Montag, 14. März 2011

Learning from failures: Activity-based KM KPIs do not work

While everywhere I am reading that we learn best and most from failures, not so many admit that they have failed. So I have decided to publish a series of my failures and the learnings that I have made.
In this post I reflect on why we have worked so long with activity-based KPIs, and why they still do not work for Knowledge Management.
Looking generically at the change management landscape from the measurement perspective, on can draw a picture below.

Looking at it from the result, a change is triggered to achieve an effect or impact. In order to do so, change applies to the employee, the structures and processes.
If we zoom in and apply to Knowledge Management, we see, most often the measurements relate to actions & activities. Do x numbers of these activities, accomplish y numbers of those actions. E.g. create 4 knowledge objects per head per year; perform 2 webinars, 1 Best Practices, do 2 re-uses ...
In Ericsson we have extensive experience with activity-based KPIs, and there were good reasons why they have been implemented, and there are more good reasons, why they fail.

Here are some of the expressed - and also unexpressed - reasons for implementation:
Easy to define
As KPIs are applied to huge organizations, and often monetary incentives are connected to them, and they need to be approved by management (non-KM experts): the expectation on the KPI is that it is simple, easy to understand (decision makers), communicate, understand (employees). Actions and activities can often beeasily defined digitally (done, not done) and thus quantified and counted, and they are very tangible.
Easy to measure
This goes hand-in-hand with the definition: The best-defined KPI is worth nothing, if it cannot be measured. But there is another twist in the argument: Knowledge Management is tightly connected to the Knowledge Management IT landscape, even often misunderstood to be the same. So if the IT department has been the driving force in the implementation (of what? KM or KM IT tools?), then it is very manifest that the success of KM is defined as the acceptance and the usage of the respective tools. And activity-based KPIs materialize often in tool actions, which can automatically, that is most easily measured.
Acceleration of KM
The idea to rely on diffusion for the implementation (better uptake) of Knowledge Management behaviour must alienate management. Not few of them have read their Keynes: "In the long run we are all dead!" The activity-based KPI steers the employees towards the behaviour and/or tool. Those who have reflected on the fragile balance between force and empowerment, realize that the activity-based KPI shall only used in the early/low maturity stage as accelerator.

And yet Activity-based KPIs lead to failure
From the generic change management landscape it is very obvious that result-oriented change management (change management with effect / impact) must be anchored in the behaviour and culture of the employees.
This will lead to the adequate actions / activities. Changing the culture by whipping employees into actions does not work. It does not work. The activity-based KPI distracts from the essential work on behaviour and culture. Working on the behaviour and culture takes much longer, costs more energy (as you face and meet the resistance of the sub-conscious) and is much harder to put into measurement (define and measure). So management is fooled by appeasing false impression of impressive achievements towards and usually over-achievement of activity-based KPIs. The numbers are great (KM is implemented, tick in the box - although we know that it is never finalized) and the management support and focus goes somewhere else - the poor Knolwedge Manager lose management support when it is most critical. Because now the activity-based KPI is not only not doing the trick, but exhibits a negative impact: As easily as they are defined and measured as easily they are short-circuited (we call it the famous December hockey-stick, close to the end of the measurement period, usually December, employees formally fulfill the KPI by a knee-jerk action, and receive the incentive connected). This actually is counteracting the behavioural change, makes it even much harder. At the same time this (mis-)behaviour is polluting the database with poor quality contributions and thus corrupting the usage of the tool (why should I use it, there is only crap in it). In this situation your efforts to work on the culture will not find any acceptance. It is scorched earth.
Losing the key contributors due to activity-based KPI is a devastating experience.



  1. I dont think the problem lies with activity-based KPIs - it lies with using ONLY activity-based KPIs (and potentially using the wrong ones as well, if you are measuring "creation of objects").

    I would suggest that we need outcome KPIs to measure whether KM is adding value, and activity based KPIs to see who is active in adding that value. Activity provides the link to value.

  2. @Nick: "Outcome" KPIs measuring value add
    What you call "Outcome" KPIs are in the picture KPIs measuring "effect&impact", and I fully agree, that is ultimately what justifies KM. NO outcome, no sense in KM. And then to do a "root cause analysis" in the positive sense to identify what / who has created the value (in order to market and multiplicate the good use) sounds very good to me.
    And if this is what you mean with measuring activity, fine.
    Totally agree, however this is not what I have observed as activity-based KPIs (because they lack - by my definition to cause and effect relation to value creation).
    Again we have the same opinion, calling them the "wrong ones" - sadly I often encounter them in action: large organizations, automised, activity-based, useless, spoiling.



Hinweis: Nur ein Mitglied dieses Blogs kann Kommentare posten.