Operations managers in many companies will tell you that if a person makes it to a certain number of days, their probability of leaving the organization becomes much lower.  Put another way, if the probability of someone leaving the organization is X out of 100 within 90 days, it shrinks to X-Y of 100 for the next 90 days (or some other time period).  This coincides with some companies’ probationary periods and when some benefits begin to kick in.

This article describes one company’s plan to get its new hires to the 90-day line.  The crux of it (and other examples provided) was, “Now that there’s high turnover, we really need to invest in our new employees because this is costing us real money!”  Just as interestingly, a person interviewed who quit a job basically looked at the first couple of months as one would while dating (“I’ll give this a shot, but if it’s not working out I’ll go find someone else”).  This is an excellent example of worker’s flexing their autonomy.

So, do you have a turnover sweet spot?  If so, how do you find it (besides urban legend, of course)?

  • You should have the data in your HR software. Don’t get hung up on whether the turnover was voluntary or not—most people jump before they get pushed, and vice versa.
  • Use the right data analysis tool. The most accurate way to measure where (or whether) turnover flattens out is using a hazard analysis, not counting things in a spreadsheet.  I won’t bore most of you with the details of why this is so, but write to me if you want them.
  • You should not be surprised that there are differences in turnover rates between people in different job titles. When doing this for a client, it was apparent that turnover for engineers was a very different animal than turnover for managers.  This will particularly be the case if your benefits program is different for salaried and hourly employees.

Seeing if there’s a turnover drop off can then inform for how long you need to keep new on-boarding programs in place.  In my experience, turnover goes down significantly at the point where an automatic raise kicks in.  However, if you are going the non-raise route, to know what should go into the programs, includes getting candid feedback from new hires and being open to their suggestions.  There may also be a bit of a Hawthorne effect—perhaps you do not need to change anything, but new hires will appreciate the additional communication and that will make it more likely that they stay.

Measuring impact is a critical part to implementing these programs.  Fortunately, the hazard analysis is very adept at doing this on an ongoing basis.

It’s unlikely that there is a “magic” day when your turnover goes down significantly.  But, people won’t stay for two years if they don’t stay for two months.  Setting this specific time period where you put in extra effort to give people the chance to be part of the team is useful for getting others to focus on turnover issues.  It is just more effective to set this time period based on data rather than myth.