Memo to the Engagement Survey Haters: Here is Why You’re Mistaken


It had to happen:  the backlash against employee opinion surveys, engagement surveys, morale surveys, call them what you will.  The sound you hear on the blogosphere in places such as HR Capitalist and Employee Engagement Network is a crescendo of voices saying that surveys are just so 1990, they are done with, finished, they never were that useful anyway.  We read of the thud of massive survey result feedback books arriving on organizations’ doormats once each year and everyone groaning that they have to, once again, wade through all this data.

As I look at the people who are saying this, there is a mix.  Some have done quite a bit of surveying and got sick and tired of it over time or had a bad experience once or twice.   Others seem to have no experience at all and yet wish to pontificate about surveys anyway.   Common themes are that:

nothing was ever done with it

–it didn’t relate to morale or engagement “on the ground”, in other words a really engaged team might come across as disengaged in a survey

–it took too much time

–it replaced one-to-one contact which is what workers really wanted

–surveys are “unscientific” in any case

This is a shame, but it is understandable.  The survey industry certainly has made some mistakes and in some cases failed to adapt to the times.  However, to throw the baby out with the bathwater, to discontinue surveying or decide not to do this if you have never done it, is a mistake.  The reason is that surveys are the ONLY method of collecting information about morale or engagement which meets all of the criteria shown in the table below.   Those criteria are:

  •  Providing a full “census” of all your people and not a sample.  You don’t want to sample because of one good reason, you leave out people who want to express their opinions.  Sampling also has myriad statistical issues which have to be resolved in order for the sample to be ……
  •  Generalizable to the whole;  you need to be able to say, the information we have here truly represents our whole organization or that part of it in which we are interested.
  •  Providing quantitative data.  As a result of having zero or minimal quantitative data, one cannot do crucial things like compare results to the past and see change (or lack thereof), make a benchmark for where things stand now, internally rank groups and find “centers of excellence” within your own organization and so on.  These are all incredibly valuable, if done by a knowledgeable practitioner with the right tools.

So let’s look at various methods of finding out the key question:  “how are our people doing?  Whether you are interested in engagement, morale, satisfaction, culture or climate, this is the basis of what you need to know.  Several methods exist for doing this:

  •  Ambush survey:  the boss (or boss’s boss) arrives in the elevator at the same time as you and quizzes you on how things are for the 30 seconds it takes to reach your floor.  Of course ambush surveys can happen anywhere, including on the phone.
  •  Casual chat:   your boss sees you in the cafeteria and sits down for a talk.
  •  Group/Team meeting:  more structured, focused on issues, etc.  May be peers only or may be attended by management.
  •  Self directed group:  this is a team which runs its own affairs, usually handling all that HR used to do, including hiring and firing, setting goals, internal people management issues, etc.
  •  Focus group:  a sample of people gathered for a task such as identifying issues for an opinion survey questionnaire.  May or may not be structured to statistically represent a larger whole, of which these people are part.
  •  Employee opinion survey:  called by many names including all those above (engagement, etc.).   Web-based and individually completed, paper and pencil and group-completed or combinations of these.

(Note: I left out social networking methods, partly because they are relatively new and we have no data on how they work and secondly because the issue of confidentiality looms so large for them I am not sure that until that is resolved, they will be so useful in gathering really solid data; they may also subject people to a feeling of being over surveyed, resulting in “survey fatigue” and the loss of good data; that remains to be seen).

Each method above has strengths and weaknesses in relation to the criteria above (full census, confidentiality and numerical data).  If you choose to run only focus groups, for example, you have several problems.  Often it is very hard to make them statistically representative of the whole.  Then you need a skilled facilitator running them so that the “talkers” do not dominate with their endless stories about…whatever.  Even then, those who are shy but may have extremely valuable information may not speak up.  Also, those who are not necessarily shy but fear their views will be “outed” if they complain about something, keep quiet.  Focus groups are very time consuming, yet even within an hour period it is impossible to cover more than a few topics in depth.  You can see, the list is long and the issues are big.  Yet people on the blogs insist they can “talk to people” to get the information they need about engagement, morale, etc.

The table below summarizes these issues for each possible method of information gathering:

Methods of Gathering Morale and Engagement Information **

Sample or 100% Census? General-izable to the Whole? Anecdotal Or Numeric? Structured or New Information Possible? Confi-dential? Allows for In-Depth Study of Issues?
“Ambush” Survey Sample (small) No Anecdotal New Information limited by “Fear Factor” No No
Casual Chat Sample (small) No Anecdotal New information, but limited No No
Group/Team Meeting Sample (very limited) No Anecdotal New information, but limited No Yes, can do over time
Self Directed Group Sample (very limited) No Anecdotal New information, but limited No Yes, can do over time
Focus Group Sample (usually limited) Limited Anecdotal New information, but limited No Yes, but limited by time
Employee Opinion Survey Typically 100% Census Yes Both Both Yes Yes

As you can see, surveys of your workforce win hands down.  They are the ONLY method of finding out “how our people are doing” which meets all the important criteria.  Anything else leaves you with less, sometimes far less.  The information you then have is second rate, even wholly misguided. Can you then base decisions on it?

The table is a summary, and does not even begin to cover the psychological factors which come into play when one human being talks to another: issues such as selective listening (filtering out what is uncomfortable to hear, or what does not fit into preconceived notions, for example).  Of course surveys can also be mis-interpreted, but professionals in this area know how to help internal or external clients face the data, and such data is much more difficult to fudge in most situations where it is openly published, as it was with all our clients as a condition of our working for them.  Try doing that with information which is collected in a quick chat between an employee and her boss’s boss in the elevator or on the telephone, which can be immediately forgotten, or turned into something quiet different, etc.  Alternatively it can be, and often is, incorrectly generalized to larger groups when in fact it represents just one person.

I challenge anyone: let’s take an organization of 5000 people scattered here and there and you go and talk to them and I will survey them.  I’ll make it so that you win if we both come up with the same information…how’s that for a deal?  But if my survey tells a different story in any area (geographic) or on any topic, we will have the issue of which is “correct”.  The tie breaker will be like cards: confidentiality is the Ace, and if you cannot provide that in the method you used to collect the data, you will be trumped.  Much research shows that confidentiality changes the data you receive, so try ensuring that with your “talks”.  I already know the winner here:  the survey.

No matter what the gurus on the blogosphere say, there is no substitute for surveying your people at work.  Far from being “unproven science”, good surveys are heavily researched and validated (see below);  they provide far better data than any other method in my Table, above.  Some bloggers make the amazing point that they conducted an engagement survey and found that the most engaged groups came out worst on the survey; this was then used to invalidate the whole process!  This is Organizational Psych 101.  Good surveys have to be carefully constructed:  they must pass tests of validity, for example, whereby they are shown to measure what they say they measure.  Using a bad survey which does not correctly identify highly engaged groups is a lack of methodological rigor for one instrument, not a failure of employee surveys as a whole, which provide extraordinary benefits to smart organizations worldwide.

When I was full time in the business of running such surveys, my clients wanted and received these benefits and took full advantage of them.  They had early warning of things in Tokyo or in their ER; they knew what was happening in the heart of their German operations or in the building next door; they knew which of their 100 hospitals was doing a great job and why (meaning they could use that to help other hospitals); they could compare and rank morale and engagement for each sales group across all operations, whether around the state, the country or around the world.  Surveys made all of this possible.

All surveys are not created equally, that is for sure.  To be any good, you need excellent, heavily tested and validated questions; you need to cover the crucial issues in that organization; you need a process which is quick and efficient; you need comprehensive and totally honest feedback; and you need the survey to be part of an overall effort which puts people first and wants to create absolute excellence in that relationship.  A survey should never be a stand-alone effort, an end in itself.   No serious practitioner would ever agree to that, and in fact if I felt a potential client was not going to follow-through, I refused the assignment since I did not want my firm associated with such a failure.  My best clients saw surveying as only a part of their goals of being world class in terms of a high morale culture, which was lived and breathed by everyone who worked there.

Managers running around chatting with individuals is a great thing and absolutely necessary on a day to day basis.  But to use it in place of a survey when you really need to know “how our people are doing” on a large scale (and who does not need that at least as much as they need to have a handle on their finances?) is like someone taking snapshots of people at their workplace. Well constructed and run surveys, like those of the really good practitioners in this field, are a whole-organization MRI or CT-scan which can be focused in vivid 3-D detail on any part of that organization.  That’s the difference.  And that difference is why Google surveys its people yearly and monthly, and every year wins the Best Places to Work For contest in the US and elsewhere.  Not a bad track record for the much maligned survey….

Follow psych4biz on Twitter

LinkedIn profile/contact:

Subscribe in a reader


**  Table above excerpted from Employee Morale: Driving Performance in Challenging Times by David Bowles and Cary Cooper. Copyright © 2009-14 by the authors and reprinted by permission of Palgrave Macmillan, a division of Macmillan Publishers Limited. All rights reserved.

Picture of Google UK from


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s