Platform Scaling, Fast and Slow

Conventional wisdom says digital platform businesses should scale quickly, but that’s a mistake in some markets.

Reading Time: 15 min 

Topics

Buy
Like what you’re reading?
Join our community
Member
Free

5 free articles per month, $6.95/article thereafter, free newsletter.

Subscribe
$89 $45/Year

Unlimited digital content, quarterly magazine, free newsletter, entire archive.

Sign me up

Image courtesy of Michael Glenwood Gibbs/theispot.com

Shortly after its 2009 founding in San Francisco, Uber executed a simple strategy that rapidly led to its expansion on a global scale. To achieve network effects by connecting as many drivers and passengers as quickly as possible, the company prioritized launches in new cities. It hired core teams of general managers, operations managers, and community managers in multiple cities at once. In each city, these teams attracted drivers by offering existing black-car services an app — and sometimes a free smartphone — to monetize their idle time. To attract riders, the teams offered subsidized fares to attendees of large conferences and other high-profile events, signing them up and then gaining thousands more riders through word of mouth.1

Rapid scaling, as exemplified by Uber, is a core element of platform strategy, with speed considered the decisive factor in the race to succeed in winner-takes-all and winner-takes-most markets.2 But we’ve found that rapid scaling may not be the best strategy for all platforms. In some cases, a more careful, incremental, and thus slower approach to scaling is more beneficial.

In studying platform businesses, including Airbnb, Amazon, Apple, Expedia, Facebook (particularly its e-payment project, Libra), Google, Grindr, LinkedIn, Netflix, PayPal, and Uber, we found that regulatory complexity and regulatory risk are two significant but often neglected factors in platform scaling decisions. Moreover, they are likely to become increasingly important in the years ahead as efforts to regulate tech companies gain momentum and as more companies in a greater variety of sectors and markets seek to capture the benefits of platforms.

Plotting Regulatory Complexity and Risk

Regulatory complexity describes the current level of legal and regulatory barriers that govern platform entry and operations in a sector. The costs of operating in sectors with high levels of regulatory complexity, such as financial services or pharmaceuticals, can be significant, but legal and compliance teams can analyze and accurately predict them.

Regulatory risk refers to the probability of an increase in legal and regulatory costs and complexity in the future. It includes a higher degree of uncertainty than regulatory complexity. For instance, until a few years ago, public opinion of tech companies and their platform businesses was largely positive, and policy makers were lenient. Recently, however, a majority of Americans have said that they want tougher regulations for tech companies, and lawmakers in the U.S., U.K., Israel, Japan, and the European Union have called for stricter antitrust, taxation, consumer and data protection, financial, and labor laws and regulations for technology companies.3 Such regulations can result in considerable expense: Witness California Assembly Bill 5, which limited the ability of companies to classify gig workers as independent contractors and threatened the platform models of companies such as DoorDash, Lyft, and Uber. Along with Instacart and Postmates, those companies spent $224 million — a record-breaking amount in a California proposition campaign — to successfully convince voters to pass Proposition 22, which exempted them from some provisions of the bill.4 Likewise, Google is likely to face considerable costs arising from the antitrust suit that has been brought against it by the U.S. Department of Justice.5

It is notoriously difficult to predict policy outcomes or even attribute odds to different outcomes.6 But there are some objective and quantifiable metrics for calculating regulatory risk, such as ongoing legal cases, probes and inquiries by government agencies, and the number and political influence of lawmakers who argue for tighter regulation.

A simple way for platform owners and operators to understand the potential combinations of regulatory complexity and risk is to think of the two factors as the axes in a 2×2 matrix. (See “Mapping Regulatory Risk and Complexity.”)

LinkedIn and Expedia are two examples of platforms exposed to low levels of regulatory complexity and risk. Compliance costs are relatively low in their sectors, as well as in sectors such as software (like Salesforce.com) or entertainment (like Netflix). In addition, there are no serious debates among lawmakers and policy makers in the U.S. and elsewhere regarding whether to restrict their business models or operations.

PayPal is among the platforms operating in markets with high levels of regulatory complexity but low levels of regulatory risk. The company is in the highly regulated financial services sector, where overall compliance spending amounts to $270 billion, and 10% to 15% of the workforce is employed in governance, risk management, and compliance activities.7 The platform successfully navigates this regulatory environment by carefully weighing regulatory complexity when it chooses which services to offer its 300 million customers. For instance, PayPal didn’t seek its first banking license for Europe — a move that would have significantly increased regulatory complexity and scrutiny — until 2007, five years after its initial public offering. But its regulatory risk is low, because the company is not at the center of current policy debates.

Some platforms face low regulatory complexity but high regulatory risk. Online dating doesn’t involve a high degree of regulatory complexity, but Chinese gaming company Beijing Kunlun Tech didn’t anticipate regulatory risk when it purchased a majority stake in the dating app Grindr in 2016 and acquired full ownership in 2018. Kunlun was planning to grow the platform and launch an IPO, but then the Committee on Foreign Investment in the United States (CFIUS) stepped in. Concluding that the platform’s data could potentially be used by the Chinese government to blackmail U.S. officials or military personnel, CFIUS forced Kunlun to divest from Grindr.8

Platforms occupying the quadrant with low regulatory complexity but high regulatory risk include more and more companies, such as Airbnb, Amazon, Facebook, Google, and Uber. A closer look reveals that many of them are operating in a regulatory void — that is, a context without established and powerful regulatory authorities, a tight net of rules, and strict barriers to entry. Accordingly, there is a high degree of uncertainty regarding how regulators may react, which makes it difficult for these businesses to develop discrete policy scenarios, attribute probabilities, and make robust assumptions on timing.9

Finally, some platforms are in markets where they must contend with high regulatory complexity and high regulatory risk. In 2019, Facebook and a consortium of international partners announced the Libra project, a blockchain-based payment system. Since the release of Bitcoin in 2009, the cryptocurrency market had been rapidly developing, with some countries, such as China and India, choosing a restrictive approach and most Western nations favoring a more nuanced approach that tempered regulation to encourage technological experimentation. Nevertheless, Libra triggered immediate alarm among policy makers and regulators, as well as a precipitous rise in the entire sector’s regulatory risk. This occurred for three reasons: a general lack of trust in Facebook; the potential reach of Libra, given Facebook’s 2.4 billion users; and the Libra consortium’s inability (or unwillingness) to explain how it would obviate negative effects in sensitive areas such as terrorism financing, tax evasion, and money laundering.10 As a result, in 2020 Facebook and its partners had to drastically revamp their institutional efforts and downgrade their ambitions for Libra to, as the Financial Times put it, “appease wary regulators.”11

Scaling Decisions in the Four Quadrants

Nobel Prize-winning behavioral economist Daniel Kahneman made a distinction between “fast” and “slow” thinking to illustrate two very different modes in which the brain operates under different circumstances. He asserted that fast thinking prevails in situations requiring rapid and intuitive action (such as if you hear a rattlesnake), whereas slow thinking occurs in situations requiring more deliberate, orderly, and computational mental work (such as when you calculate your annual income tax).

Analogous to Kahneman’s distinction, we argue that platform owners and operators should explicitly decide whether to scale their user base fast or slow. Fast scaling, which has also been called blitzscaling by Reid Hoffman and Chris Yeh, means prioritizing speed over efficiency.12 The strategic objective in fast scaling is to grow rapidly, experiment quickly to improve product-market fit, and leverage strong network effects to attain and maintain a leading market share. Slow scaling entails detailed scenario planning and actor analysis, careful risk management, incremental geographic expansion, and continual investment in the platform’s reputation and trustworthiness. It does not exclude the pursuit of network effects, which are a prerequisite of success for platform businesses, but it prioritizes analysis, iterative growth, and risk minimization over speed.

Increasingly, regulatory complexity and risk are becoming the determining factors in the choice between fast and slow scaling. Legislators and regulators were initially slow to react to the disruptive effects of the platform economy, but that is changing: Currently, there are vivid debates on the appropriate policy landscape for platform businesses in countries as diverse as France, Germany, Israel, Japan, Mexico, Russia, the United Arab Emirates, the U.K., and the U.S. These debates are resulting in legislative and regulatory changes at an accelerating pace.13

In low-risk regulatory contexts, fast scaling is necessary to activate three interrelated positive-feedback loops:14

  • A network loop, in which growing numbers of users make the platform more useful and valuable to new users.
  • A data loop, in which more data yields more insights regarding consumer preferences, market structure, and market trends, which are used to improve the platform’s product-market fit, making it more attractive to new users.
  • A capital loop, in which high growth rates make the platform more attractive to investors — generating the funding and know-how needed to support continued growth.

If any of these loops cannot be activated, scaling takes longer or becomes impossible to achieve, and the platform can become an also-ran. MyTaxi, a ride-hailing platform founded in Hamburg, Germany, in 2009 — shortly before the launch of Uber in the U.S., Halo in the U.K., and GetTaxi in Israel — is a good example. MyTaxi’s business model was similar to Uber’s, its technology was well engineered, and early feedback from drivers and riders was overwhelmingly positive. Yet MyTaxi was unable to raise the capital it needed to scale fast (mainly because of Germany’s shallow venture capital market).15 As a result, MyTaxi had to merge with a car-sharing platform, and today its rides and revenue are only a fraction of Uber’s.16

Fast scaling is also the most appropriate strategy for platforms facing low regulatory complexity and high regulatory risk. This seems counterintuitive in a context where fast scaling may arouse the attention of policy makers and regulators, à la Kunlun’s plans for Grindr. However, our analysis finds that the powerful advantages of the three feedback loops outweigh the regulatory risks, at least in the short term.

Witness the current situations in which some of the world’s largest platform companies, including Amazon, Apple, Facebook, and Google, find themselves. All of them started out in a context of low regulatory complexity and low regulatory risk, and they scaled fast. Now, because of their success and the dominant market positions they have attained, they have increasingly attracted the attention of lawmakers and oversight authorities; in essence, they have migrated to the quadrant of low complexity but high risk.

In reaching scale, however, they have also gained a powerful resource that helps mitigate regulatory risk: a huge base of users, who can serve as powerful political advocates.17 Thus, in 2017, when Transport for London (TfL) stripped Uber of its license to operate in the city because of safety failures, Uber was able to respond with a petition to renew its license that was signed by more than 500,000 people within 24 hours.18 TfL relented and granted Uber several extensions. Similarly, when the controversial Stop Online Piracy Act was introduced in the U.S. House of Representatives in 2011, more than 7 million Google users signed a petition against it.19 The bill died in committee. And when they are unable to influence legislators, as with California’s Assembly Bill 5, the platform companies don’t only have the financial resources to draft and finance measures such as Proposition 22 but can also use their platforms to influence users: Customers in California were targeted with in-app campaign messaging and via stickers on delivery bags from Instacart and DoorDash.20

These examples suggest that the ability to leverage a scaled-up user base as advocates in the political sphere provides a strong incentive for companies facing low regulatory complexity but high regulatory risk to scale fast. Moreover, in the short term, the risks of slow scaling in terms of networks, data, and capital outweigh the risk of attracting regulatory scrutiny. It is possible that the benefits of a switch to slow scaling may be substantive in the long term, but that is neither clear nor tangible given the residual uncertainty in this context.

In arenas with both high regulatory complexity and high regulatory risk, however, slow scaling is the most prudent strategy. Facebook demonstrated the pitfalls of fast scaling in this quadrant with its Libra project. Our interviews, as well as public statements, revealed that financial regulators were surprised by the project’s fast-scaling strategy, which they found highly inappropriate, especially in light of Facebook’s involvement in data misuse scandals and the project’s disruptive potential vis-à-vis national financial and monetary policies.

“Libra, like any [cryptocurrency] project with global scale and scope, must address a core set of legal and regulatory challenges,” said U.S. Federal Reserve governor Lael Brainard in a December 2019 speech. “A significant concern regarding Facebook’s Libra project is the potential for a payment system to be adopted globally in a short time period and to establish itself as a potentially new unit of account.”21

As Brainard’s comment suggests, Facebook’s fast-scaling approach in an environment of high regulatory complexity and risk led to its quick shutdown by powerful financial market supervisors.22 A slower, more careful scaling strategy would have been less controversial and more likely to have led to Libra’s success.

How to Scale Slow

Platform businesses operating in high-risk, high-complexity environments might avoid the challenges faced by the Libra initiative by using a slow-scaling strategy that has four key ingredients: analysis of the macro environment, careful risk management, investment in stakeholder trust, and incremental geographic expansion.

Analysis of the macro environment: Analysis begins with the selection of the strategy team. In contexts of high regulatory risk, platform owners and operators need to predict policy dynamics and identify potential regulatory scenarios. This requires that they supplement their legal, technical, and business teams with policy experts, risk analysts, and scenario planners.

These experts should provide in-depth analyses — and mapping before and during project development — to identify relevant institutional actors and understand their mandates and priorities along with the broader economic, social, and political effects and implications of the platform. Such analyses are a prerequisite for identifying risks, making underlying probability assumptions, and developing strategic responses.

Careful risk management: As a natural extension of the above analysis, platforms need to identify risks and develop a sound risk management system in the context of high regulatory complexity and risk. Introducing a risk management system too late can be costly in terms of time, money, and reputation. Thus, early in the process of strategy making, risk management and scenario planning should receive the same level of attention by senior management as the platform’s technology and business models.

Technology companies can identify and manage a wider range of risk by adapting the environmental, social, and governance (ESG) mechanisms already in place in other sectors. A broad set of ESG standards and risk management tools already exists.23 However, ESG mechanisms designed specifically for the digital economy are still in their infancy.

Investment in trust: Too often, companies focus their efforts on innovative technology and attractive user interfaces but neglect the potential societal consequences of their platforms. In a recent survey of 34,000 people in 28 countries, more than 60% of the respondents said they are worried that tech companies are “out of control” and that governments are not regulating them effectively.24 Such public sentiment, and the demands of investors and other stakeholders, are driving leaders to place a higher priority on seeing that their companies behave in a trustworthy and reliable manner.25

Trustworthiness is especially important in contexts of high regulatory complexity and risk. To attain this status, platform operators should understand the underpinnings of public and institutional trust, and invest resources to maintain and enhance trust from regulators and consumers.

Narrow geographical focus, incremental expansion: Because trying to achieve global scale in sectors typified by high regulatory complexity and risk is hazardous, platform expansion should be more cautious. Experimental techniques that have become a mainstay of improving product-market fit, such as A/B testing, can be difficult or strongly limited in highly complex and risky regulatory contexts.

One alternative is to test the waters in selected jurisdictions. The resulting interactions with regulators and consumers can provide important insights and highlight previously undetected risks. Before committing to geographic expansion, these findings can be fed into product development and political strategy.

“Successful tech businesses need to understand how to navigate through the complex, and not always coherent, regulation that global lawmakers are rolling out,” concluded international law firm Hogan Lovell after surveying new tech regulations in 16 jurisdictions across the globe.26 We expect this advice to apply to more and more platforms in the coming years.

Daniel Kahneman proposed a more nuanced understanding of human cognition based on the idea that thinking fast is advantageous in some situations while thinking slow is better in others. Similarly, our research suggests that a more nuanced understanding of platform scaling is needed. We think that understanding should include regulatory complexity and regulatory risk — two parameters that enable platform owners to plumb the macro environment and design sound and context-specific scaling strategies. We foresee that these parameters will become increasingly important for tech companies in the future, especially as digital disruption expands to more strictly regulated sectors, and policy makers and regulators increasingly redesign legal frameworks in the era of the platform economy.

Topics

References

1. Y. Moon, “Uber: Changing the Way the World Moves,” Harvard Business School case no. 316101 (Boston: Harvard Business School Publishing, 2015).

2. See, for example, R.I. Sutton and H. Rao, “Scaling Up Excellence: Getting to More Without Settling for Less,” (New York: Crown Business, 2014); and R. Hoffman and C. Yeh, “Blitzscaling: The Lightning-Fast Path to Building Massively Valuable Companies” (New York: Currency, 2018).

3. A. Breland, “Americans Want Tougher Regulations for Tech Companies: Poll,” The Hill, April 20, 2018, https://thehill.com; Hogan Lovells, “A Turning Point for Tech: Global Survey on Digital Regulation,” Oct. 30, 2019, www.hoganlovells.com; and G. Edelman, “Congress Unveils Its Plan to Curb Big Tech’s Power,” Wired, Oct. 6, 2020, www.wired.com.

4. R. Menezes, M. Moore, and P. Do, “Billions Have Been Spent on California’s Ballot Measure Battles. But This Year Is Unlike Any Other,” Los Angeles Times, Nov. 13, 2020, www.latimes.com.

5.Justice Department Sues Monopolist Google for Violating Antitrust Laws,” U.S. Department of Justice, Oct. 20, 2020, www.justice.gov.

6. N. Silver, “The Signal and the Noise: Why So Many Predictions Fail — but Some Don’t” (New York: Penguin Press, 2012).

7. KPMG, “There’s a Revolution Coming: Embracing the Challenge of RegTech 3.0,” 2018, https://assets.kpmg.

8. Y. Yang and J. Fontanella-Khan, “Grindr Sold by Chinese Owner After U.S. National Security Concerns,” Financial Times, March 7, 2020, www.ft.com.

9. K. Gurses and P. Ozcan, “Entrepreneurship in Highly Regulated Markets: The Battle for Pay TV in the U.S.,” Academy of Management Journal 58, no. 6 (December 2015): 1709-1739.

10. C. Condon, “Fed’s Brainard Raises Red Flags Over Facebook’s Libra Project,” The Guardian, Dec. 18, 2019, www.theguardian.com.

11. H. Murphy and I. Kaminska, “Facebook’s Libra Overhauls Core Parts of Its Digital Currency Vision,” Financial Times, April 16, 2020, www.ft.com.

12. In “Blitzscaling,” Hoffmann and Yeh also employ the term “fast scaling,” but with a different meaning than used here.

13. Hogan Lovells, “A Turning Point for Tech.”

14. For a more detailed discussion of the specific characteristics of platform business models, see, for instance, G. Parker, M. van Alstyne, and S. Choudary, “Platform Revolution: How Networked Markets Are Transforming the Economy — and How to Make Them Work for You” (New York: W.W. Norton, 2016).

15. C. Keese, “The Silicon Valley Challenge: A Wake-Up Call for Europe” (Munich: Penguin Verlag, 2016).

16. C. Kapalschinski and L. Holzki, “Free Now Will zur Roller-Plattform Werden,” Handelsblatt, Nov. 7, 2019, www.handelsblatt.com.

17. B. Uzunca, J.P. Coen Richtering, and P. Ozcan, “Sharing and Shaping: A Cross-Country Comparison of How Sharing Economy Firms Shape Their Institutional Environment to Gain Legitimacy,” Academy of Management Discoveries 4, no. 3 (September 2018): 248-272.

18. M. Farrer and N. Khomami, “More Than 500,000 Sign Petition to Save Uber as Firm Fights London Ban,” The Guardian, Sept. 23, 2017, www.theguardian.com.

19. E. Engleman, “SOPA Bill Petition Collects 7 Million Signatures, According to Google,” The Washington Post, Jan. 19, 2012, www.washingtonpost.com.

20. S. Harnett, “Gig Companies Are Making Their Workers Promote Prop. 22,” KQED, Oct. 20, 2020, www.kqed.com.

21. L. Brainard, “Update on Digital Currencies, Stablecoins, and the Challenges Ahead” (remarks at Monetary Policy: The Challenges Ahead — An ECB Colloquium Held in Honour of Benoît Cœuré, Frankfurt, Germany, Dec. 18, 2019).

22. S. Webb, “Facebook’s Libra Is a ‘Failure,’ Says Swiss President,” Coin Rivet, Dec. 30, 2019, https://coinrivet.com.

23. W. Henisz, T. Koller, and R. Nuttall, “Five Ways That ESG Creates Value,” McKinsey Quarterly, Nov. 14, 2019, www.mckinsey.com.

24.Edelman Trust Barometer 2020,” Edelman, January 2020, www.edelman.de.

25. N. Megaw, “Index Ventures’ Jan Hammer: Bringing Perspective to Single-Minded Tech Founders,” Financial Times, March 8, 2020, www.ft.com.

26. Hogan Lovells, “A Turning Point for Tech.”

Reprint #:

62303

More Like This

Add a comment

You must to post a comment.

First time here? Sign up for a free account: Comment on articles and get access to many more articles.