase decisions other customers will make. This is a wonderful tactic if you can pull it off: you are effectively getting a referral fee for these customers in the form of an up-front discount. You are in the best position to obtain favorable treatment as an “influential customer” if you can make the case that (1) you will generate a large number of additional unit sales to other customers, (2) these sales will be at a high gross margin, something especially likely with information goods and services, and (3) these effects will be long-lasting because of lock-in. We are regularly solicited by numerous business publications with offers of free subscriptions in exchange for signing up our students. These publishers recognize well that today’s MBA student subscriptions generate tomorrow’s business readers.

Oddly enough, one very effective method you can use to negotiate attractive initial terms is to convince the seller that you will bear very high switching costs later in the lock-in cycle. The greater the vendor believes your switching costs will be, the more you are worth as a locked-in customer, and the more he will invest to get your business. For example, you can negotiate a more attractive up-front price for a specialized information management system if the vendor believes you will face high switching costs in the future when you need various proprietary add-ons.

Emphasizing your own future switching costs is a tricky business, however: you will want to sing a different tune later in the lock-in cycle, so don’t reveal too much about your future vulnerabilities! The truly clever buyer initially leads her supplier to believe her switching costs will be large, thereby extracting a big sweetener. Later, she establishes that her switching costs are in fact much smaller, which helps her to avoid any monopolistic charges later in the lock-in cycle. This is a delicate game in which superior information is the key. Just as you seek to exaggerate your future switching costs, the supplier will try to downplay them. Who is better informed? You know more about your own operations, but the vendor knows more about the technology and the experience of other customers. Be wary.

You should certainly insist that your supplier sign a contract offering you protections throughout the lock-in cycle. You should be aware, however, that some “protections” are not worth the paper they are written on. Even if you can obtain a price commitment for the servicing of a machine you have bought, the seller will retain considerable control over the quality of that service, including response time, the training level of service technicians, and so on. Indeed, you may be worse off if you insist on such “partial” protections: they may merely induce your supplier to cut corners in other costly and annoying ways rather than simply raising price. Ultimately, your best protection will come from your initial discount and from keeping your options open, as we discuss in the next section.

Be wary of vague commitments offered by a supplier eager to lock you in. In the Bell Atlantic example described in Chapter 5, the company claimed it had a commitment to “openness” from AT&T in the design of its switches. But after the choice of switches was made, the two companies had a major dispute over whether AT&T had failed to honor any commitments it made.

Rockwell and Motorola wound up in a similar situation. Rockwell agreed on a specific technical standard for the 28.8 kbps generation of modems that incorporated technology covered by certain Motorola patents. To gain Rockwell’s support for this particular standard, Motorola agreed to license its key patents on “fair, reasonable, and nondiscriminatory” terms. Motorola and Rockwell then had a major dispute over the interpretation of this phrase.

These disagreements are costly to both sides. Prolonged negotiations over the supplier’s commitments absorb valuable management time. The resulting uncertainty makes it difficult to select technology for the future: as a buyer, should you make further investments that will leave you more dependent on the very supplier you no longer trust? Is the supplier truly committed to the relationship or simply milking the situation for a short-term gain? Finally, the sheer transaction costs can mount. As in a messy divorce, only the lawyers end up as winners.

Keeping Your Options Open

Whatever terms you have negotiated at the brand selection point, you’ll want to keep your own switching costs under control. Equally important, you need to convince your supplier that you can easily switch, even if you can’t! This is the best way to obtain favorable treatment once you are past the entrenchment phase.

The time to start managing your switching costs is before you even have any. In other words, from the outset you should structure your relationship to maximize your options later in the lock-in cycle. One such tactic we noted earlier is to establish a second source of supply to which you can easily switch. Your switching costs will then be the costs associated with shifting your business to the alternative source, not the far greater costs of moving to a whole new technology or architecture. A related approach is to pick an “open” system at the outset, so you will not become beholden to a single vendor. If this is your strategy, we urge you to be quite explicit about what “open” means.

Once you have reached the entrenchment phase of the lock-in cycle, or when you arrive at the next brand selection point, you will have a strong incentive to convince your incumbent supplier that your switching costs are low, thereby negotiating the best deal for yourself. One way to do this is to actually switch! Changing vendors can be expensive in the short run but may pay off in the long run if you are then recognized as a customer with low switching costs. This tactic makes the most sense if you, as a buyer, have superior information regarding your own costs of switching. For example, in some cases internal disruption costs make up a significant portion of the switching costs; you are likely to know a great deal more about these costs than would any supplier. You can credibly signal that your disruption costs are low by changing suppliers. You might send this signal to your current vendor by partially switching—for certain parts of your business, or for certain geographic areas—as a way of gaining leverage in negotiations about other parts of your business. This latter tactic is attractive only if the costs of managing a “mixed shop” are not prohibitive.

As a buyer you must be continually aware of the danger of creeping lock-in. Even if lock-in is modest during the sampling phase of the lock-in cycle, it can grow as more equipment is purchased, as more data are stored in certain formats, and as your customers in turn become accustomed to certain types of products or product features for which you are reliant on a particular supplier. As you make these additional investments, you should apply the same principles we have stressed at the brand selection phase: extract favorable terms from your supplier each and every time you become more entrenched as the result of another round of hardware or software purchases, brand-specific training, and so on. This may require coordination within your organization: if one part of your company effectively raises switching costs for the company as a whole, appropriate discounts for others should be negotiated in exchange. This is a major reason to centralize a number of information systems decisions.

In the information economy, buyers can go a long way toward protecting themselves by insisting that they retain the rights to information on their relationship with the seller. It is a lot harder to switch doctors if you cannot have your personal medical files transferred to a new doctor. Likewise, maintenance records on a piece of equipment may be very useful to a new service provider, and data about your measurements will be helpful in seeking a new cyber-clothier. Records on your telephone calling patterns can be very valuable in identifying the best carrier and service plan when you shop for a new provider of telecommunications services. The answer is either to keep records yourself or to retain the rights to transfer records from your supplier to a new one should you decide to switch or contemplate doing so.

Buyer’s Checklist

Here’s a checklist of the items in our recommended strategy for buyers facing lock-in:

• Bargain for initial sweeteners, such as discounts or support for switching from your previous system.
• Don’t be too anxious. Convey the impression that your benefits from switching are small and the costs large.
• Depict yourself as an attractive customer down the road, because of either your own future purchases or your ability to influence other purchasers.
• Seek protection from monopolistic exploitation down the road, but beware of vague promises offering such protection.
• Keep your options open via second sourcing. Partial switching is a way to gain leverage in negotiation.
• Watch out for creeping lock-in, and retain information about usage records.

LOCK-IN STRATEGY FOR SELLERS

Let’s shift perspective now and see how suppliers of information products and services can deal most effectively with the lock-in experienced by their customers. Of course, buyers’ and sellers’ strategies are closely related, and tensions are inevitable in any buyer-seller relationship. Looking over the lock-in cycle, there is a natural tug of war between buyers and sellers: sellers hope to profit from locked-in buyers, while buyers seek to strengthen their bargaining position by keeping their options open. As we will see, however, the lock-in cycle is not a zero-sum game. Both buyers and sellers benefit by structuring their relationship wisely at the outset of the cycle.

As a supplier of information systems, your basic strategy for dealing with lock-in should utilize these three key principles:

1. Invest. Be prepared to invest to build an installed base of customers. Companies unwilling or unable to offer concessions to gain locked-in customers cannot prevail in a competitive battle. Employ tactics to build your installed base at the least possible cost. Figure out how valuable different customers are to you, and tailor your offerings to match.
2. Entrench. Aim for customer entrenchment, not mere sampling. Design your products and promotions so that customers continue to invest in your product or system and become more and more committed to you over time. Incorporate proprietary improvements into your system to lengthen the lock-in cycle and convince customers to reaffirm their choice at the next brand selection point.
3. Leverage. Maximize the value of your installed base by selling complementary products to loyal customers and by selling access to these customers to other suppliers.

In the remainder of this chapter we develop these principles and show you how to put them into practice.

Investing in an Installed Base

Imagine you are about to launch a new information system, such as a new palmtop device or the latest in voice-recognition software. Perhaps you are blessed with a blockbuster technology that simply sells itself. If you are doubly blessed, buyers naturally become locked in, and you are suddenly the proud owner of an enormously valuable installed base of customers that generates income for years to come. If you are this charmed, your biggest problem is what to do for an encore; skip ahead to the section on leveraging your installed base.

Alas, we doubt that many readers can afford to skip this section. Rare indeed is the new technology obviously superior to all other alternatives and naturally giving rise to lock-in. In most cases, you will have to fight to build and retain a base of loyal customers.

Iomega is an illustrative example. In 1995 it launched its now-famous Zip drive, a removable storage device for personal computers with seventy times the storage capacity of traditional floppy drives (100 mB versus 1.44 mB). Iomega designed its Zip system so that Zip drives would accept only Zip-compatible disks manufactured by Iomega. The plan was to build an installed base of Zip-drive users and then earn profits from the sale of Zip disks to these locked-in users. To realize this strategy Iomega invested in building an installed base of Zip drives, using heavy promotional spending and offering deep discounts on the drives, setting the price below their break-even point.

Iomega realized these investments were necessary since there were numerous other competing storage devices, including tape backup systems and ever-bigger hard drives. Confident in its product, Iomega hoped that, based on favorable word-of-mouth advertising, initial sales of the drives would spur sales of more drives and that profits would eventually flow as owners of Zip drives purchased Zip disks on which Iomega earned a good margin. By 1998, Iomega had shipped 12 million drives, but its stock price was rapidly dropping in the face of stiff competition from Syquest, Imation, and others selling rival drive/disk systems.

Your product may be a technological breakthrough, or just one of many vying for consumers’ dollars. Either way, you will need to know how much to invest to build up an installed base of customers and what is the most cost-effective way to make that investment.

LOOKING AHEAD AT THE WHOLE LOCK-IN CYCLE. First and foremost, you must look ahead to the entire lock-in cycle as you invest to build an installed base. It’s all too easy to miss this basic point, which is why we have repeated it several times. At any point in time, you are likely to enjoy a steady stream of operating profits from your installed base, in the form of upgrades, maintenance contracts, sales of new equipment, sales of products complementary to your flagship product, etc. But snapshots like this are of limited value in managing the lock-in cycle. What you really need to do is evaluate the profitability of each type of prospective customer over the lock-in cycle.

The financial analysis of lock-in centers depends on present discounted value calculations, not on this quarter’s income statement. You can value a customer in your installed base by estimating the profit margins on products you will sell to that customer over time. As we learned in Chapter 5, you can also estimate the value of a locked-in customer as the sum of two components: that customer’s total switching costs plus the dollar value of your underlying competitive advantage based on product quality and cost. (This will be a “minus” if your product is regarded as inferior or your costs are higher.)

To guide your promotional investments in new customers as well as your pricing of different components of your information system, you must treat each locked-in customer as a valuable asset. Only in this fashion can you determine how much to invest in attracting new customers. For example, in using a competitive upgrade program to sell computer software, you need to quantify the likely follow-on sales to a customer you have captured from a rival to properly set the competitive upgrade price.

Traditional, static accounting data are likely to be of limited use in this exercise. Comparing your current promotional expenditures to margins on your software will simply not do the job. You need to look ahead throughout the lock-in cycle, and you need to break down the analysis by type of customer.

While you surely want to tout to the investment community the steady income stream you expect to earn from your installed base of loyal customers, looking at that stream may be of limited value in determining how aggressively to seek new customers. The reason is that old and new customers are likely to have different demographic and usage profiles. For example, as cellular telephone companies increased penetration of their products over the past decade, they typically found that average calling volumes declined with each influx of new customers. The reason is not hard to see: the early adopters were the heaviest users, those with long commutes, urgent business, and high incomes. Valuing new customers based on the calling patterns of these hard-core customers would have been a mistake. As a general rule, the stream of margins you earn from an established group of customers is primarily helpful in determining how hard to fight to keep those customers if a rival is trying to steal them away from you. Be careful using them as a guide to valuing a new group of customers.

FIGHTING FOR NEW CUSTOMERS. Some pundits would tell you that traditional economic principles, and especially the textbook model of “perfect competition,” retain no value in the information economy. You know by now that we disagree. The logic of lock-in affords a good example. Under classical perfect competition, many small firms compete on price. This intense rivalry drives price to cost and excess profits to zero.

What happens when perfect competition meets lock-in? How can we reconcile vigorous competition, which eliminates excess profits, with lock-in, which makes an installed base a valuable asset? Think about the extreme (and unpleasant) case in which you face fierce competition from equally capable rivals to attract customers in the first place. Both you and your rivals know that each customer will be locked into whatever vendor he or she selects. The result is that competition indeed wrings excess profits out of the market, but only on a life-cycle basis. The inescapable conclusion: firms will lose money (invest) in attracting customers, and (just) recoup these investments from profitable sales to locked-in customers.

In the presence of lock-in, intense competition will force you to offer very attractive initial terms to customers, so that on an overall, life-cycle basis, you would earn no more than a normal rate of return on your investments. Once you have an installed base in place, it will look like you are earning substantial operating margins, but this is merely the normal return on your initial investment in attracting and building the installed base. Economists call the margins earned on sales to the installed base quasi-profits: they look like real (excess) profits at a particular point in time but are merely a normal rate of return on prior investments.

How can you earn more than a “normal” rate of return in a market with lock-in? By and large, the key to obtaining superior financial performance in “lock-in” markets is the same as in other markets: by product differentiation, offering something distinctly superior to what your rivals can offer, or by cost leadership, achieving superior efficiency. Ideally, you would seek both differentiation and cost advantages.

In the network economy, simply being first to market can generate both differentiation and cost advantages. The key is to convert a timing advantage into a more lasting edge by building an installed base of users. Like Amazon, you may have a first-mover advantage that allows you to build an installed base before full-fledged competition arrives. You may have a superior product as with Intuit and its Quicken software. Or, like the Wall Street Journal Interactive, you may have an informational advantage in reaching or identifying customers based on sales of other products or based on a distribution network or brand name.

Within the broad categories of product differentiation and cost leadership, some distinct tactics arise in the presence of lock-in; we’ll discuss these directly below. Our point here is that you should not confuse quasi-profits with real profits. Unfortunately, in some cases the federal courts are doing just this, classifying firms as “monopolists,” potentially subject to antitrust liability, merely because they have some locked-in customers. This happened to Kodak in the copier business.

Kodak’s share of the high-volume copier market was some 20 percent and declining in 1995; it was earning such anemic returns that Kodak ultimately put its copier business on the block and sold off the copier servicing business to Danka. Despite this, Kodak was hit with a $70 million jury verdict (halved on appeal) in 1995 for “monopolizing” the market for servicing Kodak high-volume copiers. Kodak’s supposed crime? Refusing to sell its own patented and proprietary parts to independent service organizations (often former Kodak service technicians) who wanted to compete with Kodak. Kodak’s misfortune is especially relevant to high-tech firms, since the jury verdict was affirmed by the Ninth Circuit Court of Appeals, which has jurisdiction over the entire Western United States. Ironically, Xerox, with a commanding 70 percent share in the very same market for high-volume copiers, was spared a similar result by a federal court in the Midwest, which ruled that Xerox had the legal right to refuse to sell its own patented parts to potential aftermarket competitors. Kodak’s request for review by the U.S. Supreme Court was denied in 1998.

In our view, Kodak’s revenues from the service business were simply economic returns on its deep discounts on initial sales in the highly competitive copier market. Just as industry participants should look at the entire lock-in cycle, so should the antitrust authorities and the courts.

STRUCTURING THE LIFE-CYCLE DEAL. As in any complex negotiation, there are mutual gains to trade from structuring the life-cycle deal to best reflect both buyer’s and seller’s needs, tolerance for risk, time value of money, and beliefs about the future evolution of the market.

For example, buyers sometimes have separate budgets for capital expenditures as distinct from operating expenses. In selling durable equipment to a buyer with a tight capital budget, you should offer a discount on the equipment and capture a greater portion of your revenues in the form of a multiyear service contract. This approach will also be attractive to a buyer with an especially high cost of capital.

Vendors offering proprietary systems naturally shift customers’ payments over the life cycle into the future, or the back end, from the front end. Buyers obtain a sweet deal at the brand selection point, knowing they will then face real switching costs throughout the remainder of the cycle.

If your customers are worried more about lock-in than obtaining the very best terms up front, you can take the opposite tack and assure them that they will not be in your power in the future. This approach underlies companies’ promises that their products will have an “open” interface. Promising “openness” is a tricky business, however, because eventually you will want to get your customers more entrenched. We mentioned earlier that buyers should be wary of such promises. Well, the same holds true for vendors. Don’t promise more openness than you really want to deliver. The risks to your reputation, not to mention legal risks, are very real.

Even with the best of intentions, the meaning of an “open” system will be subject to interpretation in the future. Microsoft has at times indicated that its operating system is “open” in the sense that independent software vendors (ISVs) will have full access to the Application Program Interfaces (APIs) necessary to make their applications work well with Windows. Yet Microsoft’s own programmers remain in a preferred position in writing applications for the Windows platform since, inevitably, they are going to know first about changes in the operating system.

In contrast to Microsoft, Netscape has adopted an “open” strategy in the browser wars. Netscape’s approach is much like that used by Adobe when it introduced its page description language PostScript (see Chapter 8): the intent was to convince potential adopters that the product is open enough that they would not be captive to Netscape (or Microsoft!) down the road. On the other side, both Microsoft and its customers know that the customers are already locked in to Microsoft’s desktop applications. Microsoft now wants to convince the customers to extend this lock-in to the Internet by integrating Web applications with the desktop and local-area applications. Users face a clear choice: go with Netscape, open standards, and relatively low lock-in, or go with Microsoft, which offers a highly integrated system and high switching costs in the future.

Beyond this, “open” and “closed” aspects of an information system often coexist. A software publisher may have a nonproprietary, open interface with limited functionality and a proprietary interface that performs far better. For example, Cadence Design Systems, a leading supplier of electronic design automation software, has several industry standard interfaces that other software companies can use to move designs and/or data between its flagship product, Virtuoso, and other programs. However, Cadence also has a superior, proprietary interface for internal use. Another example of an “open” standard with limited functionality is Microsoft’s Rich Text Format (RTF) for word processing files. This format lends itself to fairly easy conversion, but it is quite limited in its scope, and some of the formatting attached to the original document is inevitably lost in the conversion process.

HIGH MARKET SHARES DON’T IMPLY HIGH SWITCHING COSTS. Certainly it makes sense to extend introductory offers to attract customers from whom you can expect to earn sizable profit margins in the future, after they become “loyal,” or “locked-in,” depending on your perspective. However, the quickest and surest way to take a bath in lock-in markets is to count on lock-in that does not materialize. If you give your product away, anticipating juicy follow-on sales based on consumer loyalty/switching costs, you are in for a rude surprise if those switching costs turn out to be modest. You have to form an accurate estimate of each customer’s future switching costs to determine the revenues you can expect to earn from that customer and thus the maximal prudent investment you should make to acquire the customer in the first place.

One danger is the emergence of aftermarket rivals that can serve your customers without imposing significant switching costs on them. This was Borland’s strategy in offering Quattro Pro: to attract Lotus 1-2-3 users and minimize their switching costs. Kodak and Xerox each faced a similar problem in the market for copiers: after competing aggressively to place new, high-volume copiers in anticipation of earning healthy gross margins for servicing those machines, they found third-party service providers attacking their installed bases. Hewlett-Packard has faced a similar threat from third-party refillers of cartridges for its printers.

It is all too easy for companies in the information business to downplay the likelihood that imitators will emerge and drive down prices and/or drain away their installed base. Part of the problem is that rivals often design their products to minimize switching costs. The battle of the browsers between Netscape Navigator and Microsoft’s Internet Explorer has this character. To judge by Netscape’s market capitalization, investors believed for some time that Netscape’s installed base of Navigator users was an extremely valuable asset. We are skeptical, though, since we doubt that the costs of switching from Navigator to Explorer are very significant for most users. Consequently, Netscape’s share in the browser market has been steadily evaporating with Microsoft improving Explorer, giving it away for free, and incorporating browser functions into the operating system.

Microsoft, of course, has a key strategic advantage in its dominance of the desktop operating environment. It wants to integrate its Internet browser with the file browser and other components of Windows 95 in a way that Netscape will find hard to imitate. Microsoft has said that Internet Explorer will “always be free,” but what this means is that it will simply be bundled with the company’s desktop environment, either through bundling or by some type of product integration.

Another company with an impressive market share but little evident buyer lock-in is Cisco Systems, the supplier of some 80 percent of the routers that make up the basic plumbing of the Internet. Cisco enjoys a breathtaking market capitalization: some $67 billion as of this writing, based on roughly $8 billion in annual revenues. Cisco’s value depends on its reputation for high quality, its full line of compatible offerings of networking hardware, and especially its ability to stay one step ahead of the competition in terms of product performance. Cisco’s value by and large is not based on its ability to earn profits from a captive installed base of customers. Cisco’s router designs generally employ open standards for the flow of traffic over the Internet. These open standards have done wonders to fuel the growth of the Internet, and they make Cisco’s products attractive to customers. But this same openness makes Cisco vulnerable to competition. In an attempt to reduce this vulnerability, Cisco has now branded the software that runs its routers under the acronym IOS (Internetwork Operating System).

A key question for Cisco is whether it can continue to outpace its rivals using an open architecture, or successfully incorporate proprietary features into some of its products to give it a more lasting competitive advantage. So far, Cisco has thrived in the open Internet environment. It has wisely plowed a considerable fraction of its retained earnings into acquisitions of providers of products and technology that are complementary to its core router business, such as hubs (simpler devices that link together small groups of computers) and frame-relay devices and switches, which Cisco gained in its $4 billion acquisition of StrataCom in 1996. Indeed, Cisco is widely seen as having mastered the art of acquisition: it relies on acquisitions to help stay ahead of the competition and have access to new, proprietary technologies.

The fact that a company has a large share of the installed base, as does Cisco in routers, is no guarantee that it will also have a large share of current shipments. Indeed, a divergence between these two measures of market share should serve as an early warning system for any firm: if your share of new shipments exceeds your share of the installed base, you are gaining ground on your rivals. The higher the switching costs, the greater the inertia in the market, and the smaller will be divergences between historical shares, as reflected in shares of today’s installed base, and current placements.

Indeed, recently Cisco has found itself under attack from rival router vendors such as 3Com. 3Com’s strategy is to offer comparable technology at a much lower price; it is offering high-end routers at $15,000 to $20,000 that are directly competitive with Cisco’s offerings at $65,000. We will soon see how locked in Cisco’s customers really are!

Netscape is also finding itself in this position. Netscape’s share of the installed base of browsers remains high, but its share of new “placements” is considerably lower. But this figure, too, must be interpreted with care since Microsoft is giving away Internet Explorer to so many users. Because lots of software just clutters up people’s hard drives, monitoring usage of the software is critical; the “active” installed base is far more meaningful than cumulative historical placements. In the case of browsers, it is possible to measure usage by looking at the records kept by Web servers, which record access by each type of browser.

Just as a large market share does not automatically imply lock-in profits, a company with a small market share may still have a valuable franchise if its customers generate substantial ongoing revenues and are unlikely to switch vendors. Our discussion of Computer Associates in Chapter 5 illustrates this point. Despite the fact that the market for mainframe computers has been stagnant for years, Computer Associates has performed very well. It is not an especially popular company, but many of CA’s customers would rather pay a premium price than bear the disruption costs and risks associated with going elsewhere for vital software.

Sure, having a large market share and customers with high switching costs is the best of both worlds. But you may never build a large share without giving customers choices. And a small, secure piece of the market can make for a very profitable operation. If you go with such a niche strategy, just make certain you really have a unique offering that will continue to appeal to a certain portion of the market. Be prepared to be at a cost disadvantage based on your small scale, and don’t be surprised if your share gradually erodes over time, especially if your product, like computer software and content provision, is subject to strong economies of scale.

ATTRACTING BUYERS WITH HIGH SWITCHING COSTS. The higher a buyer’s switching costs, the harder it is worth working to get the buyer in the first place. But there are two things to watch out for. First, any buyer who is expected to become locked into your products is likely to be locked into a rival’s products already, making that buyer more difficult to attract in the first place. Second, as we noted above, the buyer has an incentive to inflate his switching costs up front precisely to obtain a hefty sweetener. Don’t believe everything you are told!

Still, you can study the buyer’s operations and needs to estimate switching costs. For example, if you make an initial sale of hardware or software, will the buyer have a strong preference to buy additional units in the future as new “seats” need to be filled, so as to maintain “single vendor simplicity”? If the buyer’s lock-in is significant in magnitude and/or duration, you can expect more follow-on business, and you will be able to capture larger margins down the line.

Furthermore, you may need to fight hard to overcome the buyer’s costs of switching to you. This means that you may need to subsidize the buyer’s costs of switching. But watch out for customer churn: if you offer a deep discount to subsidize a customer’s switching costs, and this particular customer turns out to have low switching costs, you may never recoup the subsidy, since any attempt to do so later in the lock-in cycle will induce this customer to switch again. Indeed, some individuals repeatedly switch long-distance telephone companies to exploit the introductory offers made by AT&T, MCI, and Sprint to attract new customers.

Buyers with growing needs, and thus growing switching costs, are especially attractive. If you are lucky, a small incentive now will generate healthy gross margins on a significant volume of “aftermarket” needs, including demand for various complementary products.

SELLING TO INFLUENTIAL CUSTOMERS. Marketing aggressively to influential buyers can be a very effective way of building up an installed base of customers. When deciding how much to invest in capturing an influential buyer, it is important to quantify the benefits that can result from such investment. The critical measure of a buyer’s “influence” is not cash, income, or even visibility. It is much more specific. The appropriate measure of a buyer’s influence is the total gross margin on sales to other customers that results from convincing this buyer to purchase your product.

If you sell into one part of a large company, will you have a better chance of making more sales within the rest of the company? Will this buyer stimulate other sales through word of mouth or referrals, or when its employees join other companies? Will other buyers be impressed that you sold to this buyer, perhaps because the buyer is sophisticated or is known to be good at evaluating products like yours? These are the kinds of questions you should ask yourself when trying to decide how much to invest in swaying the decisions of a potentially influential buyer.

By convincing Industrial Light & Magic to use its graphics computers to create the dinosaurs for Jurassic Park, Silicon Graphics hoped to showcase its workstations and spur future sales. The resulting benefit to Silicon Graphics was especially large, since programmers at Industrial Light & Magic created new object-oriented software tools for the project that complemented SGI’s hardware and since Jurassic Park helped deepen the long-standing relationship between SGI and IL&M.

A large company may be influential because it dictates to others the format in which it insists on receiving information. If you can convince Intel to use your protocols and formats for its electronic design automation software, it is a good bet that other, smaller companies engaged in designing integrated circuits and printed circuit boards will do likewise. If you can convince a major motion picture studio to use your software for creating special effects, that may effectively lock in other, smaller customers to your formats as well. By discounting to Intel or Sony, you stand to capture valuable business from others who will pay a premium for your products. Of course, Intel and Sony know this full well.

A big buyer may also be influential because it helps establish or promote a product standard, as we discuss in Chapter 8. For example, in the modem industry Rockwell manufactures the majority of the chipsets that are the brains of the modem. Consequently, it is in a position to greatly influence the standards and protocols by which modems communicate with each other.

Even little buyers can be highly influential if compatibility is an issue. When modem speeds upgraded from 1,200 bps to 2,400 bps in the early 1980s, manufacturers offered special discounts to operators of bulletin board systems, since they realized that consumers would be more likely to upgrade if there were lots of bulletin board systems that consumers could access with their high-speed modems. For each modem given away at cost to the system operators, the manufacturers managed to land dozens of modem users who wanted to access that system. We will examine this sort of strategy further when we discuss the concept of “network externalities” in Chapter 7.

Buyers can gain influence because they are perceived as leaders, be they large or small. This has also been a driving force in the fashion industry. In high technology, demonstration effects are very important, as is the implied or explicit endorsement of respected users. Just as a highly respected hospital can lead the way in commercializing a new medical procedure, a leading-edge, high-tech information services company can induce others to adopt new information technology by using and endorsing it. A big part of Sun’s marketing strategy for Java was bringing in big-name players to endorse the product. Eventually, Sun managed to land the biggest fish of all, Bill Gates, although rumor has it that he had his fingers crossed behind his back when he announced Microsoft’s support for Java. Currently, Microsoft has banned Java from its Web site, has added Windows-specific enhancements, and is pushing Dynamic HTML and XML as alternatives, all of which are viewed as attempts to derail Sun’s Java plans.

MULTIPLAYER STRATEGIES. Selling to influential buyers takes advantage of the fact that one customer can influence others. The pursuit of several related “multiplayer” strategies rests on this same idea, although it involves different combinations of players. None of these strategies is entirely new, but they all work best for products with high gross margins and thus are especially well suited for information products.

Airline frequent-flier miles are a good example. These loyalty programs often involve three players: the airline, the passenger, and the passenger’s employer—that is, the one who is actually paying for the ticket. As you are no doubt aware, there is a temptation for the traveler to book with the airline giving him or her the greatest frequent-flier benefits, at least if the individual can personally appropriate the miles. In this way, the airline can effectively bribe the traveler, with a relatively small frequent-flier benefit, to fly its airline, despite what might be a steep price for airfare. The airline is using the frequent-flier program to drive a wedge between the interests of the payer (the employer) and the decision maker (the employee/traveler).

This pattern often arises when one customer participates at the outset of the lock-in cycle and others follow on later. For example, infant-formula manufacturers make very attractive offers to hospitals for their formula because they know new moms display a strong tendency to use the same brand at home after they leave the hospital. Likewise, automobile manufacturers have historically obtained very attractive terms from spark plug makers because many consumers displayed a tendency to replace the spark plugs in their car with the brand used by the original equipment manufacturer.

Focusing attention on one party that can lock in others also works when the decision maker and the payer are inside the same buying organization: a medical device manufacturer may try to enlist the support of a key physician to sell a hospital its brand of medical equipment, be it a catheter or a complex diagnostic machine. The manufacturer knows that the doctor will have considerable say in how the hospital spends its money and that the hospital is likely to become locked in once it starts using a particular brand or model of medical device. The attention lavished by the manufacturer on the physician can range from straightforward marketing—selling the medical professional on the virtues of the product—to an outright bribe in the form of a research grant or an invitation to a boondoggle conference in Hawaii.

Another group of multiplayer strategies includes sales to users of complementary goods. For example, when Alias Research, a high-end graphics software house acquired by Silicon Graphics in 1995, sells its animation software, these placements help promote sales of its complementary rendering software, since the two types of software work smoothly together in the production flow generating computer simulations.

One way to exploit these complementarities is to subsidize the customer who purchases first, and to recoup this investment from subsequent customers of related products who will then pay you a premium. Of course, this strategy works only if matching the first customer’s brand choice improves performance for the second customer. A variant on this theme is to subsidize the more far-sighted customer group, recouping this subsidy with revenues earned from groups less able or less willing to factor in future costs at the beginning of the lock-in cycle.

Netscape tried to employ a complements strategy, building an installed base by giving away its Web client, Netscape Navigator, in order to sell its Web server product. However, as we just discussed, this strategy is risky if buyer lock-in to the primary product—in this case, the browser—is shaky.

Encouraging Customer Entrenchment

Your work is not done once a customer joins your installed base. You merely move on to the next stage of the lock-in cycle: entrenchment. Your goal is to structure your relationship with customers to simultaneously offer them value and induce them to become more and more committed to your products, your technology, or your services.

ENTRENCHMENT BY DESIGN. You can influence the magnitude of your customers’ switching costs. Just as buyers are reluctant to become more reliant on a single source, sellers have incentives to encourage customers to invest in the relationship, thereby raising their own switching costs.

During the lock-in cycle, the buyer and the seller will perform an intricate dance, causing the magnitude of lock-in—that is, the buyer’s switching costs—to vary over time. As a seller you should attempt to incorporate new proprietary features into your products and services to raise switching costs. Buyers will try to resist this. For example, in high-end graphical software, many advertising agencies and other users purchased both Adobe’s Illustrator program and Aldus’s Freehand program, despite considerable duplication in features, to reduce reliance on either Adobe or Aldus. Alas, this strategy did them little good when Adobe and Aldus decided to merge.

Another wonderful way to get your customers entrenched is to offer them more and more value-added informational services. The pharmaceutical drug wholesaling business illustrates this point nicely. Traditionally, this business entailed ordering pharmaceutical drugs from manufacturers, warehousing them, and delivering them to customers such as drugstores and hospitals. The role of information systems and services in this industry has grown markedly in the last decade. The industry leaders, McKesson, Cardinal, Bergen Brunswig, and Amerisource, now distinguish themselves by offering sophisticated reporting services to large, national customers. To further entrench these customers, the large wholesalers have developed their own proprietary automated dispensing and reporting systems, along with consulting services to deepen their relationships with clients.

LOYALTY PROGRAMS AND CUMULATIVE DISCOUNTS. Vendors explicitly control buyers’ switching costs with the “artificial” loyalty programs discussed in Chapter 5. The key to such programs is that the reward to past loyalty must be available only to customers who remain loyal. Usually, this is done in two ways, each of which involves ongoing special treatment of customers who have cumulated substantial usage in the past. First, those customers may be given preferential treatment; this is the essence of United Airline’s Mileage Plus Premier program, whereby very frequent fliers are given preferential seating, chances to upgrade to first or business class, a special telephone number for service, and so on. Second, historically heavy users are given bonus credits when they buy more goods or services; with United Airlines, this takes the form of double or triple miles for those who travel heavily on United.

In the end, all these methods are forms of volume discounts: favorable terms for incremental purchases to customers who are heavy users on a cumulative basis. Again, we emphasize that these methods require tracking individual customer purchases over time, establishing accounts for each customer that record purchases, and maintaining a balance of some credits associated with frequent buying. As information technology continues to advance, this information processing will become less expensive, and more and more companies, including smaller retailers, will find customer tracing to be cost-effective. In an earlier era, numerous retailers banded together to offer cumulative discounts: that was the essence of the Green Stamps system, whereby customers would accumulate stamps issued by many vendors and then trade in books of stamps to earn prizes. In today’s economy, smaller vendors will again find it attractive to link arms with companies selling noncompeting products to offer cumulative discounts. We doubt your customers will be using stamps. They will likely do more clicking than licking, accessing on-line reports of their cumulative purchases from you and other companies with whom you are affiliated. Smaller and smaller businesses will find it worthwhile to set up their own loyalty programs, as the information necessary to operate these programs becomes more accurate and more easily available.

We believe that one common type of switching that arises especially in information industries—the cost of finding, evaluating, and learning to use a new brand—is likely to change markedly in the near future. These search costs are being dramatically lowered for some products by the advent of the World Wide Web and more generally by the advances in information technology that are making targeted marketing easier, better, and cheaper. The Amazon Associates Program, described in Chapter 5, is a wonderful example of a loyalty program that rewards frequent referrals. We expect this sort of program to be widely imitated in the future.

These artificial loyalty programs have the prospect of converting more and more conventional markets into lock-in markets, as consumers find themselves bearing significant switching costs in the form of foregone frequent-purchaser benefits when they change brands. For the same reasons, consumer “loyalty,” as measured by the tendency of consumers to frequent one or a few suppliers rather than many, is likely to grow. Whether the industry is clothing retailing (traditional catalog or on-line), hotels, or long-distance telephone service, the companies that can structure their charges to attract and retain the lucrative heavy users will edge out their rivals, in much the same way that American Airlines gained an edge by introducing the first frequent-flier program back in 1982. Competition is likely to take the form of sophisticated information systems and targeted promotional activities as much as traditional product design and pricing. When successful, these customer loyalty programs will have the effect of reducing customers’ price sensitivity, permitting the seller to successfully charge higher list prices in order to support the costs of the awards given when customers cash in their cumulative benefits. Rivals will soon imitate any successful program you introduce; the prospect of rapid imitation puts a premium on generating some consumer lock-in early, especially for the most lucrative, highest-volume customers.

Switching costs are a hurdle separating incumbent suppliers of information systems from would-be suppliers of rival systems. Thus, companies benefit from their own customers’ switching costs, even as they must overcome the switching costs of customers they seek. When U.S. Robotics introduced Palm Pilot, it had to convince users not only to try a hand-held computer device but also to transfer data such as names and addresses from existing databases to Palm Pilot’s format. Customer-switching costs were a hurdle to be overcome. Now that Palm Pilot has proven to be a big hit, 3Com (which acquired U.S. Robotics) benefits from the costs Palm Pilot users must bear to switch to another system. 3Com’s big challenge is to continue to grow the installed base of Palm Pilot users and to leverage its installed base by selling upgrades and new products to these customers.

Leveraging Your Installed Base

Suppose you’ve successfully built up a base of customers with switching costs. The next step is to leverage your position by selling complementary products to your installed base and access to your base of customers into the future.

SELLING COMPLEMENTARY PRODUCTS. There is no getting around the need to evaluate the likely future profit stream associated with a potential new customer in determining how aggressively to seek out that customer. You must think broadly in evaluating this “future profit stream” and make every effort to maximize it to achieve competitive success. If a rival can figure out more ways to generate profitable revenue streams from a new customer, that rival will likely “outbid” you to attract that customer. The name of the game is to be creative in generating revenue streams but realistic in terms of the magnitude of customer switching costs. One of the most effective ways to win in lock-in markets is to change the game by expanding the set of complementary products beyond those offered by your rivals. In this way, you can afford to fight harder to get new customers because you will capture more business from them later on.

We noted earlier that a customer may be locked into the purchase of various “ancillary” goods or services when buying the primary product. The example of maintenance for durable equipment fits this pattern, as does the purchase of upgrades or extensions to a computer software program.

Firms compete in lock-in markets by attempting to expand the scope of these complementary products subject to lock-in. Visa and MasterCard beat American Express in the market for payment services in this way for years. Banks that were members of Visa and MasterCard could afford to give away the “primary” product, payment services, in the form of lower charges to merchants and even rewards to cardholders based on charge volume, because Visa and MasterCard were also selling a lucrative complementary product: credit card loans at very high rates of interest. American Express was slow to recognize the need to offer credit cards rather than charge cards. In part, this was because American Express was not especially skilled at evaluating the risks associated with these consumer loans, as reflected by the significant problems faced by American Express when it first offered its Optima credit card. Visa and MasterCard and their member banks were thus able to grab a large “share of wallet” by linking payment services to something they were especially good at: consumer credit.

What is the general lesson of the battle between Visa, MasterCard, and American Express? The bank associations gained huge chunks of market share from American Express because they competed very aggressively to lock customers into the primary product—payment services—in order to make sales of a highly lucrative complementary product—consumer credit. This strategy worked especially well since consumers consistently underestimate the finance charges they will incur using their credit cards; this perceptual bias drives banks to compete in the form of low monthly fees and rebates for charge volume, but less so on interest rates. Hence the high and sticky rates charged for credit card debt.

The strategy of selling complementary products or services to your installed base has the very attractive feature that it can be executed profitably and successfully while enhancing, rather than jeopardizing, the buyer relationship, and while encouraging customer entrenchment. Microsoft has done this very effectively in selling applications software to run on Windows. For information products, with their large price/marginal cost margins, all that is needed to gain significant profits is to capture a reasonable share of the business for such complementary products at market prices. Profits are not necessarily dependent on charging any sort of “monopoly” premium for these products. Nor is this strategy dependent on any lock-in with regard to these complementary products (although Microsoft also enjoys some of that with its applications products owing to the switching costs of learning new programs).

In medical imaging equipment, for example, the companies likely to win are those that can obtain follow-on revenues not only from servicing and spare parts but also from the sale of the medium itself (such as film) and from sales of other imaging equipment. As a specific instance of this, Boston Scientific seeks to sell sophisticated imaging catheters along with the hardware and software necessary to interpret these images; Boston Scientific would have trouble placing equipment if it could not rely on catheter margins to offer discounts on the equipment. The same has been true in the field of laser eye surgery, where Summit and VisX have competed to place sophisticated equipment, knowing that they will enjoy an aftermarket revenue stream from per-procedure charges collected when the equipment is used. They are effectively selling the information contained in their patents at least as much as they are selling pieces of medical equipment. In each of these examples, doctors with the highest volume of use can expect to receive the deepest discounts on their equipment.

Intuit has done well with a similar strategy for individual Quicken users. It sells not only supplies (checks and envelopes) but complementary products (tax preparation software), on-line services (shopping for insurance and mortgages on Quicken.com), and more powerful business products (QuickBooks).

Netscape is hoping to overcome the weak lock-in of its browser customers, and to extract the most value from its installed base, by selling an integrated package of complementary products, Communicator. Communicator consists of the browser, an e-mail tool, a collaboration tool, a calendar and scheduling tool, and several other components that all work together reasonably well. All are based on open standards, but Netscape has added more functionality to these applications in one way or another. Collabra, for example, is based on the tried-and-true Usenet news protocol, NNTP, but Netscape’s version displays embedded HTML as rich text, with graphics and hotlinks.

The company that can successfully offer and sell the largest collection of attractive complementary products will enjoy a tremendous advantage in the primary lock-in market, because it will be able to set more attractive terms for the primary product. In effect, the company shares some of its profit margin on related products with the customer. The happy result is that the buyer-seller relationship is no longer a zero-sum game: the buyer is happy to buy the applications software from the same company selling the hardware and/or the operating system, so long as the applications are comparable to those offered by independent firms. Indeed, the customer may well value one-stop shopping and find highly integrated products easier to purchase and use. For example, a supplier of a computer operating system may indeed enjoy economies of scope and scale, allowing for cheaper or better integration of different pieces of software than other firms can achieve.

The prospect of employing this type of “complements” strategy will intensify competition in the primary product, since it increases the value of having an installed base. However, expanding the scope of the game by offering such complements can be an unalloyed plus for a firm that already has a secure installed base. For such a firm, adding complements to its product line is a wonderful way of maximizing the value of its installed base, bringing added value to customers at the same time.

SELLING ACCESS TO YOUR INSTALLED BASE. An installed base is a terrible thing to waste. Even if you don’t have complementary products of your own that you can sell to your current customers, you can sell access to your customers to others.

America Online is doing a great job of this. In addition to developing its own content, it is selling access to its installed base to merchants and other content developers. As of August 1997, AOL had relationships with more than seventy on-line merchants. Rent for eyeball space on the AOL homepage starts at $125,000 per year, with commissions of 5 percent to 60 percent. As we mentioned in Chapter 2, the billing information AOL obtains from its customers automatically yields valuable ZIP code information, from which it can deduce customer demographics, which is very valuable data for on-line marketing.

Microsoft is making deals left and right with content developers, in some cases encouraging them to build sites with special features accessible only via Internet Explorer software. Star Trek is a case in point: several convenient features can be used only by those with the Internet Explorer browser, although the last time we looked there was an announcement saying “expanded functionality for Netscape and Macintosh users coming soon.”

This sort of cross-marketing is hardly limited to on-line services. Supermarkets have been doing it for years, offering banking and other services to give additional value to their own installed base of customers. However, it must be remembered that it is often the additional communications and record-keeping capabilities offered by information technology that have made such partnerships feasible.

SETTING DIFFERENTIAL PRICES TO ACHIEVE LOCK-IN. Suppose you are successfully building a readership for your new on-line magazine. So far, most of your money has come through advertising revenues. But you know that sooner or later you’ll have to bite the bullet and start charging for subscriptions. You’ve done some surveys of readers and looked at competitors to help you set your monthly subscription fee. You know from Chapter 2 that you want to set different prices for different types of readers. But you’re really stumped by one basic question: Who should get the better deal on a subscription, your loyal readers or the new customers you are trying to attract?

As discussed in Chapter 2, one of the great benefits of keeping track of customer information is the enhanced ability this gives you to tailor packages of products and prices to individual customers. Tracking customers’ historical purchase patterns and tailoring your offerings to these histories very much fits into this pattern. But how should your offerings vary with customers’ purchase histories?

Approach this problem in two steps. First, figure out the prices and versions you would like to offer to customers based on their historical usage patterns. Next, see how close you can get to these target offerings in light of three factors that limit what you can achieve: (1) commitments you have already made to your installed base, (2) the amount of information you have about actual and potential customers’ past purchases, and (3) customers’ ability to engage in arbitrage.

A good starting point is to divide customers into two groups: those who are currently using your product and those who are not. If you have an obvious group of close rivals, you should further divide the latter group into those who are currently using your rivals’ products and those who are not currently using any product in this category. As an example, in the cellular telephone industry, for years there were just two cellular providers in each area, one of which was owned by the local telephone company. Each carrier could divide customers into three groups: its customers, its rival’s customers, and those without cellular telephones. We’ll refer to these three groups as (a) your installed base, (b) your rival’s installed base, and (c) new customers.

How should you price to these three groups? In most cases, you’ll be tempted to charge the highest price to your own installed base, because these customers have invested in your product and because they have revealed in prior purchases the fact that they value your product highly. If users bear costs of switching from one brand to another, you should discount to your rival’s installed base to help customers overcome these switching costs. But don’t be surprised if such efforts to “poach” your rivals’ customers trigger similar attacks on your own installed base. New customers have revealed a low willingness to pay, and they should be extended discounts. These pricing rules obey the general principles for pricing that we developed in Chapter 2.

Subscriptions to information services illustrate these points nicely. Many magazines and newspapers offer special introductory rates (for the first 90 days or the first six months, for instance) to new subscribers. After all, those subscribers may be unfamiliar with the publication, and they have not demonstrated any special taste for it. To the contrary, unless they are simply uninformed about how much they would value your information service, they are likely to be marginal customers at best. Under these circumstances, special introductory offers make a great deal of sense, including deals that are better than any you would offer to regular subscribers. Discounting also makes sense to attract customers who are subscribing to rival information services: they are used to getting similar information in another format, and they may have some time to run on their subscription with the rival information service. In this case, there is no particular reason to reward loyalty: longtime subscribers have demonstrated a taste for your publication and are likely to have a high willingness to pay for it.

Beware the “burden of locked-in customers.” If you have a large base of locked-in customers, you will be tempted to set higher prices. This, of course, is the reason why you worked so hard to attract those customers in the first place. However, if you cannot find a way to offer a selective discount to customers who are new to the market, your pricing will place you at a disadvantage in attracting those customers and thus in sustaining your market share. Price discrimination in the form of selective discounts for new customers (aided by tracking customers and their purchases) is the solution to this problem.

Be sure not to neglect the customer-relations aspects of any selective discounts to new customers. Such discounts can be offered without alienating your long-standing customers, if you are careful. One approach is to remind any regular customers who complain that they are not getting the best rates that they, too, obtained special terms when they first came aboard. An “introductory offer” sounds so much nicer than a “premium price” for long-standing customers! Another approach is to rely on versioning by offering long-standing customers enhanced services or functionality. Extra information makes a great gift: it is cheap to offer, and long-standing customers are likely to place a relatively high value on enhancements. As we learned in Chapter 3, versions should be designed to accentuate the differences between groups in their tastes. Software vendors are wise to offer an easy-to-use version for new customers along with a feature-rich version for the installed base (which also encourages entrenchment by existing users).

Whenever you contemplate special offers to groups other than your own installed base, you must consider the impact on your reputation for fair dealing. Future sales are at risk if you develop a reputation for exploiting your loyal customers. This is tricky, because the line between recouping your initial investments in your installed base and “exploitation” is not sharp. But remember that any adverse reputation could have a devastating impact on future sales to new customers, especially if you face significant competition to make such sales. Thus, an important lesson is to structure—and to communicate your prices—in a manner not perceived as unfair or opportunistic by your customers.

As we noted above, three factors apart from reputation are likely to limit your ability to extract premium prices from your own installed base. First, you must honor any commitments you previously made to attract your installed base in the first place. If you established loyalty programs rewarding existing customers with discounts, you can’t very well charge them more than new customers. If you promised current customers most-favored customer treatment, you’ll have to lower prices for them if you discount to attract new business. You may well be able to avoid triggering such most-favored customer clauses by offering distinct versions to new customers, however.

Second, the tactics available to you depend on the quality of the information you have about customers’ historical purchasing behavior. This is a good reason to keep careful records of your customers’ purchases. A customer who has responded to discounts in the past has revealed price sensitivity and warrants more discounts. In contrast, there is less reason to discount to a customer who regularly buys regardless of price. Information about customers you have not yet served is also very valuable. For example, you can use customers’ prior purchase history to distinguish customers who have been using a rival brand from those who are new to the category. In the future, it should become easier for customers to prove that they have been using a rival information service or software product and, thus, to qualify for special discounts. Alternatively, we expect that such information will become cheaper to acquire from third parties, as more transactional information is tracked, to support targeted marketing efforts.

Third, you need to anticipate and block arbitrage—efforts by locked-in customers to pose as new customers (or to buy through intermediaries) to obtain any special rates extended to other groups. As discussed in Chapter 3, a good way to handle the arbitrage problem is to offer a special version of the product to new customers. Usually, this will be a stripped-down version, both because many new customers are less likely to need the full set of functions you have developed to serve your regular, long-standing customers and because the stripped-down version will be easier to learn, reducing switching costs. Once these customers are comfortable with your product and are past the sampling portion of the lock-in cycle, you can upgrade them to a version that is richer in features and easier to use, if not easier to learn. Photoshop, Adobe’s image editing tool, is a good example. As we saw in Chapter 3, a stripped-down version comes bundled with many scanners and digital cameras. This is adequate for new users, but more serious users eventually decide to upgrade to the full-featured version.

A whole new set of issues arises when selling durable products, such as computer software. If you are selling a durable product, as opposed to an information service, your customers can just keep using what you have already sold to them. In this case, you can no longer assume that your own customers have the highest willingness to pay for your product. To the contrary, they may have the lowest willingness to pay, since they own an older version.

For computer software, which does not depreciate, you are necessarily selling the improvements to the older version. Improvements are likely to be worth considerably less than the basic functionality. So, even though the customer is locked in to using your program, making it unlikely that he or she will shift to an entirely different program, you still must price the upgrade according to its incremental value to the customer. You need to give customers a good reason to upgrade, and then make the upgrade path as painless as possible.

For replacement hardware, as for software upgrades, making a new sale does serve to further entrench the buyer and lengthen the lock-in cycle, giving you a better chance to make yet further sales or to place complementary products with this customer. This is especially true if the upgrade, or the new piece of hardware, incorporates additional proprietary features that were not present in earlier versions.

ATTEMPTS TO RAISE SEARCH COSTS. As we pointed out earlier in the chapter, the Web has generally tended to reduce search costs. You should certainly take advantage of this medium to make it easy for customers to find you and to learn about your products. By the same token, you may be tempted to make it difficult for your customers to seek out alternatives and compare your offerings with those of your rivals. This is worth trying, but we think it will be hard to do on the Web.

Remember Bargain Finder from Chapter 3? Three of the eight CD stores that Bargain Finder originally searched refused to allow it access to their sites so as to make it difficult to comparison shop. This sort of strategy will not be successful in the long run. Rather than banning searches, the CD stores should focus on reducing their costs and on providing differentiated products, as in the MusicMaker example also in Chapter 3. You don’t have to worry about consumers searching for competitive products if the product you are selling is truly unique.

EXPLOITING FIRST-MOVER ADVANTAGE. First-mover advantages can be powerful and long lasting in lock-in markets, especially those in information industries where scale economies are substantial. If you can establish an installed base before the competition arrives on the scene, you may make it difficult for later entrants to achieve the scale economies necessary to compete. This is especially true in the common circumstance in which an entrant would be able to attract customers away from your installed base only gradually. This implies that your rival will be smaller than you for some time, and very likely less efficient if economies of scale are substantial.

One way to push this strategy, especially in markets with a relatively small number of key customers, is to control the length of the lock-in cycle by entering into multiyear contracts with large customers. For example, Ticketmaster has multiyear contracts with major stadiums and other venues to handle their ticketing needs, making it harder for upstart ticketing services to break into the market in any locale. Entry is made more difficult by the need to have a network of outlets in any given area where concert goers can purchase tickets. As this crude, historical method of selling tickets is displaced by new technologies, such as on-line ticket sales and e-tickets for concerts, Ticketmaster’s grip will loosen. Selling tickets electronically to young rock concert fans who lack credit cards will be the hardest hurdle for electronic ticketing services.

One way to make the most of your first-mover advantage is to consciously stagger the termination dates on contracts with different customers. By this device, any entrant would have to operate well below efficient scale for some period of time, even while fighting to attract key customers away from you. In other words, “lock-in can lead to lock-out” when customer switching costs make rival entry unattractive. In the animal world, insects such as cicadas emerge to procreate at intervals of seven, thirteen, and seventeen years, all prime numbers, making it harder for predators to enjoy “scale economies” by emerging on the same cycle. It’s a jungle out there!

Another way to control cycle length is through the frequency and timing of new versions or upgrades. Like the weakest link in a chain, you want to prevent aggregate customer lock-in at any point in time from getting too low, as that would be the optimal time for another company to enter and attack your installed base.

You may feel the presence of competition even before it arrives on the scene. If customers expect a rival of yours to introduce a new product in six-month’s time, they will be less inclined to become locked into your current product. Similarly, a competitor may well seek customer commitments before actually launching its product. You can lock up certain customers before your rivals’ plans have solidified enough for them to credibly approach your customers. This may involve making some concessions but could yield a large payoff if your rival’s product plans prove successful. However, this type of information game is double-edged: your customers will be keen to point to new choices they see emerging in negotiating a better deal from you, and they have every incentive to be informed about those choices and to play them up. Plus, dominant hardware and software suppliers, including both Microsoft and IBM, have even been accused of “predatory product preannouncements” when announcing products (“vaporware”) before they are available. (We’ll discuss vaporware as a strategy for expectations management in Chapter 9.)

CONTROLLING CYCLE LENGTH. You can influence the duration of the lock-in cycle. Cycle length depends on such factors as the duration of contractual commitments, the lifetime of durable equipment, the presence of complementary products with different economic lifetimes that work together, the aggressiveness of outside suppliers and their tactics in approaching locked-in customers, the information outsiders have about the extent and timing of lock-in by various customers, and the frequency with which customers choose to bear the costs of putting their business up to bid.

You might think that your customers will try to keep the cycle short, while you will be pushing for a long cycle. This is not always so. American Airlines was content to sign a long-term contract with Boeing both because of American’s desire to simplify fleet maintenance and because of the price protections it obtained in the contract. Indeed, if lock-in is long-lived, the customer may well insist on contractual protections of similar duration. Indeed, as a seller you might be happy with a relatively short-term contract, if buyers are locked in for a long period of time. This pattern will leave you in a strong position when the contract expires.

Consider employing the popular tactic of truncating the lock-in cycle by getting customers to sign a new, multi-year contract well before their current contracts expire. Likewise, consider selling new equipment or an upgrade to customers before their existing equipment wears out or the upgrade is really needed. Premature renewals are certainly common in real estate transactions, in part because both landlord and tenant need to know in advance if the tenant is moving. But even when planning needs are not nearly so significant, getting the jump on contract renewal or system replacement can work well for you as a seller of information or information systems. By preempting contract termination, you can negotiate with a customer who is still attached for some time, making it less likely that a rival will be knocking at the door and engaging in serious discussions. For these very same reasons, savvy buyers will be wary of renewing a contract without going through the exercise of getting a competitive bid.

LESSONS

Consumer lock-in to specific technologies, and even to specific brands, is an ever-present feature of the information economy. Both buyers and sellers have much to gain from evaluating the consequences of their actions over the entire lock-in cycle. Short-sightedness can be extremely costly when switching costs are involved.

We have three basic lessons for purchasers of information systems and technology:

• Bargain hard before you are locked in for concessions in exchange for putting yourself in a vulnerable position. If you can’t avoid lock-in, at least get paid a sweetener up front to compensate you for becoming locked in.
• Pursue strategies like second sourcing and open systems to minimize the extent of your lock-in. Even if you must make investments in a particular technology, you can still plan ahead to avoid becoming beholden to a single supplier.
• Look ahead to the next time you’ll be picking a vendor, and take steps at the outset to improve your bargaining position at that time. Retain information on your relationship with the seller, such as maintenance records, and use patterns that could reduce costs if you have to switch to a new supplier. These will be valuable assets if you decide to break off your relationship.

We also explored a number of strategies for sellers whose customers will experience lock-in. Our key points are these:

• Be prepared to invest to build an installed base through promotions and by offering up-front discounts. You can’t succeed in competitive lock-in markets without making these investments.
• Cultivate influential buyers and buyers with high switching costs. These are your most profitable customers.
• Design your products and your pricing to get your customers to invest in your technology, thereby raising their own switching costs. Employ a loyalty program to make your product attractive to your customers at their next brand selection point. This requires keeping records of customers’ cumulative purchases.
• Maximize the value of your installed base by selling your customers complementary products and by selling access to your installed base. An installed base is a wonderful springboard for marketing new products, especially because of your access to information about customers’ historical purchases that you have gathered over time.