Employße Relations

SPECIAL REPORT

TOOLS OF
ENGAGEMENT
Employee engagement surveys
are only as good as the questions
they^re built on.
By Leigh Rivenbark

In 2001, Campbell Soup Co. faced tough times. Sales of

its iconic red-and-white food brands were slumping, inno-
vation was lacking, and employee morale was low, recalls
Chnsti Downes, Campbell's director of organization effec-
tiveness.
Just the right time, then, to ask employees "How con-
nected to the company are you feeling these days?" >
The author is afreelancewriter and editor in Vienna, Va.

'^
 Campbell, based in Camden, N.J., ate public-domain reports on their engage- cheap, it doesn't repair the equipment we
launched its first employee engagement ment research, so review their definitions need," the result could be questions ask-
survey at that low point because offi- of engagement and their focus- ing whether employees have the resources
cials needed to know what would make Benjamin Schneider warns first-timers they need to do their work or whether
workers feel more connected to the orga- not to confuse measuring engagement supervisors give them support, he says.
nization. Since then, Campbell has run with measuring job satisfaction. "Satis- How involved should department
engagement surveys annually, and today faction is largely about stuff over which heads be in designing questions? "Don't
it surveys 6,000 to 7,000 employees— the larger company has control—pay, allow the survey to be constructed by com-
about a third of its workforce—each year, benefits-—but engagement is mostly mittee," Gibbons advises. Stakeholders
Downes notes. Executives have seen the under the control of the local supervisor, may try to get pet projects onto the survey,
ratio of engaged to disengaged employees through job assignments, trust and so and keeping their questions out may prove
go from 2-to-l in 2003 to 23-to-l today. on," says Schneider, senior research fel- politically difficult.
Downes and other HR leaders who low with Chicago-based HR consulting However, some companies have had
measure employee engagement say com- firm Valtera and a professor emeritus at success in involving many stakeholders in
panies shouldn't wait until times are good the University of Maryland. survey creation. At Seattle-based outdoor
to siirvey. Consider surveying even when Engagement surveys also aren't opin- gear cooperative REI, Michelle Clements,
the economy is down, determine exactly ion surveys. Opinion surveys ask how the senior vice president of human resources,
what matters to the organization, and organization is doing on broad initiatives and HR planning and operations man-
write questions that will provide answers leaders identify as important, such as ager Chris Gardner invited ID leaders
leaders can turn into actions. diversity, hut engagement surveys focus from areas including retail, merchandis-
on "the extent you are personally con- ing and information technology to review
Define Engagement nected to the organization ... your daily questions. The HR team members made
Business leaders sometimes plunge into work, your supervisor, the probability you clear that the leaders' role was to provide
surveys without first defining engage- want to stay with the organization" and suggestions, not to make final choices,
ment. Consultants, vendors selling survey how all ofthat affects productivity, says she says. The leaders became advocates
tools and other employers all have defmi- Laura Lea Clinton, GPHR, director of for the survey because they saw firsthand
tions that may or may not apply to your HR business partnerships at CARE, an how it worked, Gardner says.
organization, notes John Gibbons, direc- international humanitarian organization. Employers emphasize that before a
tor of employee engagement research and CARE, with 700 U.S. employees and survey takes place, the sponsors should
survey services at The Conference Board, operations in 45 countries, will roll out its let executives, line managers and employ-
a membership and research organization first engagement survey this year. ees know that they will use the survey
based in New York. results to make changes employees can
HR professionals must decide for them- Who Should Help? see, "What is the intent of what you're
selves what engagement means. Gibbons Before wnting questions, HR profession- going to do with the results?" says Don
says. "Understand why you're doing this, als and their vendors should decide who Lowman, co-author of Closing the Engage-
To find out if people love us? To evaluate gets to contribute. Schneider recommends ment Gap (Portfolio, 2008) and a member
the management team? To predict discre- involving workers upfront. Employee of the board of directors of global consult-
tionary effort? To lower turnover rates?" focus groups can alert employers to issues ing firm Towers Watson & Co. "Have a
For example, a retailer might not care that should be covered in survey ques- commitment to act."
to know reasons for turnover but might tions.
want to measure whether employees are For instance, if members of a focus States of Mind
engaged in ways that improve customer group say, "This organization is really Structuring questions carefully yields use-
service. Where research and development ful answers.
are important, company officials might Asking questions awkwardly can pro-
want to measure whether employees are Online Resources duce misleading answers, notes Clements,
engaged in ways that prompt them to recalling one early survey question she
For more infonnation about emptoyee
expend extra, personal effort. engagement ano employee surveys, see simply calls "bad,"
Do some homework before talking to the online version of this articie at www "We asked, 'Rank your pay programs
vendors or consultants or striking out on . shfm.onVhrmaoazine. in order of importance from top to bot-
your own, Gibbons advises. Vendors cre- tom. ' " Employees had to rank base pay.

50 HR Magazine • February 2010

 fmployee Relations.
SPECIAL REPORT

incentives and other programs. When helps employees prepare for retirement, Beware of loaded questions or state-
some leaders saw the rankings, they not whether they are happy with specif- ments. An example is "Do you look for-
assumed that if employees ranked incen- ics of their 401 (k) matching, he says. Ask ward to going to work on Mondays?"
tives the lowest, for example, employees whether they believe their job security is Questions like that "elicit a 'no' easily."
didn't care about incentives. Not true, Cle- better at the company than it would be even from engaged workers, Federman
ments says, hut the forced rankings gave elsewhere, not what they thought of last says. Another example: "I have sufficient
that impression. "We fixed this by asking year's layoffs. time to complete my artivities. " This query
individual questions about aspects ofthe "Think about what actions the ques- isn't useRil because "no one has enough
pay program instead. like how we were tion could lead to," says fCaren Paul, time." And neither question asks about
doing delivering fair and conditions an employer can change, adds
equitable pay." Federman, also president of performance
Brad Federman, author improvement company Performancepoint
of Employee Engagement LLC in Memphis, Tenn.
(Jossey-Bass, 2009), outlines Avoid double-barreled questions.
two types of engagement These come about when leaders try to
questions. One type covers "You're trying pack two or more ideas together, often as
what he calls "core engage- a way to shorten surveys, Paul says.
ment issues, or 'Do I have to measure state Schneider offers an example; "We
what I need to do my job?' have neither the staff nor the resources
questions." The otber cov- of mind, not a to get the work done." If the answer is
ers "enriching engagement
issues, or 'Do you believe in
particular program "strongly agree." it is unclear whether it
is the staff or the resources that are lack-
the mission of the organiza-
tion?' questions."
or activity/ ing, he says.
Keep length reasonable. Campbell
HR professionals tempted to stick to head ofthe Global Center of Expertise for holds its survey to about 20 questions,
broad questions about mission should Measurement at 3M in St. Paul, Minn. A while 3M uses about 60 on its base survey.
realize tbat core issues of day-to-day "yes" or "no" answer to a generic state- Julie Gebauer, co-author with Lowman of
resources are vital to employee engage- ment like "My job demands are too high" Closing the Engagement Gap, advises clients
ment, be stresses. "If I don't have a doesn't give an employer enough informa- to limit surveys to 40 or 50 questions.
computer to do my work," Federman tion to act. Overly long surveys reduce participa-
explains, "I'm not thinking about tbe Keep language neutral or positive. tion rates and gamer skewed responses
mission ofthe organization." Careful wording can take an item from because participants tick off answers just
Survey questions should address negative to neutral, Paul says. "For to ñnish, says Federman. He knows of
enduring issues, not transient projects. instance, ask, 'Is our line-to-staff ratio one engagement survey that asked 356
Gibbons says. "It's less ahout 'Did you correct for a company our size?' not 'Are questions.
participate in training this past year?' and tbere too many staff for a company our Consider tailoring questions. If you
more about 'Do you feel you get oppor- size?' " Avoid negatively worded items. work with a vendor that comes to you
tunities to develop?' You're trying to Focus on behaviors. Good questions with a base list of questions, discuss addi-
measure state of mind, not a particular probe supervisors' and employees' every- tions to reflect particular needs.
program or activity." day behaviors and relate those behaviors Campbell uses a survey instrument from
to customer service whenever possible, Gallup and adds its own questions. "We
Tricks of the Trade Schneider says. leamed that managers are the most impor-
Employers, consultants and authors offer For example, responses to statements tant factors in employee engagement,"
some do's and don'ts for crafting engage- such as "My supervisor recognizes and Downes says, so the company added ques-
ment surveys: rewards good service to customers" or tions about whether managers link daily
Strike the difficult balance between 'When the bank gets busy, the branch objectives to business strategies, whether
detailed and generic. "Ask yourself, 'Is manager takes control" let leaders know they give effective feedback and more.
this a question I could ask every year?' If what's going on day to day between Know the audience. For example,
not, it's too detailed," Gibbons advises. employees and supervisors, Schneider asking "Are you allowed to focus on
For instance, ask whether the company says. your specialization?" may collect useful

February 2010 • PiR Magazine 51

Employee Relations.
SPECIAL REPORT

information if the person answering is a ask employees about their work/life bal- ees twice a year with a 45-question list.
specialist, but not if the respondent is a ance ... or their pay," but REI asks about The company also conducts monthly, five-
manager who no longer exercises a spe- both because those issues matter to REI's question pulse surveys that ask employees
cialist's skills, Federman says. employees and their bosses. REI asks to respond to broader statements such as
And if a question asks about the respon- employees if they find the company to "I feel like I'm making progress in my
dent's customer, it should be clear who be a fun place to work—a question Cle- career-" The twice-yearly surveys go into
that customer is, Paul says. "It's internal ments says REI's survey vendor told her more detail with responses to statements
if you're HR, external if you're in distribu- was unusual. such as "My supervisor tnakes sure I get
tion and soon." the support I need" or "My supervisor
Consider what you're gives me timely feedback."
saying about the organiza- 3M fields three kinds of engagement
tion's values. Survey ques- surveys, Paul says. One, done every other
tions put the organization's year, is for company leaders, A second
values on display, "Careful survey covers about half the employee
question selection is vital Once a survey population each year. A third set of shorter
because it tells employees surveys can be tailored for 3M operations
what you care enough to is fielded, in specific countries or business units.
ask about," says Downes,
Paul says employees
breaking down Feedback at Every Level
should be able to glance
through a survey and learn
engagement data Once a survey is fielded, breaking down
engagement data for each business unit
what interests managers.
That's one reason she orga-
for each business is important because local mianagers can
make changes that truly afîect engage-
nizes questions by topics unit is important. ment levels. Lowman says,
with clear headings—so Schneider advocates having line man-
employees easily can skim to see what If there's a characteristic of your cul- agers communicate survey results to their
matters to 3M, ture that's special, ask about it, even if it own employees—and training line man-
Ask for a few written comments. prompts questions that are not on HR's or agers to do so—while also requiring them
Some organizations include open-ended a vendor's list, Clements advises. to create action plans to respond to survey
questions—where employees can write Remember that the survey isn't set recommendations.
comments—at the end of surveys, REI in stone. Experiment, advises Ratner, "It Campbell expects all employees to
researchers use such comments to identify took us several rounds to get comfortable have engagement objectives in their per-
themes they might not have covered in the on what to ask,' she says. "It might be formance reviews so engagement is from
survey and might want to address the next some time before you get surveys that are the top down and from the bottom up,
year. Gardner says. truly useful." Downes says. Each business unit gets its
Online retailer Zappos.com provides engagement results and has to produce
space for comments on monthly surveys, Taking the Pulse "actions that are meaningful for their
and company leaders respond to some Consider doing more than one type of work teams."
comments on its intranet, says HR direc- survey, each with different questions, REI distributes survey results for every
tor Rebecca Ratner, SPHR, frequencies and audiences. 3M, hav- department, unit and store, Clements says,
Zappos.com hasn't been using engage- ing surveyed employees in various ways and forms action teams that include part-
ment surveys for very long. As a small since 1951, does this—and so does Zappos timers, hourly employees, managers and
company, engagement was easier to ,com. with far fewer employees. supervisors. Clements adds that HR col-
gauge, Ratner says. Then Zappos.com Pulse surveys are briefer, more frequent leagues in other companies tell her about
added about 300 positions in the past year, surveys that address specific issues or are engagement surveys that end up in lead-
and leaders decided about 18 months ago given to specific segments of the work- ers' hands and "the onus is all on leaders
that they needed to measure engagement force, and they can take place between to make changes ... but employees must
more formally. annual surveys. Gebauer notes, have a role in changes." she says. "We
Consider your culture. Clements says Zappos.com, whose managers create share accountability all the way down the
that in retail, "it would not be normal to surveys in-bouse, polls all 1,500 employ- employee line." QB

52 HR Magazine • February 2010

Copyright of HRMagazine is the property of Society for Human Resource Management and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.