PERFORMANCE IMPROVEMENT, PROBLEM-SOLVING AND CRITICAL THINKING

Performance improvement is one of the many names given over to the process of figuring out that something needs to work better.  There are a number of official-sounding methods for doing this---I personally like LEAN.  Its why I bought a Toyota (the 50 MPG didn't hurt, either).  But any method is fine as long as it is used correctly.  Systems should be efficient and effective--processes should not include steps that are unnecessary or nonproductive.  Who really wants to waste time doing a bad job?
Which brings me to the cause of the title.  Every day in this area, people study the data and make suggestions that don't relate to it or don't study the data and make suggestions that don't relate to the issues at hand.  My point is---the first step is seeing the problem.  Looking at what is causing the problem.  Understanding how the steps of the process work together and what is going on that is making them not work together has to be the first goal.  It doesn't matter if you read about a wonderful way to move the mail if your problem has to do with getting the right piece into the box before it is mailed..   The people that love to fix problems don't want to do the first step.  They have a fix in mind when they begin, or think something they read in a trade magazine or heard at a management meeting should be tried.  So it is tried, with no baseline data and no analysis of what that information meant.
Now we have a change, but not necessarily an improvement.  People are fired for resisting the change.  People are moved around to find a way to create the change.  AND if everyone is very lucky, the new process actually related to the original process issue and cause what was subjectively seen as an improvement.
Serendipity!
I love chance, randomness, creativity, but for some areas of work, a more scientific method (pun intended) is needed.  In science you start with an hypothesis, but we aren't doing scientific research, we are using a more rapid, goal-driven method, so we start with the problem.  And the first step is identifying the problem.  The exact problem.  Not the vague problem.  "we waste a lot of time and energy on getting information " is vague.  Ask the right questions.  Get the right information.  See if the industry has any benchmarks that give you something for comparison.  It is fine to improve against yourself, but not until you have defined your processes in a way that allows comparison.  While some of the world is subjective--everything can be quantified.  DO THAT.  If your business invests a lot of time and energy (translation--money) into performance improvement, and the end result is everyone thinks it was probably a good investment and the customers and employees think they can see a difference, will you do that again?  Will you make sure the changes stay?  Or will it be back to business as usual as soon as the improvement project  is finished.
Define everything.
Do a library search: to find benchmarks, what other companies have successfully done about the same or similar problems, read about identifying problems, see if there is stuff out there that is evidence-based.
Identify the problem:  there may be several that need to be separated and fixed individually, there may be several that are part of one process or one system.  And sometimes a problem is a person problem, but it will not be an "all people " problem.  If your research is pointing toward "all the people" in purchasing, you may need to talk to all those people, see what they say the problem is, it is probably something built into that department--part of their process which is frequently the result of policies and the people doing the work not communicating.
Administrating is not about sitting up in the clouds dictating to the workers who are so far away the two groups don't recognize each other except with large font badges.  If you are making policy, find out what is being done from the people doing it.  A policy is not a writing assignment,  it is supposed to represent the organizational process.  If the two are not related, you-the administrator- are not effective and your process can not be made more efficient if it has nothing to do with what is really going on.
Be OK with finding out that your favorite system/process/person may need to be improved.  Nothing stays the same.  Anyone that works with technology knows that.  The game is always changing.  Nothing can be too special for improvement.  We all love our babies, but not every baby is going to last forever.  My least favorite conversation is with the person that helped to create the old computer system.  The bragging never ends, every new change is compared to the greatness of that original system she made---and that every person using it has complained about daily for 15 years.  It was her pride and joy, but it was never a very useful system.
So, for this part of the job, be the detective, be the scientist, be the illustrator, be whatever is needed to identify the problem.  The original method used to identify the problem should create a data baseline. Use the same types of data collection to document the improvement.  And use numbers--they are really not that scary.