Some of the biggest startups of the last two decades are platform businesses. Google, Facebook, Uber, Airbnb and eBay are just a few examples. And platform businesses have become some of the most powerful businesses in modern economies.
Inspired by these phenomenal successes, innovators have started bringing platforms into an increasing amount of industries. And it is more than likely that they will start reshaping many of those industries where they have not already done it. And at the same time, they will continue the transformation of those where they have become massive players already.
Today you will learn fundamental economic principles of the platform business model and how this can trigger your own innovation ideas.
Types of platform businesses
(1) Economic benefits
The success of your innovation idea will be greatly driven by the economic benefits your innovation will provide to your customers. In the case of platform businesses, this will be driven by the economic benefits that your platform will provide to the multiple sides that the platform connects.
So, do multi-sided platforms add actual economic value or are they just-another-middleman that tries to capture value without creating any value? In most cases, the answer will be that they add value. Platforms are still very young and do not fit into the classic economic models. So, let’s just look at two examples as a starting point for their value-add.
Example 1: Uber
Regardless which way you measure, it seems cars are parked roughly 95% of the time. Add to this the fact that it costs an average of $8,558 per year to own a car in the U.S.
Peter Cohen, et al. state “using almost 50 million individual level observations and a regression discontinuity design, we estimate that in 2015 the UberX service generated about $2.9 billion in consumer surplus in the four U.S. cities included in our analysis.” This is for the four cities of Chicago, Los Angeles, New York, and San Francisco. Based on this they estimate a consumer surplus (=economic benefits) of $6.8 for the US in 2015 alone.
You can read more about this and Uber’s business model here.
The economic benefits summarise to:
- Average cost of ownership per year in the US: $8,558 ($23 / day)
- Cars are utilised only 5% of the time (72 mins/day)
- Economic benefit for consumer of $6.8b (in the US) in 2015 from Uber alone
Example 2: Google
How can we know if platforms add value if we don’t know how it would be without them? Finding out the counterfactual is (nearly) impossible.
On August 17th 2013, we had one of those rare occasions to check the counterfactual. For approximately 5 minutes the probably most significant platform – Google – went down for about 5 minutes. And internet traffic dropped by a staggering 40%.
Does this mean that internet traffic would be down by 40% without Google? Surely not. But let me ask this way. If you continuously would not find on the internet what you were after, wouldn’t you use it (much) less? I am certain for most people the answer is yes.
Surely, internet traffic would not go down by 40% without Google in the long-run. But even if it went down by a few percent – because people struggle to find what they are after – it shows that they add genuine value. And if people substituted Google for a different search service, it still shows that a search platform provides value to the consumer. And it adds value to the other side, advertisers, as well and the advertiser’s added-value to the user.
(2) True value creation …
Some authors compare platforms to the ancient bazaars or physical marketplaces. They note that marketplaces have connected buyers with sellers and that modern multi-sides do the same. I believe there is an important difference.
Yes, both bring together two different types of groups (=sides) that then engage in value-creating ways. And both types of platforms provide governance under which both can safely interact with each other. But I would argue that traditional marketplaces do just that and not much more. For centuries, marketplaces have not changed. Some rules may have been tweaked but by and large not much has changed.
Like marketplaces, Google brings together advertisers and search users (who happen to be potential customers of the advertisers). And they add value to advertisers in that they allow more targetted ads than traditional channels.
But Google does this on the back of a much bigger ambition. An ambition that actually only affects one side (the search user). Their ambition is to “organise the world’s information and make it universally accessible and useful”. Even if this is slightly overambitious, Google is adding massive value to the user even if the other side was not there. A marketplace adds quite little value when only buyers or only sellers are there.
But more importantly, Google very actively steers how content is being structured and what is being considered value-add content. E.g. being referenced and linked to is a very important criterium for being seen as very valuable by Google. An expectation that drives millions of content creators every day. E.g. I am structuring this article and many other elements of my web pages in ways that – hopefully – signal Google in which way it adds value to potential searchers. This holds true for any article that you read online and most blog post and most corporate web pages.
… not just another middleman
In 1998 Paul Krugman, economy Nobel prize winner, “predicted“: “The growth of the Internet will slow drastically […] most people have nothing to say to each other! By 2005 or so, it will become clear that the Internet’s impact on the economy has been no greater than the fax machine’s.”
Comparing communication on the internet to peer-to-peer communication via fax was not far-fetched. At the time, email was the prevailing channel for electronic communication among people who know each other.
But social media – especially Facebook – has proven this prediction ridiculously wrong. Facebook has massively changed how people communicate with each other. As did Google, Facebook didn’t just bring people together and left them to their own devices. Anyone seeing in Facebook just another medium (or middleman) could not be more wrong.
They have actively shaped how communication is being held. E.g., instead of writing an email to a friend, considering what the friend may be interested in and carefully crafting it, people now just blurt out what they care about in that moment. And then they wait and see who among their online friends cares to respond to the comment (or photo).
Some modern platforms indeed just bring two or more sides together, establish the rules and then let the different sides create the value. Creating value beyond that is significantly more powerful.
Connecting multiple sides is powerful
Let’s not talk down the power of (traditional) marketplaces! They do amazing things. Marketplaces bring people together, connect demand and supply and thereby determine prices. Prices then give a signal to existing (and potential) providers if it is worthwhile to invest in providing more supply.
Ancient bazaars along the Silkroad have indicated to traders which goods are in demand. Traders have taken these signals back to the producers which make medium-term decisions based on this feedback.
Modern platforms are similar. But they are also different. Pure exchange platforms are powerful, by bringing together those that are on the demand side with those on the supply side. And this in itself leads to an incredible chain (this 2-minute Milton Friedman video explains it very powerfully).
But I do want to drive the point home that the best modern platforms are different. They actively manage the interactions of the involved sides based on the data they observe. This is the most powerful (and yet most underestimated) element of modern platforms (especially by economists who try to fit the platform business model into existing theory).
(3) Network effects
Network effects are a fundamental concept that is underpinning multi-sided platforms.
Network effects are the effects of a user on the value of the network to other users (and non-users). Take retail – there are no network effects. If there are more shoppers in your grocery store or less, it will not change the value to you. Yes, prices may come down due to economies of scale, but this is a different economic mechanism and not considered as a network effect.
- Direct network effects: Take Facebook. If you were the only user of Facebook, it would have no value for you. Ok, to make it less ridiculous: if there was hardly anyone of your friends or family on Facebook, it would provide very little value to you. But as more of your friends join, so grows Facebook’s value to you. These are direct network effect: the value of the network to the user as the user base grown. Direct network effects are also called same-side network effects.
- Indirect network effects: Indirect network effects are effects on other types of participants of the network. What are other types of participants? Those, that use the network for a different economic purpose. For example, advertisers use Facebook for (drumroll): advertising. As more people join on the user side (=non-advertiser), Facebook becomes not only more valuable to other users but also to the other side, i.e. advertisers. Indirect network effects are also called cross-side network effects.
You won’t be overly surprised if I told you that indirect network effects are crucial for multi-sided networks (that’s where the whole name comes from). However (and this is a big “however”), for many multi-sided platforms same-side effect are essential as well. Social networks, communication platforms and many others are dependent on same-side effects to increase the value of the overall network and thus for growth.
For other platforms, same-side effects are not the decisive ones. Take Uber. The value of Uber does not increase for you when many other passengers join Uber. However, indirectly it will play a role. If only a few passengers join, then there will be also not many drivers who increase waiting times and reduce the value of Uber to you. It is a complex dynamic that is being carefully monitored and managed by Uber.
Here is a great illustration of the feedback loop showing the various same-side and cross-side network effects. For example, an increase in demand will likely lead to more drivers but if demand increased very rapidly in a short period of time (maybe days) the supply side may not ramp up fast enough. It could initially lead to slower pickups, hence reduce customer experience until the supply of drivers increase. (Uber have what they call surge pricing but we will ignore this for now as it is not relevant to this discussion.)
It is complex. But at this stage, all we need to know is that these considerations play an important role. And they play an important role in the design of the platform.
Negative externalities and level of control (=governance)
Networks have so-called externalities. Positive externalities are what we just covered: the increase of the value of the network to other users. But there are also negative externalities. As the network grows, so does the number of negative incidents.
For example, a number of large advertisers (Wal-Mart, PepsiCo and others) have pulled out of YouTube advertising recently after their ads were shown ahead of videos with hate speeches and potential terror content. These negative externalities reduce Google’s revenue as the owner of YouTobe (or more accurately Alphabet Inc’s revenue as the owner of both Google and YouTube). But the consequences can be dire.
In the early 80s, Atari’s game console was at its peak popularity. It had an open platform to developers without any controls of the games brought to sale. In 1982, some analysts already saw an oversupply of games. This was then exacerbated by the flood of low-quality (and offensive) games trying to make a quick buck. Buyers protested and stopped buying. Within about a year revenues dropped by 97% contributing to Atari’s bankruptcy.
This is an important lesson to any platform owner to take negative externalities very seriously.
Sources of negative externalities
Just like positive externalities, negative externalities can have their source in the same-side or the cross-side.
Examples for same-side negative externalities:
- Facebook, Twitter: spammy, offensive, rude comments and posts by users
- YouTube: Offensive, discriminating videos
- Google: pages optimised for search engines but providing only low-quality content to humans
Examples for cross-side negative externalities:
- Facebook, YouTube, Google: spammy and offensive advertisements
- Uber: Abusive drivers or passengers
- Ebay: Deceptive sellers
- LinkedIn: Fake profiles, references,
- AppStores: low-quality apps, bloatware, malware
Various degrees of control
Platform businesses have various ways to control these negative externalities. After the video game crash of 1983, game console platforms tightened the control significantly. Different platforms imposed varying degrees of controls over the games.
Any platform needs to cautiously design (and constantly reevaluate) their controls. Too tight controls can limit growth, too loose controls can lead to a sudden death (Atari) or prolonged negative publicity.
The controls can be centralised, i.e. imposed by the platform or decentralised, meaning by the users of the platforms. It will always be a combination of both but most likely the heavy lifting will likely be done by one or the other.
- Google has very strong limitations on the ads that it allows. It does not make them standout in comparison to the organic search results. It limits the length of the advertisements texts, doesn’t allow capital letters (“FREE”) or overuse of exclamation marks (“CLICK NOW!!!!!”), excludes offensive vocabulary, does not allow any pictures, etc. It is important not to underestimate this. Early search engines have lost their market share to Google because they could be very easily fooled by keyword stuffing and displayed low-value pages on the top.
- Facebook has similar controls on ads. Allows pictures but controls them tightly. It doesn’t allow more than approx 20% of text within images (used for ads), controls the content of graphics, the text of ads and so on.
- Ebay was one of the first to introduce a feedback system. The buyer would rate the seller, thereby singling out bad apples very quickly. On top, Ebay has an arbitration process that – among sellers – is assumed to favour buyers in case of doubt. The latter being a centralised control.
- Rating systems are now one of the most common, real-time systems for self-governance. But they can never be the only mechanism. Uber has a rating system both ways. Passengers rate the driver and – less known – the driver also can rate the rider.
The control design and control effectiveness will have a considerable impact on the feedback loops shown in the figure above. This is how you tune (and fine tune) your platform!
What does it mean for your innovation?
There are different ways to get started. Start brainstorming of the platform business model in one of these ways.
On a more abstract level, platforms
- facilitate better utilisation of underutilised, existing (physical or intangible) assets.
- facilitate better use of idle human capital.
- connect already existing businesses to (new) consumers in new ways reducing costs.
- Uber helps utilise idle cars (physical assets) and the human capital of the driver.
- Development platforms (e.g. app stores) help to utilise the idle human capital of the developer and intangible assets (apps).
- Google helps idle webpages (intangible assets) to be used (read) by those who derive value from the content and thus entices the creation of new content (use of human capital)
- WeChat connects banks, doctors, etc to consumers (i.e. existing businesses)
With the above in mind, you could ask, which asset (tangible or intangible) related to your business are underutilised. As you have figured by now this is not limited to assets under your control. E.g. if your industry is known for underutilised of some sort, you could ask yourself how a platform could help to utilise better the assets of your firm and your peer companies. But you could do the same with, for example, your vendor’s assets. iTunes helped to sell the music (=intangible assets) of the music labels.
Where can the different sides of your platform come from?
Look in your wider stakeholder circle. Customers, direct vendors, second tier vendors (if you want to see if you can do without the first-tier vendor), service providers, contracting providers, peer companies (=competitors, e.g. if industry profits suffers due to a mono/oligopoly of dominants suppliers), infrastructure developers/operators, alliance partners, media, governmental bodies, NGOs, consultants, legal / accounting providers, and many more. In this context, always remember we are talking about a group – never about individual companies. A side is always made up of similar constituents.
Take action now & boost your innovation skills
The platform business model is highly dynamic. And it is difficult to predict dynamic systems. I believe the best way to get started at this stage is to brainstorm the fundamental elements of the platform business model.
Brainstorms the below & add to the table provided:
- Which sides could you bring together in a potential value-adding way (start with 2 sides)?
- How can you create value?
- What are idle (physical or intangible) assets or human capital in your industry? Or in an adjacent industry (up / downstream the value chain) or related industries? What could be some ideas to better utilise the idle assets?
- Can you reduce transaction or search costs between the sides?
- What is your true value creation to actively steer the points under 2a and 2b?
- What are positive network effects that can amplify the value creation to the involved sides?
- What are the economic benefits that your platform idea would provide?
You haven’t found a platform idea yet? I’m glad you are honest. It’s unlikely this will happen in a few minutes (though you might have a brilliant hunch). Aim to get your thinking started for now. Take 30 mins to do the exercise above. From there, your subconscious mind will look out to add to the above. Start asking questions to your colleagues that may help further your thoughts, look for company or industry data, If you haven’t read our first article on the platform business model, check it out for a lot of examples.
Check out the next part of the platform business model which covers its vital elements.
If you liked this article, please share. I really appreciate it: