Make Measurements
Measure, collect data
Measure results. Start talking about what is accomplished and at what cost.
Measure costs and results over the most suitable time period, but tend to the
shortest time period that yields meaningful results. With a rapid iteration of
information and related feedback it is possible to optimize performance more
rapidly and get the best possible results at least cost.
Even the best of implementation managers will do better when there is a good
system of measurement. Measurement should happen every day or at
whatever time interval makes the most practical sense. Usually the shorter the
time interval the better. Measurement should be done in ways that are easy,
practical and reliable.
Measurement should result in information about costs, what was done and the
results arising from the activities. In a mature measurement environment it
should be possible to compare costs and results in a coherent way over
activities in various different places and at different times.
Making a lot of measurements and getting a lot of information is interesting
but not valuable. The purpose of management information is to use it to get
the best possible results.
Measure ... make it easy
The best information is often the easiest information ... but it does require
discipline. The following information should be very easy to obtain in most
practical situations:
➢ What was done today ... how much was spent today.
➢ What is the cumulative accomplishment for the month ... what is the cumulative cost for the month.
➢ What is the cumulative accomplishment for the year ... what is the cumulative cost for the year.
Distributed analysis
Some value can be immediately derived from local information, before it is
every communicated and consolidated at another level. For example:
➢ How do the measures of today compare with prior results ... this month compared to last month ... this month this year with the same month last year ... this year with prior years ... etc., etc.
Distributed data analysis makes it possible for local decisions to be made
quickly prior to analysis work being done in some remote place. This has been
described as the democratization of organization, with data and performance
and decisions all being fully integrated and accessible at the local level. This is,
of course, consistent with the community focus described later.
Organize the data
There are many works for information to be organized. The organization is to
some extent science and to some extent art. The goal is for the organization
to be easy to do and for it to be powerful so that analysis is easy.
The basic rule is to analyze and code the data once well, and then use the data
multiple times after that ... this is much more powerful than hoping that
random analysis will provide useful results. That is not to say that searching
for the unseen connections is not useful, but only that it organized data are
best used for management purposes.
Communicate and consolidate the data
The locally collected data should be communicated to a system that facilitates
consolidation of data and its comparison, for example:
➢ How the measures in one place compare with similar measures for people in other places doing the same sort of work. What can we learn so that we can do things better?
With data all sorts of things are possible.
Result of measurement
The result of measurement should be for the immediate area operators to
know much more about cost and results, and be in a better position to
improve their work. The culture of doing the most for the least should start
to be the norm rather than the exception.
The broader dialog about the operations can start to be a dialog about costs
and benefits, about trends, about behavior of cost and benefit under different
circumstances.
|