For instance, if a person’s main job is invoicing, random sampling can work because each time you check in, you get a snapshot of what the person is doing. Once you have all the data, you can figure out which task or component the person spends the most time on by looking at how often and how long it showed up in each random sample. The benefit with this method is you can observe more than one person in a set time period by rotating through each person. If you’re using this method on yourself, set up an alarm to go off at certain intervals so you can record what you’re doing for a set amount of time.
Timing a large task isn’t helpful without looking at the components of each task. If you analyze the components, you can look for inefficiencies. You won’t stop the person; you’re just breaking down the task for recording purposes. For instance, if the task is checking the mail, the components would include walking to the mail area, finding the mail, taking it back to the desk, opening envelopes, reading the mail, and discarding or dealing with each letter. It can help to have a group for observation. That way, you can have one person to work the stopwatch, 1 person to record times, and 1 person to make notes. You can also use this approach on yourself. In that case, you’ll be writing down each component as you do it.
Test the task over a given period. For example, you may want to test it over a week or even a month.
For instance, if the task is checking email and one of the components is reading email, make a record of the time it takes to read each email in the numbered boxes next to the component.
The key is finding the right level of detail. You don’t want to be overly detailed, as timing how long it takes to hit a single button isn’t useful. However, you also don’t want to be too broad, as that won’t give you enough data to work on efficiency. Say you’re checking email. You might break it down into logging on to the computer and into your email, deleting spam mail without opening, reading emails, composing replies, and organizing email into folders.
For more accuracy, take data over multiple days.
For instance, if you have reading emails as the component, the times might be 65 seconds, 210 seconds, 240 seconds, 39 seconds, and 354 seconds. Add the numbers together: 65 + 210 + 240 + 39 + 354 to equal 908. Divide by the number of times. In this case, that’s 5, so divide 908 by 5 to get 181. 6 seconds on average per email.
If you read email constantly throughout the day, you’re constantly pulling your train of thought from whatever else you’re doing. Often, it’s better to do all of a task at once, such as only reading emails in the morning, at midday, and right before you leave work.
For instance, if you must file paperwork daily in another room away from your office, consider saving it to do all at once. If you are constantly going back and forth, that takes away time you could be spending on other tasks.
