The world of platforms and ecosystems has come alive recently. If you go back five years there were very few people thinking and writing about them. That’s not the case now. So where have we come from and where are we going?
In fact, I’d like to pose three questions: What is a platform? Can they be designed? What will be their long term impact?
It might seem like that first question has been answered over and over but I don’t think so. People use the terms platform and ecosystem as if we all share the same meaning. In reality more often than not people mean different types of entities and dynamics when they use the term platform, and that becomes especially difficult in the case of ecosystems where I think there is no consensus on a meaning.
In addressing these questions I hope to give a new insight into platforms and ecosystems. I’ll conclude with some thoughts on what’s next for platforms and geopolitics and post a longer version of this article on Medium and on my platform website. I’ll be covering some of these issues in my platform workshop in Boston in February.
￼If you just want the predictions then scroll to the end.
A brief history of platform thinking
The early attempts to analyse platforms were driven by anti-trust policy concerns, or put more bluntly, how do you regulate a company like Google? That literature predates what I think of as the modern business platform and focuses instead on companies like Microsoft. It is highly specialised (it won a Nobel Prize for Jean Tirole) and didn’t really penetrate out to the wider business community.
Nonetheless it needs to be understood because today’s platforms definitely stand in contrast to the platforms that many writers have been analysing. At least the primary concerns have changed.
That’s one reason why people new to this area get confused. much of the literature is focused on a different phenomenon.
Analysis during the period of the 1990s to 2007 focused on two-sided and multi-sided platforms. That emphasis has carried over to today but as I said just now the primary concerns have changed. Nonetheless much of what you will read is infused by these old concerns and it can get you scratching your head.
In the 1990s and early 2000s, economists wanted to define policy actions to compensate for what they saw as poor market mechanisms that were allowing platforms like Microsoft’s operating system and Intel’s computer chips to dominate global computing.
Alongside this anti-trust strand of platform thinking there was a kind of revelation in the 1990s that businesses were becoming more cooperative rather than competitive. The idea of cooperating to compete began to appear along with terms like co-epetition.
The legacy of that line of analysis is that we tend to talk about platforms as being two-sided or multi-sided (this seriously limits our ability to understand ecosystems in my view), and as having network effects (they generally don’t and thinking that they do leads us to underestimate the power of content).
The two major advances of that early period were James F. Moore’s pioneering work on the business ecosystem (already emphasising the horizontal nature of the reach of companies like Apple and IBM) and Gawer and Cusumano’s Platform Leadership (focused on Microsoft and Intel). Iansiti and Levien were ploughing the same furrow successfully too.
Moore’s work on Business Ecosystems is very far-sighted. He sees the ecosystem as an emergent phenomenon I many areas of business in the 1990s and suggests that eventually it becomes part of a co-decided, cooperative strategy between participants across industries.
Think for example of all those companies researching new products in computer storage. While they can all start out with crazy ideas, eventually one or two paradigms will emerge each containing a consensus about hard storage, random access or flash memory. From here on, it pays to work within those paradigms rather than assume there are other emergent paradigms that will get you faster to market and (my interpretation here) if one company happens to have an operating system that eases the way for one of those paradigms then you have the beginnings of an ecosystem with a platform.
However, the environment Moore was describing seems to me to be what we would now call an innovation ecosystem.
A business ecosystem, like its biological counterpart, gradually moves from a random collection of elements to a more structured community. [Text Wrapping Break]
Steven Klepper pointed out during that same period that most industries have a burst of innovation activity that usually ends in oligopoly as the new champions buy up competition and reset barriers to entry in an industry. It seems to me this is what Moore is describing.
Moore’s work was indebted to the contemporary anti-trust concern about the impact of platforms and cooperation.
You might say by now that the monopoly horse has well and truly bolted. But these early works are essential reading for anybody who wants to understand business ecosystems.
The later period of platform thinking
The view Nick Vitalari and I took in The Elastic Enterprise was that modern business platforms and ecosystems differ from those described by researchers from the classical tradition in that the new business platform was not part of a well-marshalled phalanx of cooperating companies such as those in the Wintel ecosystem.
Nor were they functioning within the norms of classical economics and nor were they forming new structures in the way that many innovation ecosystems do, typically through oligopoly and high barriers to entry.
Nick and I were describing a situation where economic activity took on an entirely new vitality – for example the apps economy that grew off the back of iOS and Android is thought to have created well in excess of 2 million jobs in the US and nearly 5 million in Europe in the space of about 5 years.
The new platforms were able to scale like no company had ever scaled before and they were able to marshall ecosystems substantially bigger than anything we might have imagined was possible. There was no need to buy up the ecosystem or to raise barriers to entry – in fact a rule of platforms is that barriers to entry have to be low.
The new business ecosystem signalled a break with old economics.
In that mix are tens of thousands of new companies that have come and gone, as well as many hundreds of thousands of freelance workers. We have also seen the economy become more polarised as platforms have grown in scale and as small companies have become more globalised.
Rather than becoming a more structured community the companies that survived this process have done so through the development of powerful content ecosystems that create and distribute literature that supports progress, learning, tools’ dissemination and innovation. Content, I believe, replaces structure.
The new descriptive writing about today’s business platforms goes back to this period 2011/2012. There has been Alex Moazed and his team at Applico, probably still the leading company when it comes to some form of app-based platform strategy and implementation. Phil Simon had written The Age of The Platform in 2011. Vitalari and Shaughnessy published The Elastic Enterprise a few months later.
Over the next couple of years John Hagel produced some interesting overviews of the impact of platforms and that in turn led to the idea of platforms gaining wider currency and seeming more mainstream even though 2010/2011 was probably peak platform if judged by the number of companies that rushed into some kind of API program and tried to develop business ecosystems among the developer community.
The period 2013 to 2016 was pretty quiet on the platform analysis front, at least in terms of work that was aimed at a wide audience. Academic writing on platforms was backfilling the analytical paradigm on issues like price and growth.
In 2016 the Harvard Business Review ran a series of articles on platform business models with articles by experts such as Evans and Schmalansee (see below), and Alstyne and Parker. A weakness of the HBRs series is the sense that we know more than we really do but at least there was some mapping of the problem domains. A second issue was the US-centricity of much of the material. Platforms, I believe, cede the future of growth to China and we should be looking to how Chinese platforms perform.
I can think of a third issue. In my experience platform strategies are very organic and there is no formula for how to create them. When I do consulting projects I emphasise small steps to big changes, organic moves that draw people into new ways of thinking and working.
But in the business world, after a few years of the apps economy, “platforms” suddenly became synonymous with Airbnb and Uber.
Ironic then that as Airbnb and Uber were enjoying growth success, the world of platforms was being turned on its head by BitCoin and following in its wake companies like Ripple and Litecoin.
Platforms are essentially disruptive so it should be no surprise that the platform environment was being disrupted just as it seemed that that these two companies seemed to offer a simple model of what platforms are all about.
Around 2015/2016 there was the beginnings of a global perspective. The Centre for Global Enterprise produced a global survey in 2016; Zennon Kapron and Shaughnessy wrote about Chinese platforms and global disruption; and more recently the topic has beed taken up from an African perspective at Fibr and I have been attempting to keep some of that work updated since. I’ll be discussing it again at the Austin Forum in Texas on January 10th.
In 2017 there’s been a rush of new work on platforms. In general it is fair to say that much of it is a reflection on the longer experience of platforms in the pre-BitCoin world, beginning with Platform Revolution; and taking in works such as McAffee and Brynjolfson’s Machine Platform Crowd; Evans and Schamalensee’s Matchmakers; and Reillier and Reillier’s Platform Strategy.
Paul Hobcraft and Jeffrey Philips have also switched their work to platforms and ecosystems and Simone Cicero produced his platform design toolkit. Many of the major consultancies have now moved into the field with McKinsey and Accenture helping platform thinking to become part of the consensus.
With all this brain power focused on the platform it is perhaps surprising that we are still in the consensus forming period. But it is true to say that right now there is no real agreement on what a platform means or what an ecosystem is. In its place we revert back to that early period and talk about two-sided and multi-sided platforms or marketplaces and therefore miss some important strategic tools.
It seems to me it is time that we put some basic definitions in place so that we can debate the important concepts and move the debate on. I’ll come back with a new post on what I think platforms are and what is next for platforms. But for now I think it is worth saying that the early and the new are too mixed up. Old anti-trust concerns continue to impose an agenda on new business platform writing but times have moved on.
What is a platform?
One problem has been a lack of agreed definition. Platforms can refer to:
- Software platforms like the many that Microsoft has developed over the years including of course Sharepoint.
- SaaS or software as a service that provides centralised software functionality.
- The matchmaker market type platform like Uber or Airbnb.
- Content platforms like Forbes of Huffington Post.
- Open API programs.
- A venue for joint action.
The real platform though is none of these and yet contains a bit of each. I think the best way to understand platforms and to see where they are headed is to look at some history.
The kernel of the argument in those books is that Google’s ecosystem consisted initially of the search engine marketing and SEO communities. Without their advocacy Google would not have become the search engine of choice. Equally, by providing a group with an opportunity to make significant business income, Google won over the nascent community of search engine marketers and found itself a community of advocates.
Note that this is not a two-sided market. SEO experts where not transacting through the Google platform. In fact, initially, nobody did. As AdWords and AdSense grew, however, the seller was a content owner and the buyer was an advertiser.
The SEO community was essentially an advocacy grew, a group of experts, if you will, who purported to do this Google thing better than anyone else. They were super-users.
SEO experts, however, constantly distorted Google search returns. Their engagement with the platform essentially compromised Google’s number one promise – to deliver relevance.
Over a period of 7 years this was an uneasy battle with Google changing its algorithms to thwart its ecosystem superusers yet also to keep them in the game. At the same time the capacity for content creators to write to the algorithm became an important part of how web content was created. We all now write for SEO.
That is a very short summary! But what can we take from it?
You can see in it that the platform does not need an API, though APIs help. Nor does it need a positive relationship with all of tis ecosystem or any real “relationship” at all. Google was almost wholly hands off in its relationships with the vast majority of advertisers and content owners. It was business through terms and conditions.
But Google was solving an essential online problem. Some people think that problem was the search relevance one but it is doubtful whether Google was ever more relevant in its search returns than search engines like Alltheweb. The problem it solved was monetising web content and by extension monetising search.
The platform owner has to be adept at ecosystem management at scale. It has to find ways to replace relationships (in Google’s case by giving away tools like analytics). Google was both adept at managing SEO practitioners and at extending its ecosystem over time to include bloggers and other content makers, and then of course maps, pages and so on.
But to summarise:
- Platforms are transaction engines, particularly around micro-payments.
- They function through terms and conditions, not relationships.
- They need low barriers to entry because they rely on scale.
- They have to show a clear route to monetisation for their ecosystem members.
- They need to shape market conditions and Google’s example of creating bespoke tools for an ecosystem (Google analytics, Google translate, Google trends are all good examples).
- They need assets. Google’s future would not have been so great had it not invested in expanding the online content universe. Search would have been in limited demand without the willingness of millions of people to create content.
- They need advocacy.
- They don’t shy away from adjacencies.
I also made the argument that search gave Google the essential experience it needed to make Android work and core to that, and to Apple’s success with iOS, was previous experience of being a transaction engine.
With Android and iOS, Google and Apple trailblazed the first major disruption of industry structures by modern business platforms, becoming horizontal players rather than resting in verticals, a point Accenture and McKinsey now make.
Airbnb, as I said here on Linkedin recently, has, similar to Google, set about developing a complex ecosystem, with no open API, and has played to a number of other strengths such as raising the quality and diversity of content, maintaining very low barriers to entry and extending the range of business opportunities.
From this and from examples like Airbnb I see these as the essentials of a platform:
- Smart platforms don’t follow a set of design rules. They are able to orchestrate strategy on the fly.
- They restructure markets horizontally and they enable exponential growth. If a platform is not going to create substantial growth then something is wrong with the strategy.
- Content is critical to platform development for without good content you have no advocacy and advocacy is perhaps the single most important gain from being in a globally connected network.
- Platforms are essentially seamless transaction engines exploiting micro-payments.
- They also have a culture of restless technical innovation.
Can you design a platform?
The great platform stories have all been about great journeys. From Google’s construction of a search marketing market and the global restructuring of the content industries, to Apple’s initial rejection of the app idea on the iPhone, the CEOs of Airbnb crashing on the flow of friends as they test their market and model, to BitCoin, to Ripple’s long struggle to establish the Internet of Finance, these companies invent. They constantly invent how their new markets will function. So it’s not really about developing a product or designing a platform so much as imagining how the market can changed what role you can claim in shaping it.
The question is whether you can plan that out in advance? Can you set a strategy that will stand the test of time?
I think there are several points you can explore in strategy setting that help alert everyone to the options you have downstream but designing the future has never been so difficult.
If you are doing SaaS or PaaS, the answer is yes, you have a strong community of fellow practitioners to learn from, and they are generous with advice. Sites like Saastr are an amazing peer resource for entrepreneurs who want to build out software as a service.
The record of companies that set out to design a platform though is mixed. The reason for that, I believe, is they tend to think a plan is their most important asset and they plan for yesterday’s business environment.
A good example of that is GE. GE has tried three platform plays: ecomagination, healthymagination and Predix for industrial data.
Predix is at the most high profile data platform yet created. It is supposed to collate data from turbines, meters and other sources industrial machinery. Interestingly, when I spoke to competitor Rolls Royce recently their view of Predix was: wait and see.
Every industrial machinery company these days collates data. Rolls Royce has to collate data on every engine for 40 years or more and it uses that data to extend “miles in the sky” or operating life.
Dan Woods explains what Predix does differently using locomotives as an example:
What Predix does for a locomotive is enrich its cruise control with [Text Wrapping Break]something like Waze, but focused on far more complex systems than [Text Wrapping Break]commuter traffic. Now each train can benefit from what happened on [Text Wrapping Break]every other train, including rail, environment, weather and engine [Text Wrapping Break]performance data and factor in that information. This works because [Text Wrapping Break]information is gathered from the edge systems of the entire fleet of [Text Wrapping Break]locomotives and sent to the Predix cloud, which can find insights in [Text Wrapping Break]the information and send data and suggestions back.[Text Wrapping Break]
Now that sounds like an incredible platform opportunity. However, it contributed to CEO Jeff Immelt losing his job. It required a level of faith in software capabilities that might just be misplaced.
Ultimately, yes, these types of data services may offer the marginal but significant gains Immelt claimed for them but if he had had more familiarity with software production he may have been more of a sceptic.
The point is you can plan a platform and find the theory just doesn’t work out. Then you are stranded.
Zalando, the fashion retailer, announced two years ago that it would become the platform for the fashion industry.
in the future we want to connect every stakeholder in the fashion [Text Wrapping Break]world: retailers, stationary stores, advertising agencies, logistic [Text Wrapping Break]services, app developers and many more.[Text Wrapping Break]
There’s no question that Zalando runs a great IT shop and that many of its commerce features rival those of Amazon.
What Amazon seems to have learned over the years though is that conventional commerce is itself the problem. The huge logistics costs and cash management problems of commerce don’t weigh down companies like Alibaba.
A true platform like Alibaba had the opportunity to create data-insights into its merchant and customer investment and spending needs so it quickly became a financier with the capacity to price risk based on data banks could only dream of.
This kind of adjacency move is typical of a platforms. It’s also fair to say that technical excellence is also an important feature. But being the connection point in an industry is a tall order, theoretically fine but difficult to pull off.
It is smarter to look for the small steps that will take you to big change, as must have been the case when Alibaba offered its first merchant loan. Just as Google didn’t start out to be a mapping company or s source of traffic news or a video upload site, Alibaba didn’t set out to be financial services behemoth.
Taking these adjacency journeys is integral to being a platform. It is about being alive to how a dynamic set of relationships can change and how to influence and shape it for good outcomes. It seems to involve entrepreneurs in asking questions like: what change can I make that will restructure this industry in my favour, rather than saying – I want to be the platform of choice.
What next for platforms?
I think the story of platforms from here on is about geopolitics and cryptocurrencies. The platform is becoming integrated, taking on more services such as finance and logistics, and it is taking on a new character: participatory capitalism is the next phase of growth and as an economic value system it belongs to China. Scale too cedes the future to the East, unless western economies become much more adaptable and creative.
By participatory capitalism I mean any economic arrangement where the customer base participates in the development of products and is rewarded with features or value that would not otherwise have been available.
In the space of two years China’s DJI has achieved 85% share of the global consumer drones market forcing US companies to downsize in response. In another area of the economy, IoT specialist Ingdan is on course to be the platform of choice for IoT component manufacture around the world (having already launched 16,000 IoT projects and experienced 365% growth through the first three quarters of 2016). These numbers represent hugely dominant performance.
Innovative and very smart, scale is allowing Chinese companies to quickly amass expertise in new areas by interacting rapidly with customers. This “participatory capitalism” means Chinese companies can draw on tens of thousands of consumer advocates to help with iterating new product features quickly, a key factor in its dominance of the drone industry.
By integrated I mean the capacity of a platform to deliver services that would previously have been delivered by companies foremother verticals, like banking and logistics. Platforms have to become more adjacency driven.
China, through Alibaba, is pushing the WTO to consider special tariff conditions for small package delivery so that it can grow trade outside the multilateral system.
Shifting trade to package logistics is hurting companies like Maersk and threatening companies like Deutsche Bundespost+DHL as platforms like Amazon and Alibaba get better at it — in a pilot project in Munich Amazon took 30% of DHL’s business.
Longer term (Amazon is one of DHL’s largest customers) the concern is that platforms will dictate the terms of business. In addition they are just more innovative and will move quickly into alternatives like drones, autonomous vehicles and last-mile delivery options.
For more on those issues I will be posting an article to my platform site soon. The other side of this coin is the cryptocurrency. It seems to me that crypto, especially the ICO, is a way to bind the ecosystem not just in the sense of having common purpose but also common reward.
Cryptocurrencies segue with participatory capitalism. But both also segue with the growth of common pool resources: open source software, open design, open currencies, we are taking numerous small steps towards a different type of economy built on common resources that are not enclosed by any one company.
There is another aspect of cryptocurrencies that people overlook. China’s capacity to trade is inhibited by the lack of convertibility of its currency. Yes, you can always convert your RMB to dollars but in the world of trade companies that do business with China often still have to do this in dollars.
China will use cryptocurrencies to expand its bilateral trade arrangements, spurring even more rapid growth of its platforms.
- In summary then I think of 2018 as a time when platforms become more powerful – the mega platforms I wrote about here.
- The emphasis however will change from platforms to ecosystems and participation; and the new value system of common pool economics.
- Platforms will continue to change how we work leading to a much less structured workplace, an area touched on here.
- ￼Cryptocurrencies will spur many new platforms with a mixed and ambivalent value system at first, as we once again see a struggle between centralised and decentralised business.
- We will begin to realise the political implications of an economic structure that plays to all of China’s strengths.