The Poison Of Presumed Centralisation
On November 16th 2010 an article appeared in Harvard Business Review (HBR), written by James Allworth, about Google’s strategy. The gist of the piece can be gained from these two enclosing sentences; the first from the start, the second from the end:
1. “…But Google may regret the strategic choices that have led to this victory over Apple.”
2. “The Android operating system is, as Google initially intended, untethered to any particular partner. This was a smart way of fighting the opening battles of the smartphone wars against Apple.”
Basically the author is saying that mobile handsets with Google’s Android mobile operating system are out-shipping Apple’s iPhone due to Google’s free distribution through what is called the Open Handset Alliance, yet this openness means there is no control over third parties choosing non-Google services instead, thus diminishing Google’s competitive position.
In my opinion this critique of Google’s supposed competitive strategy is entirely dependent on a context of traditional and centralised business practice. I believe this context is fundamentally questionable. For it to be the case, Google would be displaying certain personalities of traditional centralisation, such as the need to control distribution or directly attack competitors, for example.
When things are centralised, ownership and enclosed resilience is vital; as what you have to hand defines your central unit of power. If Google traded in such a way, the HBR article would be contextually accurate and the opinion based on a valid conceptual construct.
However, Google is not a fully centralised business and their practice is fundamentally unlike many of the companies that commentators would seem eager to pitch it against. Allow me to clarify.
There are three types of organic business structures:
1. Centralised businesses (just as centralised organisms in nature) have a core hub and externally dependent spokes like a spider does. All behaviour, including competitive behaviour, is focussed on increasing the power of the core hub and decreasing the risk of the core hub being destroyed, which would end everything as all vital parts are within the core.
2. Decentralised businesses (just as decentralised organisms in nature) have all vital parts distributed throughout like a starfish does. There is no dependency of any part on any other. In fact, the further distributed it is (like when you cut off a starfish leg), the more the organism continues to grow. The starfish grows another leg, and the cut-off leg grows into a starfish. It’s like Agent Smith in The Matrix. Remember?
3. Hybrid businesses are, unsurprisingly, a combination of centralised and decentralised architecture. In Google’s case, as a hybrid business, there is centralised decision-making and decentralised execution. It is the decentralised execution that positions Google in a different way from their most obvious competitors (who tend to be extremely centralised – Apple for example).
Due to the common misinterpretation of their organic structure that leads to what I believe to be fundamentally inaccurate assessments of competition strategy and risk, I feel compelled to illustrate a different reality than portrayed in the original HBR article, and by doing so, address what I call The Poison Of Presumed Centralisation.
In 1943 Peter Drucker was commissioned by General Motors (GM) to investigate and interpret the secrets of their success. For 18 months Drucker probed and questioned all parts of the organisation and finally published his findings, as agreed he would, in a book called ‘Concept of the Corporation’.
Due to the findings GM were very angry and Drucker was very surprised at this reaction. After all, in his book, Drucker had praised GM for their way of working, even likening them to the US Government’s ‘Federal Decentralisation’.
Drucker said: “In Federal Decentralisation a company is organised in a number of autonomous businesses.” Just as the US Government ceded power to the states, GM let go of central power to autonomous, decentralised divisions. Drucker’s advice was for GM as a hybrid organisation to become even more decentralised. He claimed their success was primarily due to the level of decentralisation. Drucker suggested such measures as hard-coding customer feedback into deep strategy.
But no. GM hated it.
Their response was, in essence, “We are at the top of the game so why should we change?”
By the way, the Japanese car manufacturers took a far more proactive approach to Drucker’s advice and the rest, as they say, is history.
Around this time, Drucker spoke of an organisational position that is often referred to as ‘the sweet spot’. The place where organisations or offerings are centralised enough for control and commercial reality, yet decentralised enough for mass adoption and agility.
Google, with their decentralised execution, currently resides in a sweet spot of openness and pervasiveness.
Decentralised execution cares nothing for the supposed ‘risk’ of other players. In fact, the concentration is on creating tools for competitors to be empowered.
Did Google arm their competitors? Absolutely. On purpose.
Decentralised execution cares nothing for ‘market share’ of specific technology. Instead, the concentration is on becoming invisible, yet always there. The point of the story of GM and Drucker isn’t about the reaction. It’s about the empirical competitive advantage of decentralisation.
Despite unarguable evidence, the most common thought is that businesses are similarly structured with selfish centricity. The presumption by most commentators, including those in respected publications, is that companies are as centralised as spiders and compete accordingly. This is pure poison and shows an ignorance of business structure and market dynamics.
The poison is, however, extremely common. In the (absolutely vital) book, ‘The Starfish and The Spider’ by Ori Brafman and Rod Beckstrom, the second principle of decentralisation is: “It’s easy to mistake starfish as spiders.”
The frequency of commentators and competitors mistaking decentralisation for centralisation actually helps the decentralised compete. Put another way: The more that people misinterpret and treat decentralised companies as centralised, the bigger the threat. This is one of the main reasons that decentralised companies will rarely, if ever, correct a commentator or competitor who mistakenly uses centralised constructs in their reasoning. It is better for a starfish that others think it’s a spider.
Despite this reality, The Poison Of Presumed Centralisation can be found everywhere. Another example can be found in an article called ‘The Truth About Google’s So-Called Simplicity’. In this, the commentator writes of confusion over Google’s product range:
“A long time ago, 1968 to be precise, a wise person named Conway wrote: ‘Organizations which design systems … are constrained to produce designs which are copies of the communication structures of these organisations.’ So true: I can see this in products from many a company. Except with Google, there appears to be no organizational structure of the product. Hmm.”
The poison lives on healthily, demanding a level of naivety to exist.
In closing, and to be fair to the HBR article, one thing I would like to raise is what Google’s strategy would be if they actually felt under attack. To quote again from Brafman and Beckstrom, the first principle of decentralisation is: “When attacked, a decentralised organisation becomes even more open and decentralised.”
The message therefore is clear. Not only can we predict less predictability in competitive moves under attack, we can be assured that The Poison Of Presumed Centralisation will continue to empower the hybrid and decentralised. Faced with this reality, who would you bet on to win?
Taken from Business Poison - see 'books' on the menu.