When consumers stop buying a brand for a particular period of time, middlemen tend to

Reprint: R1406F

Almost every retailer looks to Google to refer customers, and it’s rare to find a manufacturer whose products aren’t sold on Amazon. But these and other big platforms can capture a disproportionate share of the value a company creates: Buy an app on iTunes, and Apple takes 30%. The author presents four strategies to help businesses reduce their dependence on powerful platforms.

Exploit the platform’s need to be comprehensive.

American Airlines’ strong coverage of key routes made its presence on the travel website Kayak indispensable to Kayak’s value proposition. As a result, AA negotiated a better deal.

Identify and discredit discrimination.

Public complaints that eBay was giving search prominence to suppliers who advertised on the site forced a reversal of the policy.

Create an alternative platform.

When MovieTickets was on the verge of dominating phone and online ticketing, Regal Entertainment and two other large theater chains formed Fandango.

Deal more directly.

People ordering takeout through online platforms like Foodler and GrubHub have often already chosen their restaurant. Restaurants that deal directly can exit the platform.

• Tweet

• Post

• Share

• Save

• Get PDF

• Buy Copies

• Print

In our increasingly digital world, businesses and consumers have become starkly more dependent on a number of powerful platforms. Of course, platforms aren’t new. For three decades, airlines have been relying on computerized reservation systems to reach travel agents and key customers. But platform dependence is now ubiquitous. Almost every retailer looks to Google to refer customers, and it’s rare to find a manufacturer whose products are not sold on Amazon.

In many ways, this is a good thing. Intermediaries often provide valuable benefits: They make it easier for buyers to find what they need, they help set standards, and they enable comparison shopping—efficiency improvements that keep markets working smoothly. But they can also capture a disproportionate share of the value a company creates. Restaurants, for example, typically pay 15% of each order to be listed in an online portal such as Foodler or GrubHub. Because net margins in the hospitality industry are often in the single digits, that’s a significant expense. What’s more, those costs don’t dissipate through competition, because most markets settle on just one or two dominant platforms. The economics in other industries are similar. (See the sidebar “The Value Extracted by Powerful Platforms” for some examples.)

Google’s well-known ad auction yields revenue exceeding $60 billion a year, and the costs to advertisers are commensurately high. Consider a company that sells basic web hosting, charging perhaps $150 a year for the service. To find one new customer using Google, the company might buy 100 clicks at about $1.80 each—spending $180 to get one customer. So the company doesn’t even begin to cover its advertising expense until the customer renews.

Airline reservation systems appear free to most consumers, because the prices at airline websites and at online travel agencies are usually identical. But those reservation systems impose substantial costs on airlines—roughly $3 per flight segment per passenger. A typical U.S. domestic connecting round trip entails four flight segments, or $12—about half an airline’s per-passenger cost for aircraft lease or depreciation—and those costs are built in to ticket prices.

Online marketplace platform fees are about 10 times credit card fees (which retailers routinely gripe about). Buy an app at the iTunes Store, and Apple takes 30%. Book a car through Uber or Lyft, and the service keeps 20%. Of course, sellers benefit from streamlined services; for example, Uber and Lyft find and dispatch drivers. And the mobile app concept might not even exist were it not for Apple’s efforts. Still, platform fees are the largest single expense to most sellers.

Most companies feel that they have no choice but to put up with intermediaries and their rules and fees. They are mistaken. Platform owners are far from invincible, and savvy suppliers have options for recapturing value or at least protecting themselves from abuse. In the following pages I present four strategies to help businesses reduce their dependence on power ful platforms.

1. Exploit the Platform’s Need for Completeness

Most intermediaries wield a surprisingly simple threat: If a business doesn’t accede to their terms and fees, they will exclude it from their services. That’s powerful: Advertisers compare the prospect of disappearing from Google to a death sentence.

But not all threats of exclusion are credible. Consider the launch of the travel search engine Kayak, in 2004. From the outset Kayak told users it would offer a “comprehensive, objective search” that included airlines not listed by standard online travel agencies such as Expedia and Orbitz. This approach garnered early praise for Kayak’s offering.

The new platform’s threat was obvious. If Kayak became too powerful, it could present airlines with a Hobson’s choice: pay high listing fees or watch Kayak refer passengers to competitors. But American Airlines realized that Kayak had its own vulnerability. It had promised to show users comprehensive results, and in many markets American was a dominant force, offering the most flights on key routes such as New York to Los Angeles and New York to London. To be credible to users in those cities, Kayak had to include American flights—indeed, Kayak needed American even more than American needed Kayak—so American was able to negotiate superior terms. For example, Kayak committed to link American flights only to American’s website and not through sites like Expedia and Orbitz. (A direct link to AA.com reduced American’s costs.) Furthermore, Kayak had to give AA flights fair prominence by objective criteria. By all indications, American didn’t pay Kayak a penny. It was a great agreement for the airline. Most other airlines agreed to pay Kayak for the users it refers and never considered requiring it to link directly to their own sites.

Real estate provides another good example of platform vulnerability. In most cities, agents have low market concentration: Sole proprietors remain viable, and midsize brokerages are widespread. One might expect a few powerful online platforms to extract high fees from real estate professionals. But platforms need to list all properties on the market; a real estate portal with incomplete listings is much less valuable to house hunters. What’s more, they cannot simply copy listing details from other sources: Some facts may be in the public domain (such as location and size) or noncopyrightable (such as asking price and days on the market), but reproducing photos of a property is understood to require permission from the agent that is marketing that property.

As a result, real estate websites have found that they must provide agents with significant value to induce them to join. For example, Zillow not only offers property listings without charge but also prominently names the agent marketing the property. And any agent who joins can receive messages directly from interested viewers. One can imagine Zillow’s charging hundreds of dollars per listing for these services (and some agents might be willing to pay). But given the platform’s need for completeness, the agents have the upper hand.

2. Identify and Discredit Discrimination

Imagine the backlash if it turned out that Amazon deleted negative reviews of AmazonBasics, its house brand products, or if the iTunes Store favored app makers who used Apple advertising rather than a competitor’s service. Competitors and consumers would rightly cry foul. The threat of such complaints prevents platforms from overtly favoring their own services.

But potential backlash doesn’t stop them from trying to discriminate against competitors in less obvious ways. In 2009 Google sought to acquire Yelp, but the deal fell through when the companies could not agree on price. Shortly thereafter, despite years of having been among the sites most often cited for restaurant searches, Yelp began appearing less frequently in Google results, while Google Local listings suddenly took prominent positions.

Yelp suspected that Google had fiddled with its search algorithms in order to promote its own review services. Yelp CEO Jeremy Stoppelman claimed in 2011 Senate testimony that Google “always presents links to its own consumer review website in the most prominent position regardless of whether…it has the most relevant content.” Data from the web measurement service comScore confirmed his point. During 2010 some users apparently noticed that Google wasn’t often linking to Yelp: Instances of their adding “Yelp” to Google searches increased by 50%. (If a search for the restaurant Rialto didn’t yield the Yelp link the user wanted, he or she might try “Rialto Yelp.”) Notably, users changed their searches only on Google, not on Bing and Yahoo (sites that didn’t discriminate against Yelp), which reinforced Stoppelman’s concern. Under pressure from Yelp’s complaint, regulatory inquiries, and possible user backlash, Google scaled back the changes.

Getting your customers on board is essential when making any charge of discrimination—indeed, in some cases the complaints originate with them. Beginning in 2008, eBay’s AdCommerce and Featured First programs let sellers pay to appear at the top of search results. These programs were popular with sellers because greater visibility brought them more bids and sales. But users reported finding the site harder to use, especially compared with Amazon and other simpler, instant-purchase sites. Arbitrary sorting and unwanted listings were also mentioned in users’ complaints, and focus groups confirmed that eBay’s site felt cluttered. Losing users was worse than losing advertising revenue, so in 2010 eBay ended those programs, citing the need “to keep the focus…on surfacing the items most relevant to a buyer’s search.”

Platform providers do push back on the idea that platforms should be compelled to treat all suppliers equally. Building a platform can be expensive. Constructing the famous SABRE ticket reservation system in the 1960s cost American Airlines about as much as a dozen 747s, so AA felt that it should be free to configure the system as it saw fit. The debate continues: The UCLA law professor Eugene Volokh suggested in a 2012 white paper that the First Amendment prohibits regulators from interfering with Google’s decision about where to link, and the onetime Supreme Court nominee Robert Bork, writing in 2012 with the Georgetown professor Gregory Sidak, questioned whether anyone is truly harmed when Google gives its own services top placement. (Google commissioned both papers.)

Platform providers usually get away with relatively subtle discrimination as long as consumers don’t notice or care. Nonetheless, public outcry and regulatory complaints provide an important bulwark against brazen instances of intermediaries’ favoring their own services. It’s certainly worth a company’s time to explore whether a suspected case of discrimination could become the focus of public concern.

3. Support or Create an Alternative Platform

In principle, competition among platforms can help improve suppliers’ position relative to them. When multiple platforms compete, sellers typically find it easier to get improved terms. For example, a dissatisfied seller could forgo a large platform that charges high fees in favor of several smaller ones that collectively may reach just as many users.

It also turns out that a platform’s scale and reach are often less stable than they seem, and with the right partners a savvy supplier may be able to introduce some significant competition. Take the case of Regal Entertainment, the largest movie theater group in the United States. In the early 2000s Regal was threatened by MovieTickets, which seemed to be on the verge of dominating phone and online ticketing services. In response, Regal formed Fandango in partnership with United Artists and Hoyts, other large theater chains. Their collaboration blocked MovieTickets’ expansion—indeed, Fandango ultimately surpassed MovieTickets in size.

Further Reading

• 
  Strategies for Two-Sided Markets

  Economics Feature

  • Thomas R. Eisenmann, Geoffrey G. Parker, and Marshall W. Van Alstyne

  What newspapers, HMOs, and computer operating systems have in common, and how to manage changing value chains.

  • Save

We’re seeing similar developments in the hotel business. In 2012 six large chains founded a search service called Room Key. Like online travel agents, Room Key aspires to provide comprehensive results; but rather than charging hotels a commission on each booking, it sends consumers directly to the hotels’ own sites to make reservations. Hotels may buy advertising from Room Key to obtain more-prominent placement, but its costs remain lower than those of other distribution channels, and Room Key says that hotels pass the savings on to consumers through greater flexibility as well as loyalty program benefits.

The big risk in this kind of initiative, of course, is that moving to a new platform may mean leaving users behind on the established one. The main hotel chains are still using established platforms to retain access to their long-standing customers, which means those users need not switch to Room Key. Successfully launching a new platform probably requires a deeper—and riskier—commitment than this.

The danger of losing customers is not the only serious challenge. A seller starting its own platform will surely include its inventory in the platform’s launch; but to be robust, the platform will need competitors’ offerings as well. Joining forces with competitors inevitably raises antitrust issues, and all the well-known challenges of joint venturing are exacerbated when your partners are also your competitors.

4. Deal More Directly

Many consumers buy through a platform not because it’s easy but because the seller offers no way to buy directly. With the right incentives and some investment, a direct channel can displace the platform provider for at least some consumers and make it less likely to exploit its position.

Consider a customer who is ordering food for takeout or delivery. These days many restaurants allow online orders through portals such as Foodler, GrubHub, and Seamless. If the customer wants the convenience and accuracy of online ordering, these portals are often the only options. But as we’ve seen, they charge restaurants a stiff fee. They also insist on the right to market competing offerings to the restaurants’ customers. In effect, a successful restaurant on an ordering portal is handing its e-mail list to its competitors. What’s more, once customers have used a given portal, they tend to stick with it, so restaurants—like hotels—have no choice but to stay with the portal if they want to keep their customers. All this might be easier to bear if portals delivered lots of new customers, but the fact is that customers who are using them have often already chosen their restaurant.

Although many restaurants don’t realize it, they don’t need established portals to offer online ordering. The required functions—viewing a menu, choosing items, sending the order to the restaurant by e-mail or fax, and processing payment—are well within the IT capabilities of even a small company. And some software-as-a-service firms provide those functions on a stand-alone basis, which lets a restaurant offer the convenience of online ordering at a much lower price. For instance, iMenu360 charges as little as $20 a month, with no fee on customers’ orders.

If they see it’s available, some customers will switch to direct purchase as soon as possible. But if a restaurant wants to shift away from an ordering portal, it should offer lower prices on its own site. Portals try to prevent this by requiring that the prices shown on them match what restaurants charge for direct ordering—but restaurants can get around that with coupons, freebies, and other special benefits for those who order directly.

Platforms rarely roll over when companies seek to shift the balance of power. Here are some of the ways they defend their turf:

Search engines show ads for myriad retailers. If one complains, a natural response is to demote its ads or remove them altogether. Arbitrary treatment may smack of retaliation, but that’s hard to prove; for all the advertiser knows, the change was purely coincidental. And some platforms have successfully argued that they have the right to remove unwanted listings for any reason or for no reason.

Seeking to use Yelp’s content, Google insisted in 2010 that if Yelp wanted its listings to appear anywhere in Google’s search results, it had to allow Google to copy the material into its own local services. Because Yelp depended on Google search traffic to reach users, it acceded.

If companies sign extended agreements with an intermediary, its immediate future is secure. And by structuring contracts to avoid any single day of reckoning, the intermediary can prevent a group of companies from recognizing their mutual interest in finding a cheaper alternative. In the past, major airlines’ five-year contracts all came up for renewal at the same time. If one airline threatened to leave a system that charged high fees, it knew others could soon follow. The systems later adjusted contract lengths to separate renewals and avoid this vulnerability.

Many platforms require sellers to charge the same prices whether customers buy through the platform or directly. Similarly, credit card networks prohibited credit card surcharges for decades, and many state laws continue to enforce this rule. With the assurance that prices are identical no matter where the customer buys, the platform need only add a little bit of value in order to attract customers—potentially much less than what it charges sellers.

These strategies combine economics, law, and public relations, and sometimes software design. To implement—or combat—them, you need an agile interdisciplinary team.

Companies can use a similar approach to reduce dependence on search engine advertising. When an advertiser reaches a customer via Google, advertising fees often amount to 10% of revenue. These costs obviously affect a company’s ability to offer customers lower prices. Savvy retailers need a way to get users to come to their sites without passing through high-cost search advertising. Online rebate sites are one possibility: By offering a modest rebate at, say, FatWallet, a retailer can induce users to start a shopping session there rather than on Google. That lets the retailer avoid search ad expenses, and the discount typically prompts some additional consumer spending. In this case, of course, the retailer has traded one intermediary for another. But FatWallet has numerous competitors, so the payments it gets from retailers are almost all passed along to consumers. And because rebates tend to be a fraction of search ad costs, this approach helps cut the retailers’ net expenses.

In general, powerful platforms seek to appear free to consumers—an approach the National University of Singapore professor Julian Wright and I call “price coherence.” If suppliers pay a platform’s fees, then customers perceive the platform to be free and, accordingly, choose to use it even if it offers minimal benefits. (After all, a customer who forgoes the platform loses those benefits and gets no savings.) Yet the platform’s fees may be far higher than the value the customer actually places on the platform’s benefits, making the platform an inefficient choice for both the customer and the supplier. Meanwhile, the bigger the platform’s user network, the higher the supplier’s costs, which are ultimately passed on to customers through higher prices.

The bottom line is that if customers can see that sellers offer an alternative pathway to purchase at a lower price than the apparently free but actually expensive platform, they’ll deal directly with the seller or go through a less costly platform.Powerful online platforms have important advantages in their dealings with sellers—not just size but sophistication, pricing structure, and user behavior. Thus they can make a modest investment yet enjoy profits disproportionate to those of suppliers, who hold inventory, produce products, and actually do the work. But not all is lost. Many platforms need to be comprehensive, so they must retain even small sellers. Meanwhile, when a platform rewards favored partners and penalizes others, it risks both user displeasure and regulatory concern. And with planning, sophisticated sellers can use ever-cheaper information technology to let customers buy directly. That’s not to say it will be easy: Powerful platforms have every reason to facilitate and preserve sellers’ dependence. But it’s worth the effort.

A version of this article appeared in the June 2014 issue of Harvard Business Review.

When companies use foreign countries middlemen?

Which stage of the new product development process is usually the most expensive for new products?

What is the name for the collection of activities in which businesses sell things to buyers?

What subject area should be on the checklist of criteria for evaluating middleman servicing a market?