Interview: How data helps large enterprises move to platforms

I was recently interviewed by Business Today, a leading business magazine, on the implications of platforms for large enterprises. This interview captures a lot of thinking from my upcoming book Platform Scale whose public launch is scheduled for early September.

BT: What exactly is platform thinking?

Choudary: Traditionally, businesses worked on a very linear model. Firms would assemble and deliver products and services to a market of consumers. Producers and consumers held very distinct roles. And this value creation happened inside a firm and was managed through a hierarchical organization. This was because value creation required the aggregation of resources and processes.
The real business impact of the Internet is that it now allows this aggregation to happen in a much more distributed manner. They do not need to be aggregated inside a firm any longer. This is giving rise to platform thinking. In India, we’re seeing companies like Flipkart and Snapdeal bring together an entire ecosystem of merchant partners and companies like Redbus using platform thinking to aggregate a highly inefficient and fragmented transportation market.
The Internet, today, provides an opportunity for businesses to create a digital platform and leverage an external ecosystem of producers to create value. This, in turn, allows the businesses to create and own entirely new markets when consumers come on board. It’s giving rise to a new age of monopolies that do not exercise power through control over physical resources but through orchestration of digital and physical resources using data.

BT: Will the platform model overtake all others?

Choudary: I believe that as platforms become more popular, the really large plays will be in becoming platforms while a host of smaller companies will co-exist around these platforms and operate in various niches. We already see this with YouTube that has allowed a host of new media houses to spring up that distribute exclusively through YouTube. Or Airbnb, which has allowed anyone with a spare room to run a B&B. Elance-Odesk or Freelancer allow people to turn to freelancing by offering them access to a market.
So platforms are going to be the really big plays but there is every reason for inventory-owned companies to co-exist. A platform approach helps companies create an infrastructure for all these other companies and benefit from owning all these interactions.

BT: How can large enterprises change their linear model and adopt a platform approach?

Choudary: The first step a traditional company needs to take is cultural. It needs to create a culture of data acquisition. All companies today have a culture of revenue acquisition. The sales reps who get in revenue are incentivised accordingly. The companies will need to move to creating a culture of data acquisition as well. As companies like LinkedIn demonstrate, the more data you acquire, the more ways you can make money. LinkedIn acquires way more data from its users than Monster ever did, and this helped it create a much larger job market than Monster’s.
The second change that the companies need to bring is infrastructural. They need to restructure their internal systems to be more data-porous. Today, most organisations use IT but they work in silos. There is minimal data exchange between units internally, a recipe for disaster. Companies need to move towards adopting internal APIs that allow the exchange of data and information internally between business units.
A combination of data acquisition and data porosity is the most important step towards moving to platform models. The companies that acquire and leverage data well will find new ways of understanding their users and allowing them to participate in the business in new forms. Toyota, GM and Ford are becoming data acquiring companies. Their cars constantly stream data about usage and this helps them better predict after-sales service and also allows them to monetize this data by selling it to insurance companies.

BT: Isn’t the platform model a bit risky given that it is only as good as the users who gather to use it?

Choudary: Yes, and this is why you need to be very careful about the kinds of users and partners you attract. This can be done in various ways. Some platforms restrict who can get access to start creating a culture gradually among its users. Think of Facebook restricting usage to Harvard, and subsequently to other Ivy Leagues, or Quora getting initial traction in Silicon Valley. Platforms also invest in regular user education and culture-creation initiatives. Airbnb trains its hosts regularly on best practices.
On the flip side, we’ve seen platforms like Chatroulette being overrun by the wrong kinds of users and dying an early death. It was as anonymous video chatting platform and the anonymity made way for all forms of unpleasant encounters, leading to users leaving the platform.
To ensure that the right users come on board, platforms need to start with a closely controlled launch, keep curating user access on an ongoing basis and ensure that they invest in user education and culture.

Feel Free to Share

BT: How do you scale up platforms?

Choudary: Platforms scale very well for a multitude of reasons:

Network effects: This is the most important driver for platform scale. More production activity on the platform attracts more consumption, which, in turn, attracts more production. This is a self-reinforcing virtuous loop that scales the platform. More ride availability on Uber attracts more travellers, which, in turn, attracts more drivers, leading to higher ride availability. We see this with most platform business models.

Near-zero marginal costs: The Internet has always allowed near-zero marginal costs of distribution. That is why the news industry got disrupted because the cost of distributing an extra unit of news (a news article) over the Internet, is nearly zero. This helps digital businesses scale in general. But platforms scale even better because they also benefit from near-zero marginal costs of creation. If a hotel has to add more rooms, it has to create buildings, own rooms and service them regularly. Airbnb can add more rooms at near-zero marginal costs of adding a new room. This is why non-inventory-based models like platforms scale so well.

Virality: A unique phenomenon of networks where users of a product or service help spread that product or service by spreading content about it. Instagram grew viral because new users came to know of it every time someone clicked a photo on Instagram and shared it on Facebook. Platforms benefit from strong virality, and this helps them scale even faster.

BT: You have applied rules from gaming in the platform thinking approach. How does that work?

Choudary: The platform thinking approach needs to ensure that every individual is appropriately incentivised to create value. As long as every individual user is appropriately incentivised to participate on Facebook, the news feed for all users remains fresh and active.
In The Wealth of Nations, Adam Smith observes: “Every individual… intends only his own gain, and he is… led by an invisible hand to promote an end which was not part of his intention… By pursuing his own interest he frequently promotes that of the society more effectually than when he really intends to promote it.” This is an interesting observation that shows that larger goals of societies and economies can be achieved by ensuring profitable work at the individual level.
This ‘invisible hand’ applies really well in implementing platforms. When you’re getting millions of users and partners to co-create value for your business, you need to ensure they are appropriately incentivised at an individual level. While society has always worked this way, our firms have been managed by iron-fisted internal control, not by an invisible hand that assures larger goals by ensuring individual incentives.
In the absence of great business literature on architecting such incentives, I turned to studying game design where the invisible hand phenomenon plays out. Small, seemingly innocent actions, by every individual player in an MMORPG (massively multiplayer online role playing games) are designed in a way that the gaming environment and conditions change in real-time for the good of all participants. Designing games is probably the closest analogy to designing activity on platforms. There are some rules of play, but most outcomes emerge from a combination of rules and carefully incentivized user actions.

If you did not pre-order Platform Scale, you can still get on the waiting list and be among the first to be informed of the public launch by signing up here.