Causal Relationship: Hold Times and Abandon Rates
I had an inquiry this week on hold times, asking what a good target would be. According to our benchmark data, the average hold time is under 2 minutes, which definitely shows our B2B technology support slant–I assume the consumer numbers would be much higher. Or maybe I’m still on edge from my last marathon hold session–over an hour–for our local energy monopoly. My usual guidance is that anything under 3 minutes is acceptable to customers, and when hold times stretch beyond the 3 minute mark, callers start to bail out and hang up. (And I don’t care how many times you play the “your call is important to us” message. Clearly it isn’t or you would have more staff.)
The percentage of callers who aren’t willing to wait on hold any longer and hang up is called the abandon rate. Logically, the longer the hold time, the higher the abandon rate. To see if this logic played out with data, I divided the survey responses for hold time into three categories: Pace Setters, those with the lowest hold times, Average Performers, those with median hold times, and Low Performers, those with the highest hold times. When you average the abandon rates for each group, you can clearly see the impact of longer hold times:
The Pace Setters, with an average hold time of just over 30 seconds, have the lowest average abandon rate (3.3%). As hold times increase, so do abandon rates, with the Low Performers (average hold time 4.7 minutes) jumping to a 7% abandon rate.
It is important to understand which metrics have a causal relationship, i.e., impacting one metric automatically impacts the other. This is helpful when you are trying to move a specific metric so you know what the influencers are. Another example of a causal relationship with support metrics is First Contact Resolution: the higher the FCR, the lower average resolution time is; and usually, the higher the percent of issues resolved at Level 1 is as well. Also, members usually see impacts to CSAT when critical service metrics (FCR, resolution time) improve.
I have an example in my book, Lessons Unlearned, about a US airline who wanted to increase productivity (calls per shift) and lower hold times by cutting call length. Obviously, if you spend less time on each call (or email or chat), you can handle more interactions per shift, which reduces hold times and abandon rates. Sounds like a win:win:win, right? Wrong. In this case, by putting an arbitrary time limit on inbound reservation calls, agents were cutting off calls before reservations were completed, or before customers could be routed to hotel and rental car partners. As a result, revenue dropped dramatically, and marketing and sales had to override the call center manager who thought limiting talk time was the answer to everything.
Don’t look at metrics in a vacuum. They all inter-relate, and you must carefully think through any plans to shift one metric to understand what the trickle down impact will be across the organization.
Thanks for reading!