Speed vs. Satisfaction
"Has anyone seen evidence that satisfaction rates go up when the
average speed of answer improves? We're trying to make a business
case for a few more agents on the phone, but it's surprisingly hard
to find concrete data on the potential ROI."
—Cindy from Cincinnati
Speed to answer can be tied directly to time to resolve. The
faster a call is answered by a skilled support representative the
quicker the resolve time will be—it will require fewer
interactions on the issue. We operated a direct to support rep
model using an ACD/IVR in a very complex enterprise software
company. We consistently answered the phone in less than 1 minute
90% of the time. We found our Resolve times come down and our
overall customer satisfaction and loyalty go up.
While you may need to start with a couple of more people, you will
find the lowered resolve time will improve your efficiency. Use
the freed time to do more training and watch the resolve time drop
Technical Support Management
I have some mathematical evidence from two years of ticket data for
Lucent/Alcatel-Lucent. Statisticians from our Bell Labs organization
studied the data for us. We found the number one most important
variable stored in our ticketing system that relates to score on a
transaction survey is resolve duration. The relationship is linear
with satisfaction decreasing the longer the ticket takes to solve.
Lastly, the study showed that the linear relationship holds true
regardless of the severity of the issue or regardless if the issue
is a product defect or just a need for help.
I can show a direct correlation with email speed of response to
I don’t have experience with phone average speed of answer, but
can follow our daily satisfaction rating as far as how quickly we
respond to customers via email. We call it “initial response time”.
We are primarily an e-service center.
Feel free to contact me with questions or to discuss.
—Linda von der Heide
Director, Customer Care
There seems to be two worthy measurements about speed of service
vs customer satisfaction.
1. Cleverly worded survey questions will provide valuable
information. "Please rate your satisfaction (scale – low to
high of your choosing) with how quickly you are connected to our
Reps when you call for assistance."
For this qualitative survey question, the actual number (time to
respond) is far less important than the customer's satisfaction.
You don't need to overshoot the mark, once you find out where the
2. Measure abandon by WHEN the hang-ups occur, not just the number
of abandons as a percent of total calls. For example, if the
hang-ups spike at 5 minutes, then staff to answer in something less
than 5 minutes. The problem with abandons is its very difficult to
tell if you have many different callers hanging up or is it mostly
the same callers calling back and then hanging up again. But the
spike point will tell you their tolerance level.
Depending how you create your survey, you might want to use the GAP
method. You ask the customer how IMPORTANT speed of answer is to
them (say it is on a 1 to 10 scale) and then you ask how SATISFIED
they are with speed of answer (on the same scale). The critical
measurement is the GAP or difference between those two numbers,
IMP – SAT=GAP. Generally IMPORTANCE is higher. For example if
the GAP is 2.0 or less, you are probably doing OK, When the GAP
gets larger to 2.0, you have an issue to correct. If the GAP is 3.0
or more, you may be losing the customer on this aspect of service;
serious repair is necessary. Ideally you want a GAP of about 1.
Less than that and you are wasting money on a non-issue.
Rick Kilton provides business process consulting, organizational
development, survey development, and professional skills training
to technical support, help desk and field service operations.
[If you have any other advice on this question, please send an
email to membership director Jane Farber at firstname.lastname@example.org,
and we'll post your feedback.]