Thursday, January 12, 2006
What do Google�s use of markets to disrupt advertising, Lego harnessing prosumers to amplify innovation, and connected consumers self-organizing into networks on MySpace all have in common? They�re all nascent examples of edge competencies.
Management thinkers and economists have long speculated that cheaper and cheaper information would disintegrate value chains into more and more highly specialized segments. Less often, they�ve considered the possibility that as value chains disintegrate, value creation might begin to shift outside the boundaries of firms themselves.
Yet this is exactly the world that�s emerging. Because coordination is becoming cheaper, the universe of value outside the boundaries of the firm � the resources, activities, and skills external to it � is beginning to explode: different organizational forms, like markets, networks, and communities, are beginning to emerge as powerful economic forces.
In this new world, strategy is shifting. Cheap coordination isn�t just a commoditizer � it offers radical innovators enormous opportunities as well. The same economics that give birth to a new universe of value creation external to the firm also make it possible for firms to leverage that value creation: to use it as a multiplier to the traditional activities of the firm.
Firms which can leverage cheap coordination � the new market leaders � are using the new economics strategically, to reshape their industries on their own terms. They�re building new sources of advantage which are discontinuously more powerful than yesterday�s � sources of advantage that can lead to not just supernormal returns, but sustained, explosively supernormal efficiency, innovation, or growth
In a world of cheap coordination, the edge - the boundary between the firm and the external - is the new core. That's because, counterintuitively, the strategy that dominates the shrinking core is to leverage the edge: not to simply build complementarities between internal sources of value creation, but between both internal and external sources of value creation.
That is, edge competencies are focused on learning how to utilize the universe of value outside the firm � leveraging value creation external to the firm, and, in many cases, external to all firms.
Where core competencies are the firm�s sets of deep learning about how to link internal sources of value creation, like resources, skills, and activities, edge competencies are sets of deep learning about how to link external modes of coordination, like markets, networks, and communities � sometimes to the firm�s resources, skills, and activities, and sometimes directly to each other. Edge competencies are about knowing how to most productively use cheap coordination � often, by remixing and recombining these new modes of coordination.
In many cases, these newer modes of coordination � and their offspring, as yet unseen � can undertake many of the firm�s key functions more efficiently at the edge, and so the dominant strategy is to leverage them to multiply returns to the firm�s key economic functions. It�s these modes that are driving powerful new sources of value creation, like search, peer production, viral marketing, collaborative branding, social networks, connected consumption, and prosumption.
Why are these modes of coordination hyperefficient? There are several reasons that obtain under the right conditions. First, they can build and maintain resources and undertake value activities far more efficiently than firms can. Second, they can allocate the firm�s resources more efficiently than execs, meetings, and memos can. Third, they can build and store capital that�s difficult for firms even to access, like social capital.
The bigger reason is that this hyperefficiency is difficult to achieve internally because many of the sources of value creation they enable are deeply at odds with the firm�s relatively inefficient coordination. Most of these sources of value creation wither and die when they�re traditionally �managed�: they are regulated by kinds of coordination that simply can�t be achieved within the traditional boundaries of firms. Rather, they self-organize and self-regulate outside the boundaries of firms.
Edge competencies let firms internalize the manifold gains from these new modes of coordination, which are traditionally beyond the firm�s horizon � players who have built edge competencies have learned to marry, remix, and recombine markets, networks, communities, and, crucially, the firm, in new, radical ways.
Sometimes, these gains can be simply economic � discontinuously powerful productivity, innovation, or growth. Because edge competencies let firms achieve internal as well as external leverage, they let firms tap productivity, innovation, and growth emerging outside the boundaries of firms. That is, they often let firms discontinuously more powerful productivity, innovation, or growth than they can simply by internal leverage.
More powerfully, these gains are strategic � edge competencies let firms achieve a kind of judo, leveraging cheap coordination to build new, durable entry barriers, switching costs, and market power, in their value chain segments, while slower, older, intertia-ridden players trapped upstream or downstream continue to get hypercommoditized by fallen entry barriers, vaporized switching costs, and eroding market power.
Edge competencies don�t mean that firms will vanish, as their key economic functions shift to the edge. Rather, it�s that the universe of value outside its boundaries can multiply the firm�s returns. For many of the new market leaders, this has meant integration across the value chain � not disintegration.
Edge Competencies in the Real World
In practice, then, edge competencies are about realizing complementarities between the firm, and the newer, relatively hyperefficient modes of coordination cheap coordination makes economical � like markets, networks, and communities: about leveraging complex combinations of markets, networks, and communities, to create value in new, hyperefficient ways.
Consider search. The dominant design for search, like Google�s PageRank � essentially citation analysis � depends critically on networks of connected consumers linking to and citing others. Methods of search that were solely focused inside the firm � like Yahoo through the 90s, having centralized editors annotate each and every corner of the web � were dominated by this new design, which leveraged external resources � the value created via links by connected consumers themselves.
But in isolation, this network wasn�t a powerful source of value creation. Google�s key insight was to combine the network with a market; because both were complementary forms of coordination, discontinuous, explosive growth was the result.
Or consider social networks like MySpace. Networks are revolutionizing marketing, letting smart marketers and brand strategists not get commoditized by consumer hyperfragmentation, but to leverage it instead � achieving huge efficiency gains by boosting marketing returns, achieving huge innovation gains, by co-creating brands with connected consumers, and shifting to new models of branding entirely, achieving supernormal growth.
But networks don�t exist within the boundaries of firms. Though firms may own the infrastructure, like Fox owns the technology behind MySpace, and it�s relationships with consumers, and so it may capture a share of the value created, that value is still created, organized, managed, and regulated by connected consumers themselves.
While these examples make building edge competencies sound simple - after all, many firms are slowly realizing that markets, networks, and communities can multiply returns to their key economic functions - doing so is, in practice, more complicated.
Many firms are simply diversifying into the edge - collecting scattered markets, networks, and communities. But Bubblegen's research has shown that this strategy is dominated; Yahoo, who exemplifies this strategy, is seemingly always a distant second. Yahoo isn't building edge competencies; rather, it's simply acquiring more and more space at the edge.
Rather, as the examples above demonstrate, building edge competencies requires firms not to simpyl diversify, but to actively learn how to create complementarity at the edge - to learn how to create and capture value through the deep synergies that can be realized from unique, carefully chosen combinations of markets, networks, and communities (and, often, the resources, skills, and activities of the firm).
Building Edge Competencies
What concrete steps can players take to develop edge competencies? Bubblegen has distinguished between three edge competences driving explosive success: plasticity, liquidity, and knowledge pools. They, and Bubblegen�s edge competencies practice, are outlined in this presentation in more detail.
To get started thinking about edge competencies, ask yourself:
What sources of value creation are emerging outside the boundaries of firms in our industry?
What forms of coordination - markets, networks, communities, others, or hybrids - are these sources of value creation embodied in?
What gains could we realize from remixing and recombining these modes of coordination?
For example, can we leverage a market, network, or community to atomize the key resources and capabilities of upstream or downstream players? What are the costs and benefits of making this move?
You may be interested in my writings on edge competencies, which I later renamed to call them strategic delivery points...
// Victor Lombardi // 5:25 AM
Thanks for the links. I enjoyed reading your ideas.
I think my notion of edge competence doesn't have too much in common with your thinking - it's more about how cheap coordination enables new resource transformations.
enjoyed the preesntation, but salightly unfair comparison between yahoo, google et al and wikipedia.
in yahoo and google you measured correlated revenues to number of users.
however in wikipedia you correlated number of articles to microchunks, and hypothesised combinatorial scale.
however if the relationship between articles and revenue is not linear but, say convex, then wikipedia may be very much further from demonstrating combinatorial effects.
What you're saying about measuring different thing is certainly true.
The reason is that I am explaining scale economies which operate at different levels - at the peer or user level for Yahoo + Google, and at the content level for Wikipedia - so we have to measure different outputs.
Thanks for the comment.
Different from media industry, we see a different picutre in software world. For example, http://news.com.com/Software+start-ups+feel+the+pinch/2100-1012_3-6026171.html?tag=nefd.lede
Any insights from "edge competition"?
Software is where edge competencies began.
The article you've linked is actually a perfect example of competition at the edge - sw entrepreneurs realizing that discontinuous value creation and capture flowed from leveraging the community (open-source); not the firm.
Even smarter ones are remixing and recombining communities, markets, and networks in different ways...
Key question to me is how the firm creates an incentive structure that can enable it to aggregate, consolidate, distill and leverage skills resident in the community. Traditionally, core competence has been created by deploying employees skills as required, but edge competences, on the other hand, are critically dependent upon having the support of the community. The three canonical examples - MySpace, Wikipedia and Google - have done it differently, and it almost looks like they stumbled into it somewhat serendipitiously. MySpace incentivized with reputation and social recognition rewards for bands, Wikipedia on altruism of knowledge sharing and Google on indirect leveraging of web pages created by everyone on the Interweb. In this sense, I think Zazzle is different, in that it is more about creating a market for prosumers to facilitate trading among them. I suppose this does involve creating some edge competencies, but the scope and nature of these competencies seem qualitatively different from those created by the other three.
I think the challenge for me as a strategist would be to figure out how such competences can be created for someone else. The principles of creating comp ad using core competencies are well understood, but there is a need for similar framework to emerge for edge comp's. What are the strategic/economic principles for creating edge comp? I think Network Econ - Metcalfe's/Reed's law et al - is a good beginning in that direction, but a lot more needs to be understood.
What is quite clear, thanks to your lucid writings, that competitive advantage for a firm in Web 2.0 universe will mainly (only) come from creating such competences on the edge - the internal resources and skills of the firm, by themselves, can do very little to create comp ad for the firm. Even the nature of activities they need to perform is likely to be very different from those in a traditional firm - for example, spend a lot more effort in integrating free software, evangelizing the platform, viral marketing etc rather than make huge bets on creating blockbuster music/movie/software and then sell it in millions. This is a radically disruptive phenom.
// Mahashunyam // 7:25 PM
That is exactly right...and the framework is exactly what I am working on now!!
Ummmm...in that case, forgive me if this sounds familiar, but can I be your research assistant/lackey/intern/photocopy and coffee fetching guy/general purpose bitch? ;-)
// Mahashunyam // 10:47 PM
Very interesting and nicely explained.
Blogged about this article on my site but couldn't find a TrackBack URL. This comment is just a courtesy notification.
Mahashunyam: congrats on snagging the internship!
Yeah, this a naive understanding of the idea, but worth it for the examples.
// phil jones // 2:48 AM
The way I think of an edge-focused business is one in which the edge produces some components and the center produces other components. Certain imperatives seem to emerge, such as:
(1) the edge-produced components must be distinctly different from the components produced by the center; it is not really possible for both to successfully produce the same kind of components;
(2) the two components must be very strongly synergistic - the edge-produced component, by itself, is far less valuable than that component combined with the center-produced component (and vice versa);
(3) the main implication of (2) above is that the central organization must play the role of facilitating the edge providers, not controlling, owning, or bossing them - in essence, the center must provide a structure that gives value to the edge-generated components;
(4) the main implication of (1) above is that organizations cannot normally be successful in mobilizing edge producers to act as extensions of center producers;
(5) a significant implication of (3) and (4) above is that it is extremely difficult to successfully manage an edge-focused business, particularly on a sustained basis;
tsteich336 (at) cox (dot) net
// Terry Steichen // 2:14 AM
Umair - can you share with us your thoughts on how recent work disputing Metcalfe's Law (see: http://qurl.com/rd7qh) affects the conclusions you've drawn in your Age of Plasticity presentation?
// Seamus McCauley // 3:13 PM
That article is making a fairly obsolete argument - one whose subject is telecomms networks; and one whose object is revenues.
My argument is a bit more nuanced; I am talking about resources/capital, and returns.
That said, I would treat my presentation as sets of heuristics to follow rather than a black-and-white model of reality. Reality is always messier, and no matter how nicely we try and fit curves, extrapolation without insight will lead to error.
Umair - thanks for coming back to me on this. I see your point: the limitations of Metcalfe's Law as it applies to the revenues of telecoms providers won't (necessarilly) apply to any other network, because the fact that connections between users aren't equally valuable doesn't address the fact that the variable value of the connections will differ across different types of network. But if the general application of the exponential growth law to networks has been challenged on these grounds, why doesn't it similarly affect the combinatorial growth you posit for more efficient networks? Different (potential) connections in any
network will be of different value - for example the connections hypothetically engendered on MySpace by the addition of more American users is valueless to Koreans if they don't have any interests/friends in common or even share a language. The connections that form the value of Google are quite openly acknowledged as being of different values by the pagerank and other systems that assign diffetent levels of authority (value) to different links. The weakness
of Wikipedia is in its failure to acknowledge the inherent variations in value of links/contributions. I still don't understand how the critique of Metcalfe's I referenced doesn't limit the impact of the network effects you put forward, given the fragmentation of networks into sub-networks it implies. (And admittedly my ongoing ignorance hardly obliges you to continue debating the matter! But I'm hoping you'll humour me with elucidation of how the Haque law of networks stands up to what seems to me the demolition of Metcalfe's.)
// Seamus McCauley // 4:26 PM
I think you are taking this stuff a bit too literally.
These "laws" are idealized models of what kind of dynamics we can expect all else being equal.
In the real world, much happens to interfere - all else is almost never equal.
Odlyzko chooses to use a log function to represent much of this "noise" and so moderate growth.
That's perfectly cool. But I think doing so misses the real-world point: for guys trying to really model real-world effects, a simple log function won't cut it.
You're much better off understanding what kind of dynamics will dominate your network, and then adding assumption particular to your situation (ie, influx of competitors in year 3, whatever).
Take your example - it may be that Koreans only benefit Americans on the net if Germans are there too. This is not as simple as using a log function to attenuate growth in a single model - in the real world, if real money rides on these decisions, you should model each one separately blah blah.
The whole point is that during the bubble, people used these models naively - for example, extrapolating huge growth for a decade. Odlyzko's new model again runs the risk of doing the same again (by telling us a log function captures all the noise of the real world).
I don't think it does. Rather, I think investors and strategists should be cautioned that the model will not = reality, except for those that get things really right, like Google or Myspace).
I hope this helps - basically, I think Odlyzko's stuff just muddies the waters. The whole article is a bit of a mess if you ask me (esp the Long Tail stuff).
Thanks for taking the time to clarify that Umair.
// Seamus McCauley // 5:11 PM
I have just a single line of comment on what has been said & discussed here, I guess I have said it here before. We should always keep in mind that
"All models are wrong but some models are useful" which is a tautology btw.
I have a really basic question. I'm not comfortable with many of the terms in this presentation/discussion and I *really* want to understand it. For starters, is there a good reference for the use of the terms "user" "peer" and "microchunk" as used here? I did some searching and I'm having some trouble getting a clear and consistent view of the distinctions between and attributes of especially "user" and "peer" in this context. I've got a bit of a sense of what is meant by "microchunk" from reading some of your other writings and elsewhere, but anything you can add would be appreciated. Sorry to be so behind in my 2.0 jargon.
// David Beckemeyer // 7:41 AM
I didn't get your presentation.. The post I kind of understand. Are you saying that every function in the value chain needs to become an open platform? If so, I have am interesting presentation for you.
Very curious what you think..
Your thoughts about edge competencies are on the money, Umair, but the slide with William Goldman citing Jaws (1975) as the beginning of the phenomenon has the year (1982) correct but the film wrong. The name of the film that best aligns with your observations is not Jaws. It is TRON.
TRON was pure edge competency. The film itself did not come close to a profit upon its initial release. But when you begin to total up the wealth it has generated through the various extensions of its competencies it comes well into the billions of dollars.
Disney was not prepared in 1982 to recognize or capitalize on its own edge experience. All the talent who expressed or explored competency in TRON got away from the Disney brand, most crucially John Lasseter, who, when he could not find a home at Disney for his aborning interest in CGI, went north looking for it. Disney had to buy back Lasseter's competency at a huge premium.
The most important attribute a brand can have in the Networked World is the ability to recognize and play the edge games without expecting specific outcomes. It requires a lot of courage, discipline, confidence and the ability to improvise.
Thanks for your post!
Promptly saved the file, and shared it with many people. Nice stuff. Liked the Yhprum's law too. Thought it could do with a bit of rephrasing: "if something can go right, it will". That's another key pillar of the demand-side economy. This fundamental belief, made possible by the networked economy, has interesting implications for suppliers, which is what our clients are. I see a world where demand and supply are very evenly matched. Exactly as many supplies as there are demands. Ah, the future!
// Suresh Manian // 7:33 PM
Post a Comment