Why Best Practices Don’t Work for Knowledge Work

20 thoughts on “Why Best Practices Don’t Work for Knowledge Work”

  1. Essentially, you are saying that reframing a “contextual best practice” as a “lessons learned” makes the information more transportable, and usable at the other end.

    I agree that rote reuse of practices (best or otherwise) is lazy management. But I have also seen too much of the “not invented here” mindset that wasted tons of resources and frustrated staff endlessly because management refused to consider ANY practices from other industries or peers.

    Balance and thoughtful consideration of current needs and context wins over knee jerk every time.

    Enjoying your posts!

  2. Luis,

    I much prefer the term “Testimonials” to “Best Practices”.

    I’m a firm believer taking time to consider what worked and what didn’t work on a particular project.

    Here in Oracle many of our teams use the After Action Review methodology to make this happen in a group setting.

  3. Luis, as you can image I agree with you on the fact that “best practices” are a false concept as no practice is, or at least should be seen as, the best.

    However, I do not agree with one of the reasons you point out for that: “you can’t copy best practices from one company to another because eventually it would turn itself, over time, into what [Oscar Berg] calls a “common” practice. They would devalue themselves over time with their continued reuse to the point where at some stage they would no longer add any additional value to that business”

    For me there is no issue on something becoming a “common practice”. In fact, and echoing what Mary says in her comment, it would be great if all “best” practices could be copied and used and became common practices – that would save a huge amount of time and money. The big issue is that once a practice is taken as “best” it is very easy for people / organisations to stop questioning it, trying to find ways of improving it.

    I hope this makes sense.

  4. Hi Luis. Perhaps “Best Patterns” or “Good Patterns” is a buzzword to use…with all due reverence to those who actually know and have written something about “design” and “pattern languages”. The word pattern here denotes (to me, anyway) a higher level of thinking and gets away from the implication that practices can be copied in any meaningful way. Your blog rocks. Thanks for keep us all on our toes.

  5. Luis, thanks for sharing another great piece. In his book “Collaborartion” Morten Hansen discusses how the more help a team gets from colleagues outside the project, the worse off they are. He’s talking about a cost of collaboration; we might postulate that best practices include a tax, i.e. that the further away from you the practice was developed, the more trouble it will cause you. It’s hard to generalize, of course, but as you’ve mentioned, extracting a practice from one context and embedding it in another has its risks.

    @Frank: best practices are often just stories that are generalized, turned into recipes an eventually into crutches. The semantics of “testimonial”, i.e. event-specific, descriptive, originating from an individual perspective make it such a great term. They are to be learned from, not copied.

    Here’s a good take on rules and exceptions from “Sea Kayaker’s Deep Trouble”:
    “Obeying rules without an understanding of the reasons behind them
    “creates an approximation of competence
    “which leaves one vulnerable to the exceptions.

  6. Luis,
    I remember arguing with senior managers in1998 (when best practice propaganda was in full flight) about the limitations of assuming there was one best way for all circumstances. It is critical that we attempt to share the best of what we have learnt and even more critical to know the limitations of these learnings when using them as a foundation elsewhere (in anothercontext). I agree with you that rigid best practices reapplied without adaptation is dangerous. However there aresome enviornents where this can be betterthan total flexibility. Companies in the known/simple Cynefin domain may benefit from keeping processes consistent whereas those in the comex space definately need aflexible approach, but may benefit from what has been successful in similar contexts before. The danger comes (in complex), when people think that THE answer is known. This is where best practices do most damage and why thy areso fiercely rejected. Often fueled the way they are sold in without thought of the new situation.
    Perhaps best practice might be discover what has been done and leverage it for future emergent opportunities.
    Tweeting as Metaphorage

  7. Well said.

    Humans are really only good for exception handling – when things go wrong, you need a human to analyze, to decide, and to find the ways around the problems. That’s what humans do best. It bugs the hell out of me when people get obsessed with optimizing and automating processes as though their workforce is an army of fleshy labor work-bots.

    Hopefully we can cast off the shackles of a hundred years of applying manufacturing management policy to knowledge work. It’s different. We know it’s different!

  8. Pingback: IntranetLounge
  9. Agree completely on all points, but I take a much more direct approach to get to the same “destination” in that I suggest that at BEST, a best practice has limited strategic value to begin with. A best practice is at best, “new to you knowledge.” New to you knowledge isn’t what feeds innovation which needs as its spark, new knowledge. All boats float in a storm, and all boats rise on the tide. So indeed a best practice does become as stated above a “common practice” and that doesn’t in the end really get you much (and that’s assuming that you can even apply the best practice).

    This, I believe, is exactly why KM is ill suited for common methodologies and such and why efforts by one organization to simply copy what another does are doomed to fail.

  10. Being highly tuned to metrics that discern improvement from the contrary is the only way to survive in a dynamic world. Baldrige and Deming had this figured out years ago. Make sure each morph is measured and in the right direction.

    Jim Hodson
    CEO Countdown To Buy

  11. In general I agree with your post, Luis. However, might I suggest a “meta” best practice — include feedback loops in your KM and social computing processes and systems wherever possible. Otherwise, how will anyone know when the current best practice has outlived its usefulness?

  12. I the health and development sphere I have also heard “evidence-based” practices and “promising practices”. Obviously, one comes before the other. “lessons learned” is a good terminology when you want to be context specific because the lesson comes AFTER an experience that is always relative to an individual. In closing, from now on I’ll average good and best and stick to “great”.

  13. Luis, just re-reading Anthony DiBella (Chap 9 – Learning Practices – Assessment and Action for Organizational Improvement) where he prefers “Best Principles” to “Best Pracs”. The challenge in trasnferring a “best prac” from one org to another is that often the context is not considered. The need to understand, reflect and re-invent (probe/sense/ re-innovate maybe) before responding.

Leave a Reply

Your email address will not be published. Required fields are marked *