Switching costs are another aspect of a startup’s business model that I pay attention to. Together with Network Effect, Intangibles, Cost Advantages, and Efficient Scale they form the source of economic moats. ((Any errors in appropriately citing my sources are entirely mine. Let me know what you object to, and how I might fix the problem. Any data in this post is only as reliable as the sources from which I obtained them.)) In this post I will discuss switching costs; what they are, how they develop and evolve, and how switching costs can help or hurt a startup.
To ensure we are on the same page, I will start with some definitions. In the rest of this discussion I am primarily focused on early stage technology startups. Also, the customer I have in mind is one whose present known needs are adequately served by the current product. Finally, I assume government intervention is not a significant factor.
Definition #1: What is a startup? A startup is a temporary organization built to search for the solution to a problem, and in the process to find a repeatable, scalable and profitable business model that is designed for incredibly fast growth. The defining characteristic of a startup is that of experimentation – in order to have a chance of survival every startup has to be good at performing the experiments that are necessary for the discovery of a successful business model. ((I am paraphrasing Steve Blank and Bob Dorf, and the definition they provide in their book The Startup Owner’s Manual: The Step-by-Step Guide for Building a Great Company. I have modified their definition with an element from a discussion in which Paul Graham, founder of Y Combinator discusses the startups that Y Combinator supports.))
Definition #2: What is an economic moat? An economic moat is a structural barrier that protects a company from competition. ((Heather Brilliant, Elizabeth Collins, et al. Why Moats Matter: The Morningstar Approach to Stock Investing. Wiley. Hoboken, NJ. 2014; p. 1))
Definition #3: What are switching costs? Switching costs refer to the expense in cash, time, convenience, risk, and process disruption that a customer of one product or service must incur if they change from one product from an incumbent Producer A to another product from Producer B. Switching costs can be explicit or implicit, and confer the benefit of customer lock-in to incumbent suppliers if the customer perceives the cost of switching to outweigh the benefits that would be obtained by making the switch. ((In economics switching costs are defined as the disutility that a customer experiences in switching between products.))
How do switching costs develop?
Switching costs develop and become stronger when an incumbent product becomes “mission-critical” for the purpose for which the customer acquired the product in the first place. An incumbent that combines network effects with high switching costs in the same product line is well positioned to build a durable moat around its business. Economists describe at least three main assumptions about switching costs. Exogenous switching costs are believed to evolve without any intentional influence from the incumbent producer – for example; customers independently create or enhance switching costs by becoming more skilled and experienced in applying and adapting the incumbent product towards solving a wider variety of problems than the producer had originally envisioned. Endogenous switching costs evolve through deliberate actions by the incumbent – for example; volume discounts to encourage wide adoption within a company of a new software product, coupled with long-lived license agreements and punitive charges if the license is terminated between license renewal dates. Also, deeply entrenched incumbents will typically opt for incompatibility with competing products while new entrants will prefer to build in compatibility with the incumbent product that they seek to displace. Lastly, switching costs are symmetrical between all the producers competing within a given market. ((Pei-yu Chen and Lorin M. Hitt. Information Technology and Switching Costs. September 2005; p.9. Accessed online on Oct. 19, 2014.))
What are the types of switching costs that lead to buyer lock-in?
One might consider switching costs to exist along a continuum that is characterised most distinctly by how intertwined each of the categories identified by economists is tightly intertwined with nearly every other category below. ((The following discussion is based largely on; Paul Klemperer. Competition When Consumers Have Switching Costs: An Overview With Applications to Industrial Organization, Macroeconomics, and International Trade. Review of Economic Studies, 1995; p. 515 – 539. Accessed online on Oct. 19, 2014.))
Compatibility Requirements make it difficult and expensive to switch between products. Consider an individual or organization running MS Windows contemplating a decision to switch to Linux. The implications of this choice are generally non-trivial. Compatibility requirements are largely an implicit cost that is borne by the customer.
Transaction Costs impose an explicit cost on customers who decide to switch from one product to another. For example, cable-tv subscription agreements typically impose a high penalty on subscribers who decide to terminate their agreement before it has run its full course. Any cost than can be measured explicitly and that has to be considered in switching between products falls under this category.
Cognitive Costs are the perceived hurdles customers feel they will have to overcome when they switch from one product to another. One example is the dichotomy between people who prefer Mac OS and those who prefer MS Windows. I understand the practical reasons one might have for preferring one operating system over the other; for example, one platform is more compatible with a wider variety of products in a certain category of software applications than the other. What often surprises me is the speed with which conversations between those two groups quickly devolve past anything one might consider rational, logical or practical to become an exercise in name-calling and ad-hominem attacks. Such episodes suggest that in some situations there are significant psychological issues at play that have nothing to do with the reality one might face if one tried to switch products. ((Klemperer calls these psychological costs.))
Uncertainty is the apprehension that the customer has to face regarding the quality of the new product. Uncertainty is minimized only if the customer believes that, at a minimum, the new product will match the old product in quality. As an example, consider a small business that is trying to decide if it should migrate from MS Exchange Server to Google Apps for Business at a time when its license for the former is up for renewal. Uncertainty works in favor of the incumbent product when customers have very little information about the relative performance characteristics of the new product. ((Ibid. Pei-yu Chen and Lorin M. Hitt. p. 4.))
Learning Costs are the known hurdles that a customer must overcome in order to attain mastery of the new product that is at par with that customer’s mastery of the incumbent product required to accomplish the tasks the customer needs to complete. Learning costs need to be considered on their own, independent of other categories of switching costs. High learning costs tied to adopting a new product increase switching costs in favor of the incumbent. Minimal learning costs tied to the adoption of a new product lower switching costs in favor of the new product. On one hand, an individual customer might be willing to face high learning costs in situations where the consequence if things go wrong is non-fatal; for example switching from one messaging app to another. On the other hand, enterprise or small business customers who face loss of business and revenue if things go wrong will exhibit high levels of inertia in the face of high learning costs; for example switching from one company-wide CRM system to another.
Lost-Benefit Costs are costs suffered by the customer because certain benefits that have been earned but not yet consumed by the customer as a result of its historical relationship with the incumbent are non-transferrable in nature – the customer who decides to make a switch suffers a significant loss and must start to earn such benefits from scratch with the new provider. An example of this is found in the various loyalty programs that are used to induce customers from switching from one product to another; airline travel points, for instance. Mobile phone subscription roll-over minutes are another example – my roll-over minutes accumulated on AT&T’s cell phone network are not transferrable to another carrier if I choose to switch. ((Klemperer calls these discount coupons and other devices.))
How might switching costs become a disadvantage?
Switching costs lock in customers who face the highest opportunity costs of switching from an incumbent product to another product offered by a rival. It is often the case that these customers comprise the most profitable segment of customers for the incumbent, and it is not uncommon for the incumbent to continue optimizing the product to meet their requirements, with each improvement in the incumbent product being reflected in an across-the-board increase in prices for all the incumbent producer’s customers. This iterative cycle of feature upgrades and attendant price increases will continue until it gets to a point at which the following things happen; first the product becomes too advanced for a large number of customers with “only moderate” needs and therefore, commensurately moderate switching costs. Second, at this point the medium- to long-term cost of switching is perceived by this group of customers to be less than the commensurate benefit of remaining locked into the incumbent product. Last, the rival product has matured such that it satisfactorily meets the needs of customers considered to be low-margin customers based on the business model implemented by the incumbent producer. However, these same customers comprise an attractive, high-margin customer cohort for the rival product because the rival producer is pursuing a business model that features significantly lower overhead costs than the comparable costs reflected in the incumbent’s existing business model. ((This is the process described by Clayton Christensen in The Innovator’s Dilemma.))
As a result the incumbent producer faces a dilemma; stay and fight for low-margin customers, or cede that ground to the rival product? The most typical response from incumbents is to cede the unprofitable customers to the rival. This gives the rival a toe-hold in the market, a position from which the rival can gradually strengthen its position and eventually migrate upstream until it poses a direct and powerful threat to the incumbent. The effect high customer switching costs have on an incumbent producer is that they lock the incumbent into a pattern of sustaining innovation. Sustaining innovation improves on already existing products, and focuses on squeezing more out of a large base of existing and and a comparatively small base of new customers. An example of sustaining innovation is Microsoft’s line of Windows and Office products; annual sales to new customers is small compared to sales generated from the installed base of Windows and Office users. Disruptive innovation seeks to satisfy non-consumption by developing products with features so simple and inexpensive in comparison to the status quo that a disproportionately large number of new customers enter the market. ((Or, a large number of “unprofitable customers” abandon the incumbent product for the new, disruptive product.)) The key is that the customers that flock to the disruptive product are very unattractive to established incumbents. With time, the disruptive innovation matures to the extent that it becomes a viable substitute for the incumbent’s most profitable customers at a price point that is extremely hard for them to resist. It is at this tipping point that the incumbent’s fight for its survival begins. It is easy to dismiss disruptive innovations at the outset because the performance measurements that have become customary for the market in question do not apply in the same way for the new wave of consumption that the disruptive innovation enables. For example, consider an investor trying to decide if an investment in Facebook was a good idea in 2006. On the basis of CPMs this investor would probably have decided to pass on the opportunity to invest in Facebook, reasoning that it did not have the qualities necessary to build a highly valuable business. Except, CPMs were the wrong metric by which to judge Facebook at that point. ((Read: Andrew Chen. Why I Doubted Facebook Could Build A Billion Dollar Business, and What I learned From Being Horribly Wrong. Accessed online on Oct. 19, 2014.))
What are the competitive strategies at play in markets in which switching costs matter.
It is important for technology startups and early stage technology startups to understand the dynamic that might evolve as they seek entry into a market characterized by an incumbent who benefits from customer lock-in. Fortunately, substantial economic research exists on that topic. ((Joseph Farrell, Carl Shapiro. Dynamic Competition With Switching Costs. RAND Journal of Economics; Vol. 19, No. 1, Spring 1988. and Joseph Farrell, Paul Klemperer. Coordination and Lock-in: Competition With Switching Costs and Network Effects; Handbook of Industrial Organization, Volume 3. Ed. M. Armstrong, R. Porter. Copyright 2007, Elsevier B.V. Accessed online on Oct. 23, 2014.))
The incumbent sells to existing customers, rival new-entrant serves new buyers. This happens in markets that are relatively mature. The incumbent focuses its efforts on its existing customer base, with growth in revenues arising from endogenous growth within that customer base – e.g. Bloomberg revenues increasing because the average number of employees of each of its existing customers is increasing with time. New entrants meanwhile utilize new technology to serve new customers, initially ignoring the incumbents existing customer base. This is especially true in markets in which the incumbent producer has a high level of power relative to customers in that market – typified by dominant market share, giving it pricing power over its existing customer base. To use the parlance of Farrell and Shapiro (1998) the incumbent sells to the oldsters while the entrant sells to the youngsters. ((See for example; Aaron Timms. The Race To Topple Bloomberg; Institutional Investor, Jan. 30, 2014 and Startups Estimize and Kensho Take Aim at Bloomberg; Institutional Investor, Jan. 30, 2014.))
The incumbent excludes the new entrant. This happens when the incumbent’s fixed costs per customer are greater than the switching costs per customer. The strategy works under conditions in which the incumbent is in a position to set a price that makes it unattractive for any new entrant to enter the market. Where this is not possible the incumbent will choose to set a price that allows the market to be shared between the incumbent and the new entrant. This is why freemium business models are so powerful, especially when a freemium business model is coupled with a product that embodies network effects and switching costs. For example, think about how dominant Facebook has become because it gives its product away to users for free. Clearly, it not possible to compete with Facebook on the basis of the price users pay in exchange for the value they derive from it. The startups that will ultimately compete with Facebook do not have a cost-leadership strategy available to them, and so must instead seek an alternate path. ((Examples; Whatsapp, Instagram, Pinterest, Snapchat, Line, Kik etc. Most recently Ello has tried to carve a niche for itself by emphasizing privacy. It is too early to tell if that tactic will work.))
New customers are won with bargains, then they are “ripped off”. This happens when customers are offered low “introductory offers” in order to entice them to adopt a product. Prices increase once lock-in has been established. As an example, consider a product that is free up to a certain usage threshold but for which continued use beyond the set threshold requires customers to pay. In this scenario, various mechanisms might be used to ensure the onset of customer lock-in, and improvements in the product’s features and capabilities are designed to nudge users over the threshold beyond which they have to become paying customers. ((Examples; Google Apps for Business, now renamed Google Apps for Work started used this tactic to build a beach-head in a market dominated by Microsoft. This scenario excludes predatory pricing practices.)) This tactic is common among cable TV and satellite TV providers, and also among internet service providers.
Customers are paid to switch. Consider three segments of an incumbent producer’s customers; Existing locked-in customers, unattached or new customers, and customers locked into a rival. In this situation, rival producers will implement price discrimination. Existing locked-in customers get one set of prices, new or unattached customers get another set of prices, while customers locked into rivals are paid to switch. ((Ibid; Farrell and Klemperer.)) Recent reports of the battle for market share between Uber and Lyft are a great example of this tactic being applied in the real world. ((See for example; Alison Grisworld, Uber Rival Gett is Making a Risky, Clever Play in The Ride-Sharing Game, Oct 15, 2014. and Avi Asher-Schapiro, Is Uber’s Business Model Screwing Its Workers?, Oct 1, 2014.)) This tactic is common with cellular phone service providers and credit card issuers.
A portfolio of products is bundled together in order to increase total switching costs. This tactic is especially effective because in order to make a switch, the customer must deal with nearly all the switching costs we have previously considered at the same time and it works especially when the incumbent producer offers a product line that is so broad that most customers simply deal with the incumbent as their single supplier for the entire line of products that they use. ((Ibid; Farrell and Klemperer.)) For example, Microsoft’s strategy of giving away Internet Explorer in a bundle with Microsoft Windows reportedly led to the demise of Netscape Navigator. I would guess that beyond merely bundling Explorer with Windows, Microsoft built-in a number of features that made Navigator less compatible with the Windows operating system than Explorer. ((Wikipedia; United States v. Microsoft Corp. Accessed online, Oct 23rd, 2014.))
Switching costs play an important role in retaining customers, and motivating repeat purchases in the future. Technology startups can’t survive without user lock-in and incumbent suppliers with strong customer lock-in typically earn monopoly profits. Early stage startups thinking about spend some time understanding the features that create value for the customer while building customer lock-in for the startup early in product design process. The existence, or lack thereof, of switching costs amongst the incumbent’s customers will play an important role in determining the competitive response that is likely to occur once the new-entrant’s intentions become undeniable. In which case speed of market entry is critical for the new-entrant. In a market with low switching costs, one might expect vicious price wars to ensue. Generally, such price wars will always favor the presumably better capitalised incumbent. Moreover, price wars are a bad idea for the incumbent as well as the new entrants. In a market where the incumbent enjoys significant customer lock-in with ensuing monopoly profits, one generally expects new entrants to find a foothold from which they can eventually migrate up-market.