This lesson is from Digital Transformation by Indian School of Business

Online Business Models

• especially relevant to digital goods
• At the heart of these business models is the concept of network effects.
  • Network effects are also known as demand-side economies of skill.
  • And they are a very important phenomenon in the information economy.
  • It is argued that while the all industrial economy is driven by economies of skill, the new information economy is driven by economies of networks.
• Platform mediated business networks
  • leverage demand side economies of scale and are ubiquitous to the world of digital goods.

Monster.com

• two distinct user groups that affiliate with Monster
  • Job seekers who are looking for jobs across diverse organizations.
  • Job recruiters who are looking to the platform to hire individuals for open positions in their firms and institutions.
• Advantages to job seekers:
  • access to a broad network of job recruiters,
  • low job search cost
  • You can use a variety of platform features to identify a set of jobs that best aligns with your needs.
  • And it’s free.
• Advantages to job recruiters
  • Recruiters can include more info than in print ads,
  • the online ads include sort and search functions that reduce your candidate search cost much like they reduce search cost for the candidates themselves.
  • You receive responses a few minutes within posting the ad and the online ads are less expensive, priced at one-third the cost of the print ads or less.

• Therefore, Monster has two distinct user groups who require very different functionalities from it and use it to interact with each other.

• What does Monster do in all of this?
  • They provide the platform that is the technology required for these two groups of people to interact with each other.
  • They also established the rules of engagement on this platform. How many days can a job posting be active?
  • Procedures for dispute resolution between groups and a whole host of other protocol.
  • Therefore, Monster provides the technology and rules of engagement for the two user groups, job seekers and job recruiters to interact with each other.
• Do you the job seekers have a preference for the number of recruiters who affiliate with the Monster platform?
  • Yes, the value of Monster increases to you with the number of recruiters who affiliate with it.
• Similarly, do you the job recruiters have a preference for the number of seekers who affiliate with the platform?
  • Yes, the value of Monster increases to you the recruiter with the number of job seekers who affiliate with it.

• And this is the basic structure of a platform mediated business network.

• Basic elements of two-sided platform mediated business networks.
  • The very first element of the network is two distinct user groups who require very different functionalities from the network and use it to interact with each other.
    • In the case of Monster, these were the job seekers and the job recruiters.
    • Keep in mind that the two user groups must require distinct functionalities from the platform for the platform to be a two-sided business network.
    • For example, consider credit card networks with customers and merchants as two sides.
    • What do the customers require? They need a unique account, they need a card, they need access to customer service, a monthly bill.
    • Very different from merchants who require from the credit card company, terminals for authorizing transactions, procedures for submitting charges, and receiving payments, and so on.
    • Similarly as we discussed in the case of monster, on one side, its users are firms that have job openings and are looking for hires, on the other side, its users are individuals who have skills and are looking for jobs.
    • These are all two sided networks.
    • On the other hand, think about a phone network with callers and receivers.
    • Is this a two-sided network?
      • No, because the roles of callers and receivers are interchangeable and converge to the single need to communicate with one another.
    • Similarly Facebook originated as a one-sided network with users who were looking to socialize with one another.
    • However it evolved to being a two-sided network with users on one side and advertisers on the other.
    • So the first important component, two distinct user groups on either side of the network.
• The second important component in a platform mediated business network is the platform itself that facilitates the interactions between the two groups.
  • The platform company provides the technology and rules required to bring together the two distinct user groups.
  • For instance, we said Monster provides the technology that is the website as well as rules of engagement such as dispute resolution procedures, listing policies and other elements of engagement protocol.
  • So, all of these technological components and the rules together comprise the platform.
• Finally, the most important concept in platform mediated networks is that of network effects.
  • What network effects means, is that the value of the network and the platform to any given user depends on the number of other users with whom they can interact.
  • In the classic economic example, the first fax machine sold was useless until someone purchased the second fax machine.
  • The subsequent arrival of each new fax machine increased the value of each existing machine and the willingness to pay for the product by the consumer.
  • We noted in the case of Monster that firms would be more likely to join this platform if there are more users on the other side because it gives them access to one bigger pool of likely hires.
  • At the same time more individuals are likely to join the platform if they recognize that the platform will give them access to a large number of firms, and therefore more job openings.

  • The value of the platform increases with the number of users, distinct users who affiliate with that platform. That is at the heart of network effects.

  • Network effects in a two-sided platform mediated business network are of four types.
    • In a two-sided network, members of each group exhibit a preference regarding the number of users in the other group. These are called cross-side network effects.
    • And each group members may also have preferences regarding the number of users in their own group. And these are called same-side network effects.
    • That is, your preference as a job seeker for the number of recruiters on the other side is a cross-side network effect. Your preference as a job seeker for the number of seekers on your side Is the same side network effect.
    • Network effects would cross as well as same side, maybe positive or negative.
      • In two-sided networks, cross-side effects are usually positive but can be negative, for example with consumer reactions to advertising.
      • You may have a negative preference in the sense you may have, you may have a dis-utility from having a number of advertisers on the other side of the platform in which case the cross side network effect is negative.
      • Same side network effects may be either positive.
      • The benefit that you get from swapping video games with more players or it could be negative.
      • The desire to exclude rivals from an online business to business marketplace or even a platform like Monster.
      • For instance, you may not want too many job seekers applying for the same job. Therefore the greater the number of users on your side of the platform, applying for a given job, the same side network effect maybe characterized as negative.

Google

• What are the two distinct user groups that Google brings together?
• On one side, we have users.
  • What do you value Google for?
  • While users value the Google platform because it provides relevant and valuable information across the vast world of the Internet.
  • It’s become a verb in our lives.
• What’s the other side of the Google platform?
  • The other side of the Google platform are the advertisers. What’s the value to advertisers?
  • Advertisers value the Google platform because it gives to them a global audience of potential buyers who will likely to click on the ads, visit their virtual stores and buy their products or services.
  • But is Google just another advertising agency?
  • What other benefits do advertisers see in the Google platform beyond access to a global audience?
  • There is no direct marketing platform that does it better than Google.
  • Google gives you access to a vast range of potential consumers, while also providing you the ability to foster very rich interactions with these customers.
    • As an advertiser what information would you like to have?
    • Would it help if you could measure the ROI on your individual ads?
    • Would it help if you could have greater control over your campaign?
    • Would it help if you paid not for just the display of the ad as you would in television but for only when a user clicks on your ad and views its content and better yet buys the product or service.
  • These value added features make the Google platform more beneficial for the advertiser and are the reason why more than 50% of the current advertising money has migrated to the Internet and Google has been the huge beneficiary.
  • The company’s advertising revenue accounted for nearly 97% of its profits in 2012.
  • Now Google’s advertising has two components.
    • An AdWords model that is keyword based and
    • an AdSense model that is display advertising.
  • That is Google may choose to display an ad either as tag to a keyword which a user is searching for or on one of their partner sites whose content is related to the landing page of the advertiser.
  • To that extent, Google may be viewed as two two-sided network.
  • Google is atypical in that the users don’t directly interact with the advertisers. So it’s not your typical two-sided market.

Management Challenges for Networked Businesses

• Network effects are a critical element of these business models. Question is:
  • How can a platform mobilize the two sides of its network to create network effects that can subsequently leverage to generate revenue?
  • The willingness to pay for an existing fax machine goes up with each subsequent fax machine that comes along.
    • Well, fax machine’s are part of a one-sided network.
  • In the case of two-sided networks, or two-sided markets, network mobilization is more challenging.
    • This is because it represents a chicken and egg problem.
    • The amount that prospective users are willing to pay to participate on each side of the network is a function of the number of network users with whom they expect to be able to interact with on the other side.
  • In the case of Monster the more the number of job recruiters the more the number of job seekers will affiliate with the platform.
  • Similarly on a platform like eBay the sellers will come if the buyers will come, and the buyers will come if the sellers come.
  • Therefore, a challenge that platform providers face is which side to mobilize first and how.
• Let’s look at some of the strategies that platforms use in this regard.
  • Subsidization of one side of the network.
    • Think of Adobe, manufacture of the popular Acrobat Reader.
    • At the launch of the product in 1992 the company charged $50 for reader software and $195 for writer software.
    • It earned only $7 million from both products in its launch year.
    • In 1994 Adobe changed its pricing.
    • They gave away the reader software for free.
    • But continued to charge $195 for the writer software.
    • PDF quickly grew to become the dominant universal format for document presentation.
    • A decade later by 2005 Acrobat sales had grown to $640 million.
    • Almost one third of Adobe’s total revenue.
    • Why did the company give the Acrobat Reader away for free?
    • And why does it still continue to do so?
    • Because it realized that in order to convince publishers to use its software for producing documents, it needed to show them there was large mass of people who would read the documents using Acrobat.
    • Therefore the readers were subsidized and mobilized first.
    • The greater the number of readers, the greater the value of the Acrobat platform to the publishers.
    • And now the publishers would be willing to pay for the software through which they could create and publish Acrobat documents.
    • The writers were the money side for Adobe.
    • What we discern through the Acrobat example is a pricing strategy where a platform provider prices products or services to one side below marginal cost.
    • Sometimes even giving the product away.
    • Something it would not do if that side were viewed as an independent market, rather than part of a two-sided network.
    • While in some cases, these subsidies may be provided for early adopters, the subsidies are rarely lower than high penetration pricing.
    • The discount of subsidy, as we see in the case of Acrobat Reader, is often permanent.
    • Value seeded to the subsidy side is recovered through charges to the other side of the network also known as the money side.
    • The money side remember shares strong cross side network effects for the subsidy side.
    • To that extent pricing in two-sided networks departs from familiar economic rules that equate price to marginal cost in competitive markets.
    • In two-sided networks a platform provider prices so that its marginal revenue equals marginal cost across the entire network rather than separately for each side of the product.
      • you are talking about pricing below marginal cost to an entire set of consumers for an undetermined period of time.
    • You are going to burn a large amount of cash in this pricing strategy.
    • For example, you remember PayPal, the electronic payments company.
    • When they launched their service, in order to build a subsidy side, they offered a $10 credit to new accounts and a $10 payment to account holders who referred a new account.
    • PayPal was burning a million dollars a week.
    • They were ably aided by buoyant capital markets at the time, but not all firms have that kind of cash.
    • Of course, aggressive subsidization like this yields all the benefits of strong network effects as we will see, increasing the terms to scale, high switching costs, all of that.
    • But, as I said, I want you to observe that it is very costly and requires significant financial capital.
    • Digital goods such as software are still good candidates for subsidization strategies because of near zero marginal costs.
    • However when products have non-trivial marginal costs firms must be more careful.
  • If network effects as less than expected, subsidization can quickly rack up large losses.
• The next important question in the context of this subsidization strategy is to determine which side to subsidize.
  • If the network effect is discernibly stronger for one side, it is profitable to subsidize this side while charging a premium to the other side that values this side.
  • As a rule of thumb
    • subsidize the side that is price sensitive and demands quality,
    • and price the side that demands quantity.
  • A related strategy, in building either the money or the subsidy side of the platform, is to secure users’ exclusive affiliation.
    • That is the agreement not to have affiliate with a rival platform.
    • For many years, such exclusive licensing arrangements, where the core of Visa’s marketing campaigns.
    • And they don’t take American Express.
  • On the subsidy side, Google’s platform model attracts users through its world class search engine.
  • The search engine is based on the algorithm initially developed by Larry Page. Beyond their search engine, Google also attracts users to a variety of free services such as Gmail, YouTube, Google News, etc., etc. How does Google make money?
    • Through the fees that it generates from advertisers.
    • By providing the unsurpassed advertising services that we discussed earlier.
    • On the money side Google has pioneered the AdWords auction model to decide which ads will be displayed alongside search keywords and Adsense, its display advertising engine.
    • Google AdWords was a significant innovation of its time that combined a proprietary quality rank for each landing page, along with a second price auction to determine the final price paid by the bidder.
    • For now, let’s just absorb that the users of Google are its subsidy side and the advertisers, its money side.
    • For purposes of comparison, can you identify the subsidy and money side for eBay?
      • marketplace buyers of products and services on eBay are its subsidy side
      • marketplace sellers of these products and services are its money side

Winner-take-all nature of competition amongst platform models

Multi-homing costs

Multi-homing costs imply the costs of affiliating/ maintaining presence on multiple platforms at the same time. My most popular example is the case of internet-based email services. Even though it is literally free for anyone with an internet to have an unlimited number of email accounts, most of us cannot really maintain more than three email accounts. The monetary costs of creating and operating multiple email accounts may be zero, but the effort required to remember passwords, periodic logins to each of the accounts, and ensuring that you are communicating using the right email account is too much for most people. These are multi-homing costs.

eBay

• First question, are multi-homing costs significant for at least one side?
  • What are the multi-homing costs for sellers?
    • That is, what are the costs of auctioning wares on two platforms at the same time?

    • Well, to some extent, eBay’s policies themselves make it difficult to auction wares on two platforms.

    • For example, there are limits to when you can de-list a product or service. Fewer than 12 hours before the listing ends restrictions apply to de-list.
    • Similarly, sellers may find it difficult to manage reputation scores on multiple platforms.
    • So once they establish a good selling history, they have little incentive to switch.
    • Sellers may also lose out on volume discounts with both the eBay platform and its payment platform Paypal, if they auctioned on multiple platforms.
    • All of these represent significant multi-homing costs.
  • What are the multi-homing costs for buyers?
    • Well, similar to the sellers, the buyers face non-trivial costs in managing their profiles and security on multiple platforms.
    • So to the extent that buyers are able to find multiple products across diverse categories on a single platform, they don’t even have the incentive to multi-home.
    • So from these analysis, we can conclude that multi-homing costs are high for both sellers and buyers affiliated with the eBay marketplace.
• Now the second question. Does the side with significant multi-homing costs experience positive network effects?
  • Yes
  • Both sellers and buyers share strong cross-side network effects on the eBay platform.
  • As we noted earlier, the value of the eBay platform to the buyers increases significantly with the number of sellers that they can interact with and the value of the platform to the sellers increases significantly with the number of buyers that they can interact with.
  • Cross-side network effects, strong and positive in both directions.
• Finally, are opportunities for differentiation are limited in this platform market?
  • eBay has sales in multiple product categories, ranging from collectibles, to electronics, to apparel, to home decor, to houseware.
  • To the extent that sellers and buyers complete transactions across all these diverse categories, there is little space for another vertically differentiated auction platform.
  • Indeed, a variety of auction participants converge on eBay.
  • So what this table is telling us is that eBay displays strong Winner Take All dynamics.
  • No wonder it dominates the auction market the way it does.

Google

• Multi-homing costs?
  • what are your costs of affiliating with multiple search engines?
  • Not high at all. It’s quite trivial.
• What about cross-side network effects?
  • We documented very strong network effects for the advertisers with respect to the users who use the search engine.
  • However, they don’t exist very strong network effects in the other direction.
• So, if Google’s users, for whom multi-homing costs are low, switched to another platform, the advertisers will move with them.
• And finally, differentiation. Do you believe that your search experience has potential for improvement Are there opportunities for differentiation in search?
  • intelligent search that reflects an understanding of the search context.
  • search that uses machine learning to provide much greater levels of personalization by including user’s response to search results in determining ranking factors.
  • search that provides grouping functions in semantically related terms.
  • Vertical search etc.
  • In other words, the potential for differentiation in search, the potential for your search expedients to be improved is fairly high.
• And to summarize, our analysis of Google reveals:
  • low multi-homing costs,
  • strong network effects for the side that has low multi-homing costs, and
  • a fighting potential for differentiation.
• So in contrast to eBay, what we are discerning in the case of Google is very low Winner Take All dynamics.
• In fact, Google’s position in its market is quite fragile.
• You will discern that Google’s dominance in search and advertising is not intrinsic to its business model.
• It is an outcome of an agile organization that is continually looking over its shoulder and responding to competitive threats.

• It makes dominance look easy, but it works very hard to maintain that position.

• Microsoft’s Bing search engine that was launched in 2010.
  • This slide tells you that within the year of its launch, Bing captured 12.5% of the search market.
  • Place yourself in Google’s shoes. Would you be concerned?
  • Would you perceive Bing as a threat to your Winner Take All dynamics?
  • What would you do?
  • The question that Google sought to answer was what was driving the growth of Bing?
  • Remember opportunities for differentiation in search? One of the things that we mentioned was vertical search.
  • Well, comparisons of the two search engines showed that Bing did particularly better than Google for travel and healthcare information search.
  • Microsoft acquired this company called Farecast. A fare prediction startup for $115 million in 2008.
  • This company would tell you the likelihood of cutting flight prices dropping or increasing over different periods of time.
  • Microsoft incorporated the technology in its Bing search engine, creating the Bing travel section that was prominently linked off the homepage.
  • When any user attempted to do a travel search on Bing, the user was prompted to visit this section. Here, the results were far superior to a Google search on travel.
  • When travelers searched for tickets on a specific date, Bing travel would return a detailed overview of available flights including their prices, estimated departure and arrival times, airports of origin and destination, the number of connections that the user would need to make, layovers, estimated flying time etc.
  • Picking the right flight was made easy by visually showing the best fares within the specified time frame, as well as Farecast’s,remember them?
  • Farecast’s predictions of whether these prices were likely to fluctuate and in what direction.
  • As Microsoft sought to differentiate its browser in vertical search, there was a likely threat to the existing Winner Take All effects of Google.
  • And Google, sensing this threat, responded.
  • What did they do? Google acquired the leading provider of flight data, ITA software to blunt any differentiation that Bing might seek in the travel search space.
  • It also acquired review sites like Zagat and Frommer’s to boost the quality of information it could provide in travel searches.
  • Google’s acquisition made strange bedfellows of Microsoft, Expedia, Kayak, Travelport and Amadeus because Bing relied on ITA data for its travel searches.
  • Nevertheless, Microsoft was outflanked by Google which swiftly launched Google Flight in response.
  • In 2014, Microsoft turned off its flight prediction tool powered by Farecast.
  • Rumor has it, they did not want to pay Google for the data required to power Farecast.