Category: Four pillars

Just one. The best

I love chess for a variety of reasons. The sheer breathtaking beauty of the game, as evinced here, in “Fatal Attraction”, Edward Lasker v Sir George Thomas nearly a century ago.  The characters it throws up, as in Jose Raul Capablanca and Efim Bogoljubov. And the way chess teaches us about cause and effect in a complex adaptive environment.

By the way, I’ve written about it before, but if you have any interest in chess at all, do play out Fatal Attraction. It’s a mesmerising game.

One of the reasons I like Bogoljubov is the outrageousness of his statements. In that outrage is truth. Examples: Bogoljubov had just won a remarkable game with the black pieces, and was asked how he’d done it. And he said “When I play white I win because I play white. When I play black I win because I am Bogoljubov.” Another time he was asked how many moves ahead he thinks. His answer? “Just one. The best”.

My interest in chess, largely kindled by an old schoolmate, Devangshu Datta, stayed steady through the years largely as a result of my interest in complex adaptive systems. When it came to analogies for root cause analysis and prevention of recurrence, I found chess hard to beat. I could sit down after a game and work out precisely when I started down the wrong path, what real options I had, how I could make sure I didn’t do it again. Chess was also good as a way of learning damage limitation, what to do when you have made a mistake.

I was reminded of this recently with all the brouhaha about healthcare in the US. Somewhere along the line, the focus of discussion appeared to deal primarily with the effciency of the curing process rather than the preventing process. Too often the same happens in the enterprise world. People are so busy getting better at fixing problems that they forget the real point, which is to stop doing what causes the problem in the first place.

Improving the speed and quality at which you fix things is a worthwhile objective: that is, if (and only if) things break down less often as a result. So when you look at repair processes, it is more important to look at why things break down, and to prevent them from breaking down,  than to focus on getting better at fixing things.

For some time now, we’ve been focused on the customer experience at BT. We looked at the way we dealt with customer requests, how often we delivered what the customer wanted, when the customer wanted it and how the customer wanted it. And we would take a close look at how often we got that right. A very close look. Because it affected what we took home.

That extreme focus has now begun to pay off: that’s why CEO Ian Livingston could tell the world last December that our complaint calls had halved since we embarked on the RFT initiative. Halved. In fact, that’s partly how we’ve been able to cut costs sharply.

The complaint calls coming in were a useful proxy for the number of problems we were causing. But we have to be careful. In large organisations, it is normal, understandable, even tempting to create an environment where the focus shifts from preventing problems to curing them. So before you know it, all the energy is deployed in fixing things, and not preventing the occurrence of the problem in the first place. That’s why you have to be careful what you measure, and how you use the measure. Finding out that you’re solving problems faster and faster is a good thing ….. provided the absolute number of problems is going down, and the problems aren’t repeating. Don’t get seduced by the message that you’re fixing things faster, cheaper better. They shouldn’t be going wrong in the first place.

So look ahead, be a Bogoljubov, and play that best right move. Concentrate on making sure you don’t make the mistake in the first place; the introduction of automation is a commonsense way of achieving this. As long as the environment is in steady state, this should be tractable. When you introduce change, then mistakes can be introduced, found, dealt with. As Esther Dyson says so often, “always make new mistakes”.

Parallel lives

For the next few days, while I am at the World Economic Forum at Davos, I’m going to be spending my time guesting on the Telegraph blog. My first post went up this morning, and can be found here. Let me know what you think.

Walls and bridges: even more on Facebookisation

Whatever gets you through the night it’s alright, it’s alright
It’s your money or your life it’s alright, it’s alright
Don’t need a sword to cut thru flowers oh no, oh no
Whatever gets you thru your life it’s alright, it’s alright
Do it wrong or do it right it’s alright, it’s alright
Don’t need a watch to waste your time oh no, oh no

John Lennon, Whatever Gets You Thru The Night, Walls And Bridges, 1974

Note: The song was Lennon’s first and only US solo number 1 during his lifetime. (Just Like) Starting Over, the only other Lennon single to make it, didn’t actually make number 1 until after his untimely death.

Do you remember the “high street” banks of the UK in the Sixties or Seventies or Eighties? What wonderfully decrepit institutions. A place where 85% of the space was devoted to “admin”; where customers didn’t exist, only account numbers; where the plysical manifestations of these account numbers, the human beings, were sardine-smashed into the remaining 15%; where staff took lunch breaks to make sure queues were at their highest at the only times customers could come in; where bank managers ruled, and considered it an insult if you actually came in to withdraw some money.

Do you remember?

The key thing that struck me during that time was the enormous amount of space given over to the internal workings of the firm, and the tiny area allowed for the customer.  A tiny area that was usually not air-conditioned, smelled of damp and dank, looked like a check-in queue for prison.

You know something? We still have them now. In most firms. They’re called e-mail systems. [Stowe Boyd, come in, I can hear you calling. Time to tell everyone about your Publicy, Privacy, Secrecy. And your love for e-mail :-) ]

Think of e-mail within a firm as a physical space consisting of 100 square units of “stuff”. Then divide that stuff into two parts. Stuff that stays within the firm. Stuff that goes out of the firm, or comes in to the firm: stuff that crosses the firm’s boundaries. And think of these two kinds of stuff as represented in a ratio.

In many firms, “internal” stuff far outweighs “boundary wall crossing” stuff. Check for yourself. In my experience the ratio is close to 9:1. Ninety per cent of e-mail is generated by the firm and never leaves the firm.

Note: I did not count external spam in this measure. Spam is not mail.

Think about it. The majority of collaborative conversations taking place in an enterprise mail world do not involve the customer. Words fail me. Well actually some words come to me. Words like “unbelievable” and “circle” and “jerk”, but then I’m too polite to string those words into a coherent sentence.

Why do I spend an entire post on this point? Because it is really important. Facebook is facebook because it is multitenant. Multientity. Not to do with a single person or home or firm.

Whatever we do in the enterprise, we need to ensure that the walls of the enterprise do not keep customers out. Think Cluetrain.

Collaborative systems are still evolving, people with collaborative instincts are still evolving. Whatever we do, we must ensure that the model of collaboration we build is a holistic one, one that encompasses staff, customers and supply chain.

Building a walled garden in your own enterprise is the equivalent of taking your computer and burying it in six feet of concrete. [On the other hand, since that is precisely what so many firms are wont to do, perhaps I should not be surprised].

More on the Facebookisation of the enterprise

Note: This is a follow-up post to one I wrote a few days ago, The Facebookisation of the enterprise, given the kind of interest it generated. People seriously interested in the subject may wish to read my nine-part series on Facebook and the Enterprise from 2007. The first part remains my most-read post,  apart from the kernel for this blog: Building Society for the 21st Century, which is a page and not a post. You may also be interested in my Twitter in the Enterprise series, a sample of which is here.

If the IT department was made to behave like Facebook, what would an enterprise look like?

You join the “company”. You do this by using a personal token like an email address, choosing a password for all your activity in the company, then filling in some basic profile info. You’re all set. At this stage no one has given you a computer or a phone or anything like that. You’re Generation M. You come fitted with these things as standard. The first thing the IT department would need to provide is simple self-service signup. Access.

Because you’re in a new “social network environment”, one of the first things you do is look for a way to discover which of your friends is already here. So you look around. You need tools to do this looking around. You’d also like to invite the friends you already had into this space. Importing contacts, address books, that sort of thing. So you need tools to do this as well. The second thing the IT department would need to provide you is a set of directories, and ways of adding to them, searching them, extracting from them.

You don’t want the directory to be a firehose, so you want some ways of managing your lists. Making sure that you can group people the way you want to. Friends, family, group, company, department, location, whatever. So the third thing the IT department needs to provide is tools to classify the elements of the directory.

Knowing who’s around is a fat lot of good if you can’t connect with people. So what does Facebook do? It provides you with ways to send people messages, chat with them, converse, communicate with them. Publish stuff, upload stuff, read stuff, view stuff. The IT department must therefore provide communications tools, that’s the fourth thing.

Everything in business happens because people talk to people. [Even black-box trading is people talking to people, but delayed and via proxies]. It helps if people could plan when they were going to talk to each other. Facebook calls these things events. Meetings are nothing more than events. Tools for scheduling events is the fifth thing that the IT department needs to provide.

[Strangely, telcos used to have a stranglehold on the first few items: directories, groups, modalities of communication. But for some strange reason they never bothered to provide scheduling tools. Microsoft were the first to fix this gaping hole.]

Profiles. Directories. Groups. Events. Things published, like links, videos, photos. Relationships between all these things. None of which is static. Which means there needs to be a way of telling people what’s new, what has changed. Who’s joined, who’s left. Who was born, who died. Who joined up together, who broke up. Hatches, matches and dispatches as they used to be called. Which is why Facebook has a News Feed. The sixth thing that the IT department would need to provide is a News Feed. And ways of managing the firehose.

These are just the foundations. When people use Facebook, they use a series of other applications. Applications built by third parties on the developer platform. Applications accessed via Facebook, using the identity and relationship and profile and activity data provided by Facebook. Applications whose access to that data also requires the permission of the person whose profile it is.

Which is the final thing the IT department has to provide: A developer platform with the appropriate controls and service wrap around it.

Access to the environment, directories, ways to group people,  modalities of communication,the ability to schedule events, the publication of records of changes. And a developer platform that allows people to build edge applications that use this core in a safe and controlled way.

Was I talking about Facebook? Or was I talking about the IT department?

Which brings me to my final point. Facebook does not invest in the edge apps, build them, host them, amend them. They don’t support them, maintain them, back them up. I think IT departments would do well to learn from this. Let the people at the edge build what they want, within a 21st century enabling framework. They know what they want better than any IT department can. What the IT department should do is their utmost to guarantee safety and security of access, privacy and confidentiality, search and subscription tools, scheduling tools, data migration tools, visualisation and mashing tools, prioritisation and ranking tools.

Sometime later this month I want to spend time talking about the semantic web, linked data, the Web Science Trust and related subjects. I will also spend time on publish-subscribe and enterprise buses, on augmented reality. On mobility. On opensource. And bring it all back to Platforms and Stewardship.

In the meantime, I’d love some feedback.

Musing about trust

Everywhere around me I see more and more examples of resources, interactions and even entire marketplaces converted into virtual constructs. Abstracted. Expressed in ways that allow for sophisticated models and simulations. In fact that’s one way of looking at what’s happening at the Singularity University.

Everywhere around me I see more and more examples of situations where the core problem people are trying to solve is that of trust. There appears to be a lot of work being done trying to distil trust into something formulaic, data-driven.

And this is good. Data-driven is good. Feedback loops are good. Abstractions and models are good.

But.

There’s something very human about trust. Something more related to the Age of Biology rather than the Age of Physics.

We’ve seen what happens when we rely on mathematics for ratings and values and decisions. Last time round it was called the Credit Crunch. A decade earlier it was called LTCM. Whatever.

You cannot legislate for ethics. Enron would have been SOx-compliant. Basle II may well have triggered some aspects of the recent financial crisis. Michael Power at LSE has been banging on about the implications of “managing” second-order risks for some time now. And he’s been right.

Some of us believe passionately in the power of what’s happening today, in terms of democratisated tools and access and community-based approaches to many things, from home to work to government and beyond. In fact, I’m personally somewhat at a loss as to why no one has really put together the right community-based vehicle for “climate change”, built as an open and transparent platform, on opensource principles and in a global inclusive manner.

Trust is about covenant relationships, not about contract relationships. In a contract you await breach and effect recourse. The question answered is “who pays?” In a covenant the question that’s answered is “how do we fix it?”

I think we’re going to spend a lot of time in 2010 learning about covenant relationships and their role in society. At home. In the community. At work. As a nation. As the world.

Stewardship, my word for 2010, is based on platforms. Those platforms need to be underpinned by trust. Not the trust of physics but the trust of biology. Because that is how value is going to be generated.