Measures
Definitions of measures within the area of Knowledge Management
- A measure is something that is used in relation to a rule or a principle. Here we assume that a measure is
the smallest building block of organisational software, i.e. the formalised perspective of an organisation.
- When we discuss measures we always refer to measures as being used in the formal management processes of large
organisations.
- The most simple example of sociometrics can be found in weekly magazins. Measure yourself concerning your ability
to X
- The world of metrics is marketed as being objective and neutral but anyone who knows the low degre of reliability
also knows that the type of measures you want to use is a specification of the politics in the organisation
- However, we need to measure because if we do not we are victims of complete subjectivity instead of partial
and standardised subjectivity
- The measure of performance can be seen as an agreement of how an organisation want to standardise its evaluation
of results of work
- What you cannot measure you cannot control (Curt Nicolain, Director of SAF)
Background
Measures and machines in relation to trust
Antithesis of measures is personal trust in human beings motivation
The antithesis to using mesaures is trust, communication and closeness
Intuitive agreements, personal relationships
All such appreciated human features only works if:
- There is an engaged an caring management
- There is a low degree of change in the organisation
- The organisation is not distributed over large areas or divided into virtual parts
This points at measures as being a shell of protection for large abstract structures
Relating formalisms to non formalisms
There must be some kind of chain of command in an organisation. There must be some structure that guides the
employees in their work. Within any department of an organisation there is either authority by person or by rules
Rules versus Humans
|
Rules |
Humans |
Creativity |
Spontanuous creativity of employees |
Spontanuous creativity of management |
Flexibility |
Flexibility by adjusting weight on rules |
Flexibility be reorganising fixed routines |
Delegation |
Delegation is automated |
Delegation is personal |
Forms |
Software of tomorrow |
Today is based on tradition |
Change |
Change is meta-rules |
Change is enforced |
|
|
|
|
|
|
Quality of rules
Rules should be as simple and as generic as possible
The better the rules:
the less need for control
the less execution of power from management
Reorganisations
- The reasons why the CEO must be allowed to reorganise
- otherwise she does not get in control
- otherwise old hidden powerstructures that prevents change
- What is bad with reorganisations
- It takes time from important productivity
- The resource consuming conflicts may take away motivation
- How to avoid non-productive continuous reorganisations
- Select the flexible organisation once and for all. Then reorganisations willnot be needed
- Base the organisation on field theory instead of authority.
- Motivation instead of force
- Market instead of plans
- Introduce formal rules. Football would not be entertaining for the audience if there were no rules
- Make sur that the internal competition of the organisation is controlled and well structured. Everyone is competing
with everyone according to the rules
Is it possible to monitor competition within an organisation
Any type of game can be controlled if this believed to be productive
- The best monitoring of a game is never to change the rules but only to slightly modify the way the rules are
applied. For instance by modifying the reward for desired behaviour. Like Alan Greenspan monitors economy.
What constitutes good rules and measures of their fulfillment
They are:
- Global
- Simple
- consistent
- As few as possible
- Flexible
- Synchronized
Three levels of synchronizing organisational activity
- 3 Reward structures:
- People can do what they want but are rewarded if the results are beneficial for the organisation
- Abstract formalisation and centralisation of principles enables impersonal control
- The power of information technology enables virtual control
- 2 Management By Objectives
- Telling people which results are expected
- 1 Monitoring
- Controlling that people do what they are told to do
Culture of agreements and trust
- Intimacy is created in a community when there are many shared agreements and basic trust about people behaving
as is expected
- Trust works when there is a physical presence in the same building or any kind of person to person relationships
- When there are no relationships and no continuous commuication there is no trust
- In virtual organisations the relationships are anonymous which creates entirely different roles and games
- In such an organisation there is a need to develop formal stuctures that makes people feel safe in what they
do. People need to feel that they are seen and that what they do matters. If they do not know that they are seen
sub-optimization will start.
Any deliverance of anything within the organisation could be associated with organisational learning:
- Assess the value of the delivered product/service by giving feedback concerning how needed the product/service
was
- Why is it good or necessary
- How can it be improved. If we do not relate the quality of something to any kind of genereal standard the person
who delivers soemthing will not know if it can be improved or not.
- There should be a formalisation of possible alternative ways to improve it
- Establish a measure for the product/service that shows the quality of it
Integration means that a transaction should be defined by already existing generic definitions
- Whenever something unique is defined this increases the risk of information overload
- "Specialisation may create a growth of a type of stupidity called knowledge"
- Every time any piece of information is generalised thiw makes it easier to:
- Index the information
- Reduce unnecessary information from it (compare this with inheritance in object oriented programming)
Measures is a natural part of the formalism needed in large organisations
The more a cooperation is growing the greater is the need for standard routines. Such standard routines should
be as simple as possible to implement. The organisational software for measuring performance should be so simple
that anybody could do it
An exampel of how complex measures can be utilized
Peer review
Every employee has a certain amount of points they can give out each month. Quality, motivation and expert responsability
points.
Who gets it must be secret. It is all handled by a system. Those who do not do it do not get a bonus.
Professor Robert B. Cialdini's 7 principles forhow human acts in organisation. Harward Business Review 2001
- People like those who like them
- People pay back with the same as what they got
- If you give a favor it creates a dept, but only if this is made explicit to the receiver
- People want to be like those who are like themselves. This works better than authority
- Public conversation. Make people vote on things to motivate them to be loyal
- People trust consultants/experts
- People want those things that are scarce and that they have a limited access to
- People inform only those it is worth to inform. you do nto inform somebody if it does not pay to inform this
person.
- People do not read emails that are not sent only to them
Types of measures on the management of knowledge
Knowledge Management could preferably begin with with Personal Knowledge Management PKM and then Group Knowledge
Management GKM before Organisational Knowledge Management OKM
Levels of measures
- The basic level is numerical measures
- The next is love and trust
- Then there is a distinct and lofty communication
- Nobody cares who does what since all have an abaundance of the oxygen
- Everyone supports each other
Specific advantages of using measures:
Game efficiency
The efficiency and creativity of any game is related to the cost of making spontanous "on the fly" commitments.
The entrepeneurs must be allowed to improvise.
Knowledge transfer
Knowledged transfer should be as simple as posible and coded into the culture of the organisation. For instance:
Each friday afternoon we select the one who tells the best story in our five person self-help group. We write a
number on a slip of paper.
How the measures makes the organisation more balanced
- Organisations are alive and they need to be balanced. In order to balance an organisation you need to discover
the imbalances
- Measures creates a simple standardised map on an abstract level
- If the measures are numerical they can be added into general appreciations of value. This was how money was
invented.
- Administration should be simple, organic and automated. Decision should be done on a general level. Automation
on a specific level
- The transaction costs of everyday relationships decreases if they are based on automated relationships = trust
- If the company manages to formalise the power game there is less risk that the organisations falls into the
trap of suboptimization due to hidden personal interests
How the measures increase the flexibility and efficiency
- Organisations are not efficient until they are machines, but few leaders know how to formalize the abstract
into machines (Kamprad, H&M, McKinsey, Aurthur Andersson, Accenture, Ernst and Young....)
- The large change towards the large growth will not happen until the owners understand how to formalise an organisation.
- This formalization should be so simple that it penetrates the whole organisation as clearly visible symbols
- Efficient organisations can be alive within the schell of machine like bureaucracy
The efficiency of measuring
- Evaluation is expensive
- Evluation can rarely be rational
- Evaluation should preferably be on documentation
- All documentation should be standard
- Market can be evaluator if the market is able to compare products services. In order to compare they need to
be able to measure
- Talking with friends is never enough. Only open competition makes it.
Routine reward
Any routine without an added value is the death to the organisation
Routine hierarchy
Small measures should be added to large measures by immediate boss and colleagues. By units
Development of measures in an organisation
Simplifying all methods and standards that can be automated in large organisations
All large organisations need to automate their information processing in order to get an efficient administration
It also need to coordinate all automated processes. This is where most of them fail. They lose advantages of scaling
when they have not standardised the information processing
All complex methods should be inside the heads of individuals. The larger the organisation the more simple should
the principles be for measuring the value of the performance within the organisation. Shared methods should be
easily communicated as subclasses of internationally well known methods.
Formalise everything that can be formalised if this can be done within a simple set of rules and measures
Everything should be formalised into measures of control (either individually or in organisation). Every theory
should be:
- Broken down into transaction parts for how it is used in a routine transaction
- Analysed what the two parties of the transaction is getting out of the transaction
A description of a measure should:
- Explicitly show the flow of products/services/information within the transaction
- How a rule can be related to a general rule. How is the value of a measure added on to a general measure
- The frequency or reports of the measure.
- How is it weighted in relation to other measures
- Be defined by its users. The machines should be designed by the ones who use them and modified by thsoe who
use them. The machines are here to allow us to be more spontanous. Form enhances creativity
Developing prototype measures
- Getting a working set of initial hypothetic measures
- The company creates initial measures
- Each employee agrees with personal goals with each boss
- Refinement phase
- All measures are refined
- Each employee makes frame agreements with boss
- Completion phase
- Measures are standardised globally
- Each employee writes reports
- Management governs by assigning rewarding weights on targets
The relationship between local and global measures
Even if global measures are preferable to local measures in an organisation there should be individual agreements
about which measures to go for, i.e. which measures are preferable in this part of the organisation. The boss
needs to know what you intend to do in order to be able to support your plans.
Development of measures globally in a large corporation
Once general principles and general measures are accepted on the management level, each manager can introduce
them on his/her unit.
There should be a bookkeeper of KM at each corporation. Doing the following:
- Creating standards
- For every type of job carry out interviews: When have you done a good job?
- Refining standards of what is expected of employees. Not only concerning the specific output but also concerning
the cooperative behaviour
How can the measures be created by using interviews
How do you want to report your performance in a numerical way.
- Must be general measures. Select among all accepted in the company's list of certified measures
- Must be possible to make benchmarking with all work processes that can be related to your work process
Interview steps of the knowledge champion's (facilitator) interviews for extracting the essential knowledge:
- Warm up
- what do you work with
- who is your client
- who gives input
- Directions
- Doy ou have any goals in your work
- when do you feel good
- Specific goals
- Who is your end client
- When is your customer/end client satisfied (everybody has one)
- How do you get fast results at your client
- Where do you find you clients
- Where is your market
- How can you make the product better
- How can you make the employeed to like their work (hygien)
- How can they be motivated
How to introduce measures of knowledge contribution step by step
- Quarter 1. Those who fulfill certain basic criteria get a bonus. During a conference all is publicly scrutinized
against criteria. All must be informed about this from the very start. New process templates are introduced
- Quarter 2. Some get bonus and honor. Those who have not filled in get a warning. New templates are introduced
concerning how to write plans in the calender with references to processes.
- Quarter 3. Calenders are checked and bonuses are handed out. An introduction of measures when evaluating the
weekly calender with the excel-system is introduced
- Quarter 4. Results are compared
- Quarter 1. A complete evaluation is done
Continuous refinement of measures
Measures can be manipulated
Ittner and Larcker, Harward Buisness Review
Kaplan and Norton: balanced scorecards
Accenture: Performance prisms
Skandia: Intelectual Capital Navigator
Measures of satisfied customers
The only way to measure is through benchmarking standard figures
- with other companies in the branch
- with all companies
- with yourself over time
They should be more and more general.
More and more soft
In a future society we may all asses we we are standing and nobody would want to have a higher assessment than
is realistic. the game does not work that way.
Criteria for measuring the quality of knowledge structures
The measures should be:
- Easy to:
- understand
- apply
- communicate
- summarize => numeric
- General
- Valid
- Reliable
- Pedagogic examples
Examples of criteria for measuring the quality of a thesis structur
These criteria should be used to secure that the knowledge structures meet the requirements for constructing
knowledge descriptions. Especially the requirements of the knowledge being related to generalisations of the knowledge.
All requirements must not be fulfilled but the criteria can be used to assist you in analysing the knowledge descriptions.
Domain independent headers and concepts
Read through all headers and concepts and see if there is any of them that cannot be used in any domain. If
they cannot they are not domain independent and should not be headers.
Specific examples
Does every single piece of knowledge have a specific example from a specific situation/domain or is it possible
to get a clear understanding of the knowledge without such an example.
Background
- Stand alone. Is the background sufficiently described to secure that the knowledge object can be used
"stand-alone". This means that it shall be possible to understand the knowledge content of the description
without any previous knowledge of the background of the knowledge.
- When. Is it clearly stated when the knowledge shall be used
- Who. Is it clearly stated who shall use the knowledge. Is it clearly stated which is the role of the
receiver or the target group.
Problem
Can the reader see what type of problem the knowledge is intended to solve
Purpose
What is the purpose of the knowledge. This may be redundant with the problem but can be a clarifying inversion
of the problem.
Pattern classification focus
Is each piece of knowledge in the knowledge description classified into any of the classes below:
- Logical relation. The described knowledge consists of Rules, Laws or something that can be proven as
being logically related. In natural sciences one finds logical relations that are always the same and that can
be proven in all possible circumstances.
- Correlation. The described knowledge consists of principles, guidelines, checklists, measures, criteria,
requirements, success factors, rankings or any kind of relations that are not absolutely true or false but based
on estimations of their value or usefulness. This type of correlation is found in social sciences
- Process. The described knowledge consists of changes over time. It can be a method, a system, a story
or any episodic sequence of states.
- Object. The described knowledge consists of static structures such as models, hierarchies, taxonomies
Criteria for organisational software
- It must be simple
- It should preferably be domain independent
- It should follow some global standard
- It should be based on based at report structures
- It should enable benchmarking
- The reason KM is difficult to implement is that it must be rooted in accepted tacit routines of people. You
cannot force new routines on people but you can show them ways to prove they have done something good
- The art is to create habits of formalising these tacit routines into 1) decisions 2) measures
Examples of general measures and evaluations
- Personal
- Evaluate a person for a job
- Evaluate a person for selecting the best
- Evaluate a persons knowledge
- Organisational
- Degree of centralisation
- Degree of creativity
- Degree of efficiency
- Degree of motivated employees
- Degree of business intelligence
- Value of organisation. This is the most important area
- Results from new investments
- Results from selling production units
- Deficit from investments
- Cost of maintenance
- Economy
- All possible quotas
- Results of politics. Should be a department
- Comparing governments
- Benchmarking governments
Structure and management of measures
Meritocracy
There should be levels for every person and areas of rights
For every position a person can be certified to this position by:
- Quantitative additions of merits
- An example in hte academic world is: Number of publications
- Temporary merits that are valid a short time period after an examination by a board of experts
There are many ways to divide an model of an organisation into parts
- By functions. Economy, production, marketing, etc..
- By products. Cars, trucks, boat engines, etc..
- By markets. Young children's clothes, women's clothes, men's clothes, etc..
- By geography. Sweden, Norway, Denmark part of corporation.
- By processes. Chemical industry.
Whichever way you divide it it should be from a perspective of making report structures automated
Rules should be centralized
- The higher up in an organisation the more difficult to have personal control
- All principles and rules should be centralised and all operative decisions should be decentralised
- Successful corporations are known for global rules
- All formalizations should be centralized. IT, Databases, etc...
- Abstractions can be centralized when they are impersonal
Risks and dangers when introducing formalisms
There is a risk that reward systems conserv machine bureaucrazies
The measures should be instantly created in every negotiations
Is all that should be discussed at yearly conferences
They should be related to return of investments
When are the rules counter productive
Decay begins at the most abstract level. Suboptimization is initiated when higher level control does not work.
An analogy with decay in a human is presented. Decay in:
- Abstract belief structures (moral)
- Mental disorder
- Emotional disorder and stress
- Psycho somatic illness