Only Dead Fish

Platform Economics (2) - Network Dynamics

I thought I'd write some more about some of the basic economic fundamentals of platform business models (links to the other posts in this series are collected at the bottom of this post). 

It's worth delving a bit more into network dynamics and why (perhaps) the shift from more traditional market dynamics and approaches can be hard to grasp for incumbent businesses. For this I'll draw particularly from this Wharton piece by Barry Libert, Megan Beck and Jerry (Yoram) Wind (who in turn are drawing from their book: The Network Imperative). As always, it's worth reading the original, but I'm summarising my take outs here for my own benefit.

To begin with, the authors have defined four key business model types in the modern economy:

Asset Builders: a value proposition based on physical capital (manufacturing, distributing, marketing, selling or leasing physical goods). Example include Exxon, Ford, Walmart, Boeing

Service Providers: a value proposition based on human capital (providing expertise and services to customers). Examples include Accenture, JP Morgan Chase

Technology Creators: a value proposition based on intellectual capital (developing, selling IP such as software, analytics, pharmaceuticals, biotechnology). Examples include Pfizer, Microsoft, Oracle

Network Orchestrators: a value proposition based on network capital (facilitating interaction or transaction between different members of the network, value coming from building relationships, collaboration, sharing advice or reviews). Examples include TripAdvisor, Uber, Airbnb, eBay

In research conducted in 2014, it's been shown that when this framework was applied to the S & P 1500 (which captures a range of company size), there were notable differences in performance, with network orchestrators outperforming the other models across a range of measures including profit, return on assets and sales growth. 

Networks are of-course easier, faster and more cost effective to scale than physical products:

'Not only are many of the most valuable goods in our market — such as ideas, intellectual capital, and access — digitizable, but also our digital networks allow them to proliferate with great ease. The scaling cost is close to zero. When you add the network effect, where each additional participant (or node) in the network increases the value for every other participant, the network drives its own growth.'

The authors advocate for a recalibration of focus and investment to reflect this shift, but also recognise how hard it is for leaders to move on from decades of entrenched industrial age thinking that places greater value on physical assets. They have a nice summation of the key shift from firm-centric to network-centric thinking:

Whilst network orchestrators still represent a small proportion of publicly traded businesses, networks will proliferate across industries (albeit in different ways), and also blur the boundaries between how we think of different sectors. But the organisational orientation of a platform business is different to a traditional model:

'The companies that have truly built network-orchestrating organizations don’t just do one thing differently; they do everything differently — from leadership to recruiting to production to advertising.'

I've captured this shift before as one from resource control to resource orchestration, from internal optimisation to external interaction, from a focus on customer value to a focus on ecosystem value. I also believe that there is opportunity in creating multiple platform models across the different capabilities of a single organisation, in the same way that Amazon is systematically 'platformising' its different competencies.

This is a big shift. The idea of platform businesses may not be new, but digital technologies have brought a scale and potential comprehensiveness to this the like of which we have never seen. 

Other posts in this series: 

Defining Platform Business Models

The shift from pipeline to platform models

Platform Economics (1) - managing the demand curves on supply and demand side, cross-side and same-side network effects

And there is more in my book of-course.

Platform Economics

I've been writing a bit recently about the shift from linear pipeline to platform business models, in part to help me be more precise in how I define and describe it (to clients), and in part because I think the degree of shift in the dynamics of scalability, ownership, data, and interaction is under-appreciated (particularly in large, incumbent businesses).

I'm reading Andrew McAfee and Erik Brynjolfsson's Machine, Platform, Crowd at the moment, which does a good job summarising some of the big shifts. The a16z podcast above featuring Andrew and Erik is a really good listen, and discusses some of the key ideas in the book. These are positioned in the context of the rebalancing that is needed between the dynamics of mind and machine, product vs platform, core vs crowd. They also discuss a few of the key principles inherent in platform business economics, notably network effects, supply and demand side economies of scale, the changing nature of corporate ownership, the idea of complements, and the red queen phenomena.

To complement this I re-read Chris Smith's excellent two-part post on platform business economics, which set out in simple terms some of the distinctive dynamics involved. As blogging is my way of assimilating and thinking aloud I'm going to summarise some of Chris's points here (largely for my own benefit but as always you're welcome to tag along of-course).

To remind us, platform businesses are essentially orchestrators. They typically have some key characteristics including serving multiple types of customer, facilitating efficient value exchange (between all the constituent players in the ecosystem), and exhibit shared common standards (the terms that ensure the participants can transact easily and fairly). Chris quotes Alex Moazed of Applico who defines two key types of platforms: exchange and maker:

…both platforms that create value and generate revenue by facilitating interactions between consumers and producers. But each platform type has a slightly different core value proposition. An exchange platform creates value primarily by enabling direct exchanges between its consumers and producers. In contrast, a maker platform creates value by enabling its producers to make content and broadcast it out to an audience.

The key thing here is being the facilitator - value being derived from access and interaction (because data). Being a thin layer that enables exchange of value means high leverage but more limited ownership (and as Tom Goodwin has said, that the battle is for the customer interface). The interaction between distinct user types in the platform ecosystem can act as its own market (facilitated by the platform of-course) but this activity strongly affects the other players and markets in the ecosystem, leading to exponential growth through network effects.

The example I used before is Google and Android. Google is the owner of the operating system, and it partners with providers (mobile device manufacturers who serve as the carrier) but it's a win-win relationship (great hardware carrying great software, everybody wins). App developers are the producers (creating apps for Google Play) but again it's a win-win (the more great apps, the more Google Play becomes a differentiator for the OS). Consumers are part of the ecosystem too, but there is a value exchange here too with access to great phones carrying great operating systems and access to a great app store, and money and data flowing the other way. A change in one of these market dynamics positively impacts the other dynamics in the ecosystem e.g. more people buying apps through Google Play doesn't only mean more developers wanting to put their own apps on the platform, but it benefits the OS through more people choosing Android and the devices that carry it.

Chris uses the standard demand curve, which shows the relationship between price and volume, to show how platforms yield network effects. 

With a platform business you have to manage multiple demand curves simultaneously - the supply-side and the demand side - and a change in one is tightly linked to changes in the other. A platform like Airbnb for example, needs to have critical mass on both sides if it is to work. But the more accommodation providers are on the platform the more attractive Airbnb becomes to consumers. The more consumers join because of this, the more attractive Airbnb becomes to other accommodation providers, and the better Airbnb gets:

 

There are two key types of network effects in this context - cross-side and same side. Cross-side is like the scenario above where user dynamics on one side impact user dynamics on the other - for example where realising demand-side economies of scale helps to achieve supply-side economies of scale and vice versa.

Same-side network effects are when user growth on one side of the platform affects other users on that same-side. Chris uses the example of where positive advocacy from great customer experience can enhance user growth, but Chris has a neat table showing how cross-side and same-side network effects can have positive and negative demand curve impact:

Key to the effective orchestration of a platform ecosystem then, is understanding how the demand curves shift in response to each other. 

I've liberally lifted from Chris's post here for my own benefit but worth reading the original. He's also written a follow up that deals with growth dynamics which I'll aim to write up in due course.

Google Firestarters 24 - Deciphering Digital Transformation - The Event

On Wednesday evening we had the 24th Google Firestarters which took digital transformation as it's subject. The event had been well oversubscribed within 48 hours of registration opening, demonstrating the high level of interest that there is around this topic in agencies right now (this was supported by the IPA Future of Agencies research that I did which revealed the disparate range of agencies and consultancies that are coalescing around this burgeoning area). It's a huge subject but as always with Firestarters we wanted to take some different but challenging perspectives on the theme and we had four excellent talks that I think achieved that really well.

First up we had the brilliant Emily Webber, who has worked with people and organisations in both the private and public sectors to develop their agile capability for sustainable change, and gave a fascinating talk on scaling learning through communities within organisations. Developing a organisational learning culture in the context of transformation is something that I talked about a lot in my book, and Emily brought that to life through the lens of enabling better cross-functional progress and learning at scale within a business. She talked about how the number of connections between people in a team (and therefore communication traffic) increases exponentially as the team grows bigger, and how the organisational response to these kinds of scaling issues relies on hierarchy and the creation of functional silos.

Yet learning across the organisation is an essential element in achieving real transformation: 'learning is done better when it's done together'. Social learning theorist Albert Bandura said: 'From observing others one forms an idea of how new behaviours are performed, and on later occasions this coded information serves as a guide for action'. Collaboration is about building on top of each other's ideas, and yet so many organisations treat learning as an individualistic thing. She referenced Simon Wardley's concept of Pioneers, Settlers, and Town Planners, making the point about the different types of people needed in an organisation to bring innovation and transformation to life, and Dan North's thoughts around Delivery Mapping, talking about 'skills liquidity' and the cross-over of business need, current skills and desired skills. Emily's concept of 'Communities of Practice' enables learning horizontally across the business, and a collective intelligence that is greater than the sum of individual intelligence. When combined with a 'safe to fail' environment, fast feedback loops, and an empowered supporting network, people can really fly, it allows for decentralised learning and scales good practice.

Scaling learning through communities in digital transformations from Emily Webber

Tim Malbon followed with a witty, insightful talk on how the process, practice and culture of building digital products and services can support business transformation. In that sense, the product is the means to an end, with the process actually serving to impact a wide range of organisational practices including workflow, resourcing, processes, hiring, training, learning and so on. If transformation is the aim, the product is the how, with multi-disciplinary product teams learning (he echoed Emily's point about how people learn better together, so agency and client staffers are combined) how to prioritise, innovate and iterate - how a lean/agile product innovation team works. This becomes particularly important when you consider that real transformation comes from not only customer experience, but employee experience as well. 

I'd thought it was important for this event that we had a client perspective, from a company that is undergoing digital transformation, and so was happy to have Johan Enstrom, Digital Transformation lead at Pentland brands (Speedo, Canterbury, Mitre, etc). Johan adeptly captured just how hard the change process is in reality within companies, particularly in managing priorities across disparate levels of maturity, knowledge and capability in the business. He talked about the importance of qualities such as honesty and humility in enabling the culture that supports change and inspires creative solutions, and of understanding the nuanced levels of skills needed at each level of the organisation, and how decisions are often hard in this context because they are not always binary choices. But a shared sense of purpose that allows more fluid application of strategy is key.

Our final talk was from transformation consultant Lucia Adams who drew a lot on her experiences transforming a big global media owner. I've written before about Lucia's wisdom in saying that: ‘Digital is 10% tech and 90% human. Organisations talk about digital as if it is 90% tech and 10% human’. Her talk built on this premis, describing seven myths of transformation (drawing a lot from Steven Johnson's excellent book on Where Good Ideas Come From) . Change management, she said, is often an expensive, waterfall beast that treats people as objects to be shifted this way or that and the effort that goes into ‘transformation programmes’ is usually met with an equal and opposite cultural force. Change is not single thing, it is a 'swarm', and yet a leader needs the courage to set a direction. Eureka moments and epiphanies are a fallacy - true insights require you to think something that no one has thought before in quite the same way, serendipity needs unlikely collisions and discoveries, and error sets off the truth. Platform building she said, is by definition an exercise in emergent behaviour.

The event was broad-ranging and extremely insightful so my thanks to our brilliant speakers, to Google for hosting, and all those who came on the night and joined in the debate. As always, Scriberia did a great job of visualising the talks. If you have any themes that you'd like us to cover in future Firestarters, do let me know.

Agencies as Sensemakers

It seems fitting in the week when the ad community has decamped to Cannes, an event which seems increasingly dominated by the ad technology players, to write about agencies and technology. This is also the week of-course in which Publicis Groupe have announced that rather than spend money on flashy awards and events, they are investing in creating an internally-focused, AI-based 'professional assistant platform' to connect their 80,000 employees. It's laudable to want to empower different ways of working and to bring dispersed disciplines and resources together in new ways, but the cynic in me worries that this just another example of an overly enthusiastic focus on shiny/proprietary technology at the expense of focusing on the people stuff that really drives change - behaviours, culture, the very fabric of the way in which you work (this has been a key theme in the digital transformation work I've done with clients over the past years). I'm sure we'll see.

Anyhow, there was some really interesting feedback around this theme that came out of the interviews I did for the IPA Future of Agencies report which came out recently. The overarching metaphor/model I used for the report was Charles Leadbetter's Systems and Empathy matrix, which captured neatly the dichotomous nature of the dynamics in the market right now and the spectrum of capability along which agencies are positioned, from the systems end (characterised by data/technology/operational and business models and efficiences) all the way through to the high empathy end (founded in human insight, creativity). So it is with agencies. Large, traditional consultancies schooled in business and strategic consulting, organisation and operating model design, moving to expand design and creative capability to create more of a 'full stack' approach. More creatively and digitally led agencies broadening out from marcomms and front end into consulting, technology and digital transformation. A disparate set of consultancies and agencies converging on that top right quadrant. 

One of the key challenges in this shifting dynamic (particularly for the creatively-led end of this spectrum) is to understand the new context that technology creates for marketing and advertising. Whilst some agencies seem determined to ignore it, the reality is that this context is changing dramatically. And I'm not just talking about the growth in ad tech. Each year (friend of ODF) Scott Brinker does a heroic job of tracking and mapping the burgeoning marketing technology landscape. In fact 'burgeoning' is something of an understatement. Back in 2011 when he first started mapping the market he was already tracking 150 vendors. That number has roughly doubled each year since then so that in the latest survey which has just been released we now have over 5,000 different marketing technology solutions across a huge variety of competencies. Solutions that are increasingly platformised, enabling marketing teams to plug together a variety of different specialised and vertical solutions to create their own 'marketing stack'. It's a huge degree of complexity that is fundamentally changing the discipline, function and operations of marketing. 

More than one interviewee for the IPA research noted that agencies simply do not have enough understanding about this technological shift (and that clients themselves are struggling with this level of complexity). This is a challenge for agencies, but it's also an opportunity. It may not be necessary for many agencies to understand the operational intricacies of specific technologies (unless of-course that is a key part of their proposition), but there is an imperative on every agency to appreciate the technological context within which they are operating. Work and ideas that make sense in the context of the client organisational focus on all aspects of customer experience, a broader set of customer touchpoints, more diverse brand content delivery, and growing marketing automation and technology-driven execution are critical. 

With so many martech vendors convergence is an inevitability but there is an opportunity for agencies to make sense of new waves of technologies for clients and to distil and enable the realisation of the value that they present. In the report I focused this into three key areas:

1. Potential:-  the ability of agencies to enable clients to better understand the potential of specific technologies and trends to enable better prioritisation of client attention and investment and properly assess strategies and focus
2. Integration:- to understand how different combinations of technologies might be brought together for beneficial effect (and be larger than the sum of the parts)
3. Application:- to appreciate and advise how technologies might be applied for business benefit

As successive waves of new technology (AI, VR/AR, IoT) take hold, this role as ‘sense-maker’ for clients creates further opportunity in the operationalising and optimisation of new technologies and capability. There were a number of agencies I spoke to, for example, that were doing some really interesting work around creating the equity/practices/processes designed to help a client operationalise and scale a new capability around the globe. 

There is enormous value to be had in being the trusted adviser and enabler. But without an understanding of the broader context in which advertising and marketing now operates, agencies will really struggle to play that role.

You can download a copy of the full IPA Future of Agencies report here.

When Robots Bluff

Accepting the (current) limitations of machine learning (and some recent hiccups), it's still fascinating how AI capability is developing so fast despite the field being far from new. The 'AI beating humans' narrative (as Tom Chatfield talked about at last year's Google Firestarters on the theme of AI - describing it as the ‘usurpation narrative of human-machine interactions…a creation is pitted against its creators, aspiring ultimately to supplant them’) is not always that helpful. But still, the landmarks along this storyline (AI beating humans at Jeopardy and more recently the famous victory of Google Deepmind’s AlphaGo against GO world champion Lee Sedol) are quite compelling.

So now we have the AI system (developed by Carnegie Mellon University) that has just accumulated $1.7m worth of chips over twenty days playing against four of the world’s top professional poker players. What was particularly remarkable about AlphaGo's victory over Lee Sedol (which I remarked on at the time) was the way in which AlphaGo won, playing in highly unexpected and creative ways, with moves that left even its creators 'pretty shocked'. Wired described how:

‘…with its 19th move, AlphaGo made an even more surprising and forceful play, dropping a black piece into some empty space on the right-hand side of the board. Lee Sedol seemed just as surprised as anyone else. He promptly left the match table, taking an (allowed) break as his game clock continued to run.’

The commentator at the time noted how Sedol appeared to have trouble dealing with a highly unusual move that he had never seen before. It was a real moment in the development of AI/human relations.

And perhaps we've just passed another one since what’s again interesting about the poker example is that on its way to victory, the AI succeeded in out-bluffing its human opponents. Poker, the FT piece notes, tests different mental muscles since it involves 'strategising using imperfect information in a way that is more akin to the real world'. The AI started not so well, but then focused on fixing its own weaknesses and filling the holes in its strategy as it went. It was able to use an additional feedback loop to respond in real time to the human activity. The implications are that AI can take other imperfect information situations and derive a good strategy for that situation.

Another new milestone.

Image courtesy