|
|
 |
|
|
Response Rates
|
|
As people get inundated with more and more surveys, they are
becoming more discerning in deciding which surveys to complete and which to
ignore. Surveys get better response rates if you:
- Have a close
relationship with the people that you are surveying.
- Guarantee the
confidentiality of responses.
- Use compelling survey
announcement and invitation messages.
- Present the survey in
the respondent's native language.
- Give your respondents
feedback on what you learned and the changes you are making in response to the
findings.
- Provide incentives for
completing the survey (e.g., entry in a sweepstakes).
Employee surveys tend
to get the highest response rates. Often, it is possible to get
response rates in the range of 75% - 85% completion for employee surveys.
This is because your closest relationships tend to be with your own employees.
For workgroup and manager feedback surveys, such as 180o and 360o
degree feedback surveys, response rates will be high only if members of
workgroups or direct reports are given ongoing feedback on the response rate as
the survey progresses.
First Customers have close relationships with
their suppliers and the closer the relationship is, the better the survey
response rate will tend to be.
Customers that invest large amounts of money on products and ongoing services
(as in the IT industry) will typically have response rates in the 20% - 30%
range. Smaller consumer surveys (e.g., retail customers) will tend to have
lower response rates. The actual rate obtained will depend in part on the
method of recruiting participants. E-mail survey invitations for
example, will fare better than surveys for which the link to the survey is
printed on a customer's receipt or appears as a pop-up on your website. If
you have large numbers of retail customers (or website visitors), then the low
response rates for these types of approaches can be countered by the sheer
number of potential invitees.
However, these response rates can be maximized by
including an incentive, such as a sweepstakes entry, and the large number of
issues that most companies close can result in robust data sets even when
response rates are not high.
In a study of more than 6,000 survey invitees, for example, we compared four
conditions: (1) no incentive; (2) an unspecified donation to charity
for each survey completed; (3) a $1 donation to charity for each survey
completed; (3) a sweepstakes for a high tech prize (this particular study used a
flash memory key, which was popular at the time); and, (4) a sweepstakes for a
Mont Blanc pen. The winner? Mont Blanc pens increased response rates
by 1/3. We replicated this finding in a second study and now use Mont
Blanc pens as one of our more common incentives. Even if people already
have one, they make nice gifts! Other research groups have conducted
similar studies with comparable findings. Paying each respondent
(generally now $5 each or a gift certificate) also works well, but this is more
expensive.
Generally, if there is a close relationship with the customers and the groups
are convenient for them, then a personal invitation from Account Managers or
other suitable staff in your organization will prompt participation. If
this is not possible, then you may need to pay participants. For those in
high demand groups (e.g., CIO's, physicians) the rates required can be
high.
When the nature of the research does not require that
you have group discussions or need to make audio-visual presentations, then
these can be a very effective alternative for qualitative research with high
demand groups.
|
|
|
