Time Observations - 10 Common Mistakes

Stopwatch picUnderstanding the current reality within the context of time and space is extremely critical. The time observation form is a powerful tool to facilitate direct observation. The form is instrumental in the identification and understanding of waste elimination and variation reduction opportunities.  It's a staple of kaizen and feedstock for standard work combination sheets and process capacity tables.

If the time observation form is so important, then everyone knows how and when to use it, right? WRONG. Unfortunately, there are a bunch of common mistakes that practitioners regularly make. In no particular order, here's an incomplete list of time observation mistakes:

  1. One form for multiple operators. It doesn't get much more confusing than this. The individual operator's work sequence and work content can't be discerned.
  2. Component tasks not broken down into the smallest observable elements. Summary tasks like, "assemble part" or "room patient" does not give the observer usable insight.
  3. Incorrect or missing cumulative times. The lap button is on the stopwatch for a reason. And don't pretend that you're accurate enough to use decimal points.
  4. Insufficient number of cycles observed. Unless we're talking about multi-hour cycles, the observer(s) should observe and document as many as 7 to 10 cycles. How else can you identify variation and understand most repeatable times?
  5. "Interviewing" operators during the observation. Not a good way to conduct accurate, real-life observations...unless their work normally includes responding to interview questions.
  6. Improperly determining component task times. No averages and throw out abnormal values (but try to understand them and the reasons for them). Make their sum equal to the lowest cycle time observed. Above all, use common sense.
  7. Not communicating the what, how and why to the operators and other stakeholders BEFORE the observations are conducted. This is respect for the worker and helps ensure that the observed cycles are reflective of reality (no rushing by someone out to impress the observer, no slow down to taint the observations, etc.).
  8. Not following the operator. If they leave the immediate area, go and follow them. How else will you directly observe?
  9. Not using the "Points Observed" column. This is the place on the form where you can record the reasons for abnormal times, variation and capture improvement ideas. These are pearls.
  10. Not completing the form header. Without this information, later on it may prove difficult to determine who made the observations, what process was being observed and when the observations were made.

So, do you have any additions or corrections to this list of common mistakes?

Related post: The Truth Will Set You Free!

There are 10 Comments

Panu Kinnari's picture

Requirements are ofcourse different for different kind of time study. And these points are valid for normal time study. For observation time study for example you need to look at component tasks and decide the accuracy based on what you are trying to accomplish. And in observation time study decreasing the accuracy gives you ability to study several people at the same time thus giving better view at big picture.

In continuous time study, if done manually, shortest component task observable is decided by how fast you are able to write it down. Special equipment can be used to improve accuracy some what.

But I do agree with your points and they are mostly applicable to all three forms of time study.

markrhamel's picture

Hi Panu,

Thanks for the comment! You have made some good points. If the time observation is strictly for something like a value stream analysis (mapping), then typically the team is just looking for the total cycle time and is not worried about the component tasks. That level of detail is reserved for process kaizen. If multiple people need to be observed, we should then have multiple observers and/or we should employ the use of video. The video can then be reviewed repeatedly (offline) thus facilitating the necessary detail and accuracy of the time observation.

Best regards,

Dale Savage's picture

Thank you for covering this subject. One of my greatest pet peeves is when people rush through time studies and then think they understand a process enough to suggest improvements. When this is done by management, they often try to justify what they think should be happening based on what their plan was on paper instead of what is actually happening. This ends up holding associates to a higher expectation than what is actually achievable which leads to distrust of and disgust with management. Often, those doing the time studies fail to capture the non-value added portions of a process and therefore they fail to recognize and eliminate much of the waste.

When I facilitate an improvement event, I require participants to spend a whole day in observation and time studies so they can get a clearer picture of the actual current condition, based on facts. Only then do we have the right to suggest changes that will truly deal with eliminating waste. A ten minute time study cannot capture everything that must be done at an operation - time studies cannot be rushed. When everyone realizes that, then there will be less anxiety in the associates when they see someone walk up to their process with a stop watch.

markrhamel's picture

Hi Dale,

Thanks for your comment! I am amazed that when executives participate in a kaizen event and the necessary rigorous direct observation of the current reality, they often state that they never had any idea that the process had so much waste. Well, the waste had been there all along. The "surprise" occurs when leadership's idea of how things are meets objective reality.

Best regards,

Ron Pereira's picture

I am personally a big fan of using a video camera since not only can we capture times easily (i.e. hit pause, write down the time, or rewind if needed) we can also identify wasted motion, transportation, etc.

Evan Durant's picture

To #2 I would add not fully understanding that component tasks before trying to do the time observations. You need to watch a few cycles first, completely understand the tasks, and then start the stopwatch.

markrhamel's picture

Hi Evan,

Right on! I instruct people to first put their hands in their pockets and just observe the process, gain an understanding of the component tasks and their sequence and THEN start using the time observation form. Of course, if the cycles are very long (an hour or more), then this approach may not be pragmatic. In this case, I suggest documenting right away.


Panu Kinnari's picture

When I do time studies it is usually for whole shift. Only rarely shorter period of time.

What I have noticed is that proper preparation really pays of. Just observing few cycles isn't enough, I recommend doing short test study before starting for real. Especially when doing observation time studies where deciding observed component tasks in advance is really important.

Tim McMahon's picture

I have found it is easier if you can do the time observations with 2 people. One to watch content and use stop watch. The other to record the information. Two sets of eyes are better than one.

Great advice Mark.

markrhamel's picture

Hi Tim,

Absolutely! The two person team approach is tried and true. Thanks for the comment.

Best regards,