Types of Platform Businesses
As mentioned in the previous article, platforms focus on facilitating interactions between consumers and producers instead of selling a standalone product. Their model is based on three pillars:
- The network effect allows for exponential growth, increasing the value of the platform for everyone involved.
- Technological progress has made this growth possible and sustainable.
- Uncompromising focus on customer value attracts and retains consumers and producers.
Despite these similarities, platforms vary widely in their structure and purpose. Some create value by enabling a direct exchange of goods or services between participants. Others encourage the production of content and deliver it to an audience of consumers.
At the same time, both types can be subdivided into platforms that aggregate existing offerings of goods, services or content and those that enable producers to create new offerings altogether.
This can be shown in the following matrix. the transitions between different fields are not always clear cut since many platforms provide secondary interactions that may reach into other quadrants.
Exchange of Existing Offerings
Platforms in this category often position themselves in between traditional pipeline businesses and their customers. They offer value to customers by making a variety of offers comparable and simplifying the order process. Since these platforms usually offer multiple different products, they are able to effectively cross-sell by utilizing innovative data analysis.
Amazon Marketplace: This platform collects existing Offerings of different companies from a variety of industries and makes them easily comparable. Customers can then order these products directly via Amazon.
Aggregation of Existing Content
These platforms aggregate existing content from different sources. While the content can usually be accessed independent of the platform, it is enhanced by its presentation in curated form. Often, the content is made searchable and filterable as well. In contrast to exchange platforms, however, no direct exchange takes place on these platforms.
Feedly: This platform combines existing news articles from different sources into a single news feed. Similar to a search engine, it does not enable direct exchanges but presents a channel for producers to market their content.
Exchange of New Offerings
While all types of platforms have a certain potential to disrupt existing businesses, this potential is particularly evident for this type of platform. Not only are new offerings created, which usually replace existing ones, but the direct exchange between producers and consumers often represents a form of direct competition to existing pipeline models that cater to the same demand.
Uber: This platform enables drivers to offer their services directly to consumers. Since most drivers would not have offered their services without the platform, this represents a new offering.
Aggregation of New Content
These platforms give producers the opportunity to create content and make it available to an audience of consumers. Since most social networks fall into this category, users on these platforms often take on the producer and consumer role at the same time. These platforms often take advantage of third party innovation by granting much creative freedom to their producers.
iOS: This platform enables developers to extend the platform’s functionality through new apps and to offer them to consumers.
When operating a platform, it is important to strike the right balance between the freedoms and restrictions granted to or imposed on consumers and producers. If the platform is too restrictive it might lose its value for users since innovations cannot be implemented. If the platform is too liberal, on the other hand, abuse can run rampant or producers can even try to compete with the platform’s objectives.
Apple is currently facing an example of this situation in relation to its service Apple Pay. Apple Pay uses a closed, private interface on the iPhone to enable contactless payments. Opening this interface would benefit banks, developers and consumers and thus increase the value of the iOS platform as a whole. At the same time, however, the resulting new apps would compete directly with Apple Pay. Therefore, Apple is resisting demands to make the interface publicly accessible, even in the light of litigation.
The appropriate level of restrictions also depends on the type of platform. Content platforms, such as Android or iOS, often have to be more liberal in their approach than exchange platforms because they rely to a large extent on third-party innovations to make their platform more attractive. Thus, in order to preserve their platform’s value, operators constantly need to keep an eye on the balance between freedoms and restrictions. The optimal levels for which, can change along with the platform’s level of maturity.
The right balance between restrictions and freedoms is particularly important in the beginning stages, so as not to stifle the crucial growth beyond a critical mass of consumers and producers.
The above-mentioned inter-dependence of the number of consumers and producers confronts the platform operators with the chicken and egg problem. Without a certain number of producers, the platform is useless for consumers because it usually does not offer any intrinsic value. Producers, however, will only be attracted to the platform if they can find an audience for their content or products.
This phase in the development of the platform is referred to as the seeding phase. While it should not be confused with seeding-rounds startups go through in order to find investors and build up an initial pool of capital, they are similar in the sense that the seeding phase aims to build an initial pool of users. This phase continues until the critical mass of consumers and producers has been reached so interactions can occur on a regular basis. These regular interactions make the platform attractive to new users and lead to the desired network effect.
There is a range of strategies for the seeding phase that can often be employed in conjunction with one another.
Platform Operator as First Producer
With this common approach, the operator will act as the first producer on the platform, overcoming the problem that most platforms do not offer any value in and of themselves. The goal is to attract a critical mass of consumers to the platform, who, in the next step, will lure in additional producers and start the network effect. It is not uncommon for the operator to be the only producer during this stage and to appear to be running a pipeline business rather than a platform. This goes to show that even enterprises that start out with the pipeline model can transform themselves into a platform with the right strategy and leadership.
A prominent example for this approach is Amazon. It started out as an online book retailer and only opened its platform, Amazon Marketplace, six years later. By then it had already established itself in four countries. This international customer base helped draw in third party retailers faster.
Comparison platforms often use web scraping to automatically extract relevant data from the compared websites. Once such platforms reach a significant number of consumers they usually open up their platform to the companies they are comparing, offering direct sales partnerships.
Customer incentive or refer-a-friend programs are not inventions of the digital age. Nevertheless, these enticements can be used effectively to boost the initial number of users. PayPal, for instance, heavily relied on these types of incentives, spending up to $20 per customer. This allowed them to aggressively grow their network and to reach 100,000 customers within their first month. Once the platform became increasingly valuable through the number of users alone, they began phasing out the rewards.
Most platforms also rely on the integration with social media. Even though it can be difficult to get consumers’ attention through these saturated channels, Instagram proved that the right strategy can work wonders. Instagram’s users were able to share their creations seamlessly on other social media sites, giving the platform’s producers access to millions of consumers from the start. Viral growth followed promptly, helping Instagram reach their first million active users within only two months.
Keep it Simple, Stupid
The Instagram example also serves to show that it pays to keep it simple when launching a new platform. Initially, Instagram focused solely on their simple concept of sharing edited photos. By simplifying this value proposition as much as possible, they were able to draw millions of users to their platform. After a certain user base was established they were able to branch out and add new interactions.
For platforms that, by their nature, are more demanding than Instagram, simply enabling interactions is often not sufficient. The principle of easy access, however, still applies. Operators of complex platforms, such as Android or iOS, often provide prospective producers with special software (software development kits), documentation, courses, conferences and similar aids to help them get started. While this may require a significant investment in a relatively small group of users initially, it becomes more proportionate as the platform matures and gains additional producers.
Focus on Producers
Unlike the previous approaches, platforms can attempt to get producers on board before focusing on consumers. Even though this approach is less widespread, it allows platforms to quickly accommodate large numbers of consumers later on.
The restaurant reservation service OpenTable implemented this strategy. The company started by selling digital customer management systems to restaurants. They could then integrate their consumer sided online reservation service with their management systems already used by many restaurants.
Another, counter-intuitive, strategy entails restricting access to a platform to a limited number of users. This restriction is often based on the users’ location. Aside from lower costs, this approach has the advantage that initial problems with the platforms can be fixed quickly and without much negative PR. Furthermore, the critical mass of users to make interactions happen regularly can be reached quicker with a smaller, easier to control user base.
Uber used this strategy by initially only offering its higher priced black car service UberBlack in San Francisco. This tackled two problems: It allowed them to circumvent the lack of producers by partnering with existing, professional black car services, who could bridge down times by accepting additional jobs on Uber. This approach also let Uber develop their platform further while continually adding an ever-increasing number of cities to their service. After gaining a foothold in several countries and cities, they opened their platform to the masses with their now most popular services UberX.
Data is the new Oil
Collecting and analyzing user data goes hand in hand with the platform model. While companies usually pursue a variety of concrete goals through data analytics, one of the most common denominators is customer value. Especially platforms whose business model relies on advertising revenue have a self-evident interest in it. The concrete applications of the technology are very diverse and vary greatly depending on the type and strategy of the platform.
Improvement of Service
Platforms can use the collected data to gauge promising areas of future development, the quality of their user experience, uncovered use cases, potentials for optimization or to reveal unwanted behavior. This information can then be used internally to make data-driven decisions about the future of the platform.
Increasing Value for Producers
Producers on most platforms will have a great interest in judging how their offers or content compares against their competitors’. Therefore, most platforms offer detailed metrics, tailored to the needs of their producers.
These metrics can range from simple numbers of Likes on social media sites or target audience analyses to detailed Dashboards. This lets producers discern quickly what works and what doesn’t to ensure they deliver the highest quality content, products or services.
Amazon, for example, offers their sellers detailed reports about their shipments, delays and performance on customer service requests. This data can, of course, be used in multiple ways. Sellers can use it to improve their service, while Amazon can control the quality of service by setting standards and terminating partnerships with underperformers.
Increasing Value for Consumers
Continuous improvement of the platform for consumers is every bit as important as the optimization on the producer side. The improvements can either be realized internally or, if appropriate, passed down to producers in the form of directives.
This point is illustrated by Uber’s efforts to evaluate their consumers’ GPS data even after a trip has ended. With this effort, the platform is investigating how frequently their customers are dropped off on the wrong side of the street, making it necessary to cross. By reducing these kinds of drop-offs, the company is hoping to enhance both, customer safety and convenience.
Comparison platforms often offer a wide variety of different product categories. This holds the potential to generate value for everyone involved by analyzing the data in order to enable cross-selling. If a customer changes their address, for example, they could receive additional offers for changing electricity, insurance or telecommunications providers. Even if traditional pipeline companies collect the necessary data, they seldom make use the potentials of thorough analytics. Often their product range is not broad enough to justify this kind of cross-selling or they are lacking the necessary infrastructure to automate these tasks.
Machine Learning is used increasingly to automate complex tasks in a dynamic environment. Popular areas of application range from relatively simple tasks like spam recognition in email applications to self-driving cars. Regardless of the application, however, all machine learning systems require huge, reliable training data sets. Relatively reliable, learning translation systems, for instance, require several billion bilingual sentence pairs to study the syntax, grammar and vocabulary of the two languages.
Successful platforms with a high number of users can be used to create training data sets for a variety of different tasks. Popular uses for such systems include product recommendations on e-commerce sites or tailored ads on social media sites.
This concludes the discussion of some of the concrete uses and potentials of the platform business model. In the last part of this series, we will focus on how traditional pipeline businesses can prepare themselves against new platform competitors.