Enterprise 2.0 Forum: JP Rangaswami

I’m here at Enterprise 2.0 Forum in Cologne, enjoying the conference even though a lot of it is in German and I am entirely incompetent in the language. Luckily, JP, like me, is speaking in English even though he says he can ‘listen German’ rather than ‘speak German’.

JP Rangaswami
Favourite artist in the UK is Banksy, hidden within his “Graffiti Removal Hotline” piece is the message “Make sure that the cost of repair is kept equal to or below the cost of damage”.

Thanks to mcfer2k

However you implement whatever you choose to use, whether it’s Sharepoint (or Don’tSharepoint), Confluence, Twiki (and people get very polarised about it), you must watch the cost of repair.

Chewing gum – the cost of a stick of gum is about c5 but the cost of removing it off the pavement is c15, because the cost of repair is too high. So Singapore bans chewing gum all together.

Same is true about graffiti removal is the same. The cost of buying a can of paint and spraying a wall is low compared to the cost of fixing it, of cleaning the wall.

The power of Wikipedia lies in how easily you can undo attempts at vandalism, lies, errors – the magic is how quickly you can revert to a previous version. When we implement wikis in business we forget that because we come from an environment of permissions, authorities, firewalls. We’ve built a very complicated world. There’s something warped about how we build walls then tunnel through them all the time.

But the keeping the cost of repair lower than the cost of damage is essential.

Why do people not use manuals? Mainly because they are out of date. The pace of change is faster than the pace of updating the manual. If you know the manual is out of date you won’t use it. Same happens for employee handbooks, guidelines policies. The larger the enterprise the more of these documents get produced. These got moved onto the intranet, but there were few people with the right to edit. So you had a wall around the intranet that you couldn’t get through – you needed special permission and tools to change it. Unless you were the expert you were not allowed in. And people who were allowed in were a small team and the editing capacity of the firm was sharply restricted. They kept the cost of repair high. Cost of damage was low because information decays over time. Even doing nothing to the manual decays the manual, but the cost of repair was high – high cost of access, high barriers to entry.

Whatever you implement this for, watch the cost of repair. Magic of a wiki comes from allowing people to amend things. What if they amend it wrong? Who cares? Even an investment bank can allow people to use a wiki. Why? If you can prove to a regulator that you can capture the date and time that something that was put on, and that you can prove how fast errors were corrected, that’s what makes it valuable. You get a perfect audit trail of who did what and how fast things get corrected.

Whatever the content, it doesn’t matter. Don’t replicate the historical cost of repair. Don’t pave cow paths. When you move from cattle to roads, but just pave the cow paths, you’re just making an incremental changes. Need to carve a new path.

The space shuttle’s rockets are based on the shape of a horse’s back. The place they are assembled is linked by rail to the next assembly place. Rail gauge is related to horse paths. Ergo, space shuttle’s rocket design is influenced by legacy decisions from 150 years ago.

Kevin Kelly. Interesting chap. The internet is a copy machine. Printing presses of 15th century were about two things – cheap repeatability and cheap standardisation. Hand-written manuscripts never look the same and have errors, so every one is different. Add time, distance and culture, and the corruption gets worse.

But now, what we attach to email is just like the manuscript problem. Because we have version mismatch all over the place. Instead of people going to a single source, we are attaching documents to emails. And people spend a lot of time reconciling versions, checking that people have the right version. But we don’t have to do this because whilst we learnt this with the printing press, that standardisation and repeatability is critical, we haven’t learnt it with modern technology. Internet is a great big copy machine – a wiki should not allow people to diverge versions. You have to be looking at the same thing even if you disagree with it. Don’t want 100k copies of the same thing in everyone’s emails.

Doesn’t care which platform you use, but does care that you don’t raise the cost of copying and the cost of transmission. If policies force you to do it, throw the policies away don’t throw the value away. The value of the web is in copyability. Web became 2.0 when it became writable.

We’ve been able to create structured data for years, then came the ability to consume unstructured data via search engines, and now we have he ability to produce unstructured information. When someone comments on a blog they are uploading text – there’s no difference between text data, video data, audio data, other than size and file type. A comment is an upload.

Do not throw away the value of wikis by not understanding that you have to keep the cost of transmission and reproduction low.

Open source question. This has one other value for global organisations. It’s very easy when you have the web as basis for your architecture to change language. Speed of innovation for OpenOffice in different languages outstripped the work done by Microsoft, because the community is interested in solving its own problems. OS people don’t look for the business model, they look for problems they can solve. They don’t ask how they are going to make money or who they are going to hold captive.

Slide: Advise for spies in the war about how to sabotage organisations and production.


But the advice is similar to how many businesses are actually run. E.g. “Insist on doing everything through “channels”. “Bring up irrelevant issues as frequently as possible.” “Haggle over precise wordings of communications, minutes, resolutions”.

This is very, very effective disruption.

Life has changed. His father had one job. JP will have had 7 jobs. His son will have 7 jobs at once. The concept of the contract between employee and enterprise has changed. Enterprises exist because they can get better capital than individuals, but this is changing. Enterprises were meant to have global scale and reach, but we have that as individuals now too. They were supposed to be about security – that has changed so dramatically. There is no security, unions don’t matter, culture of the labour movement doesn’t guarantee jobs. Enterprise is supposed to provide benefits, but these have decayed over last 20 years. So what is the enterprise? What does it mean in the current world?

Youngsters now don’t see things the same way as us. They are used to Google. They are used to ways of working that don’t require them to become institutionalised. 100 years ago, if you worked in a bank you could only use the company’s quill pens because they wanted to to standardise writing.

You can’t tell the upcoming generation that they can’t use their own computer, you have to use the company’s, because that device has become personal. They don’t want to carry two or three devices. So enterprise has to become device agnostic.

But only 60 years ago, people thought the spire piece was sabotage, now they think it’s normal work. Now they are wrong, it *is* sabotage. Wikis allow interconnection, lateral movement, movement beyond the departmental silo (see point 1 of sabotage advice). Point 2 is Twitter, SMS. You can get a lot done in short words. Speeches are not a good thing. And so on.

Have to stop being hypocritical. See a few people with computers. Normally at conferences most people have computers and they are listening to talk whilst they are checking on things, or if they’ve remembered something based on association to what I’m saying, but they are free to do what they want and they do it openly. But in a meeting, they hide the Blackberry and check it under the table, and think no one is noticing. New generation knows when someone is checking a Blackberry.

The things we do in large enterprises today, many of these things would have been considered not just unproductive, but sabotage 60 -70 years ago.


1. Keep cost of repair lower than cost of damage
2. Underlying value proposition of the web is copyability and standardisation
3. We are providing these tools to a new generation who don’t believe in the hypocrisies of the exiting generation

Open Source
DrKW went open source, and now many people are. If a problem is generic, allow the OS community to solve it. Is a problem is a commodity, then the community will scale and find a cheap and easy way to solve it. That’s why OS software tends to be generic tools that are built by people saying ‘this is too generic to be proprietary’. Proprietary = cost.

If a problem is specific to a vertical market, pharma, education, financial services, then go to the proprietary community, because someone will take the risk to solve it, because it’s too expensive to solve for one and the OS community won’t solve it because it’s too specific.

If the problem is unique to your enterprise. No one has an incentive to build it. So look to your own developers. Try to avoid unique problems.

When wanted to sell the idea of OS, we did it with the economics. Generic, non-contraversial, commoditised software is logical place for OS. So don’t deal with politics and emotion, but economics.