cow Fast Company‘s Seth Godin is writing about purple cows . What’s a purple cow? Something that is remarkable, worth talking about, worth paying attention to, something that stands out compared to “perfectly competent, even undeniably excellent cows”. Here’s an example that Schindler Elevator came up with:

When you approach their new elevators, you key in your floor on a centralized control panel. In return, the panel tells you which elevator is going to take you to your floor. With this simple presort, Schindler Elevator Corporation has managed to turn every elevator into an express. Your elevator takes you immediately to the 12th floor and races back to the lobby. This means that buildings can be taller, they need fewer elevators for a given density of people, the wait is shorter, and the building can use precious space for people rather than for elevators. A huge win, implemented at a remarkably low cost.

Matt Mower of Curiouser and Curiouser thinks Knowledge Management (KM) needs a purple cow, a product, concept or innovation as remarkable in its way as Schindler’s presorting elevator. He suggests KM is moribund, and says “the whole field of KM is dominated by the idea of being good enough“. Matt is talking specifically about KM products, but what he says is true of the whole, newly-boring field of KM. Five years ago, six of the top ten best-selling business books were about KM, and the field was hot: today none of them are. KM gurus are blaming the economy, the unfortunate name “knowledge management”, and each other for the sad state of the discipline. But the simple truth is, nothing remarkable and implementable has emerged in KM in years.

If a purple KM cow could revive the discipline before it goes the way of TQM and BPR, where could we find one? Seth suggests ten ways to raise a purple cow:

  1. Find the customer group that’s most profitable, or most likely to influence other customers. Figure out how to develop for, advertise to, or reward either group. 
  2. Launch a product that does nothing but appeal to, and let you dominate, one underserved market niche.
  3. Create two teams: the inventors and the milkers. Put them in separate buildings. Hold a formal ceremony when you move a product from one group to the other. Celebrate them both, and rotate people around.
  4. Get the email addresses of the 20% of your customer base that loves what you do, and make something extraordinary for them.
  5. Remarkable isn’t always about changing your #1 product. It can be the way you answer the phone, launch a new brand, or price a product.
  6. Test the limits. Ask what it would take to be the cheapest, the fastest, the easiest, the most efficient, the most x.
  7. Think of the smallest conceivable market and describe a product that overwhelms it with its remarkability. Go from there.
  8. Find things that are “just not done” in your industry, and then go ahead and do them. 
  9. Ask, “Why not?” Almost everything you don’t do has no good reason for it. 
  10. Ask what would happen if you simply told the absolute truth inside your company and to your customers?

Think about the different aspects of KM in your organization: intranets, extranets, communities of practice, external database purchases, research, push/pull distribution. Think about the internal and external customer segments for each aspect, and how the ten ways above might apply to create a product, a process, or a tool for one or more segments that is really remarkable.

I’ve pulled together a few possible purple KM cows from discussion with a couple of front-line KM practitioners. I’ll share them here next Friday.

This entry was posted in Working Smarter. Bookmark the permalink.


  1. Chris Dent says:

    I can’t resist. This seems like the perfect time to mention PurpleNumbers as a simple but powerful Knowledge Management tool. See some pointers to more info.In brief, purple numbers are a way to provide granular access to content. This allows for straightforward references. Much of knowledge management is hinged upon having an awareness of (but not necessarily using) handles to information. Purple numbers provide such handles.

  2. Dave Pollard says:

    Intriguing. Thanks, Chris. I know both Wired magazine and the Cap Gemini KM Journal use circles and arrows from key text to the margins, and hence by pointers to other pages of related text. PurpleNumbers would seem to me to be closer to that concept than, say, footnote numbers. Their real value in my view would be as a way of parsing large documents, so that instead of the corporate intranet search engine retrieving an entire document (with search keywords in a different colour if you’re lucky), it would retrieve only the relevant Knowledge Objects lassoed by the PurpleNumber. The big question of course is: Can purple numbering be automated, so that the horrendous job doesn’t have to be done by authors or, worse, taxonomists?

  3. Chris Dent says:

    Lassoed index thoughts: current implementation of Purple and PurpleWiki (reachable in various ways from the links in my previous comment) automate he process of generating the PurpleNumbers. Purple will do it for most valid XML documents (includ XHTML).There’s also something called PurpleSlurple that takes a different approach: it processes existing HTML documents, anywhere on the web, to add the PurpleNumbers: concerns itself with static documents that aren’t likely to change, so it doesn’t matter that the numbers do not stick to granular elements.In Purple and PurpleWiki the numbers stick with the granular elements, so paragraphs can move around.

  4. paulapalooza says:

    i’ve never seen a purple cow i hope i never see onebut i can tell you anyhowi’d rather see than be one-ogden nash

Comments are closed.