Monday 20 January 2014

Employee Opinion Surveys and Trust

Fact: Lack of trust causes employee opinion surveys to fail.

Employees do not trust that the process is anonymous, or trust that actions will be taken based on their responses. This leads to low participation and insipid results.

Listening to employee voice is a critical part of any culture transformation and employee engagement program - so getting this part right is crucial. As with any data source you need to focus on getting the very best quality.

I was once told by a concerned employee that the company recorded all keystrokes,  so responses could not be anonymous. He went on to explain that all negative or controversial comments within the survey were decoded and traced back to the originator using this method. For those done outside the company firewall, trackers were in the computer, and if you used your own computer you could still be traced using language and pattern recognition. 

My jaw dropped - I figured the guy had read WAY too many spy novels.

I asked him how anyone would have the time and resource to do this, and why it would be a priority for the company? Surely it would be easier just to ignore the comment than employ a team of covert black ops tech guys in the basement. He just smiled at me, and told me that he knew people who had been fired for being honest on the survey.

I gave up.



At another company, I was asked about the anonymity problem and invited a number of worried people to a meeting to discuss it.

I sat with the administrator of the survey and walked everyone through the process of invitation, collection, summarisation and distribution - in an effort to spread the message through the company that these things were genuinely anonymous.

Email adresses for participants were entered into the system - which then distributed questionnaires automatically (in this case to around eleven thousand people). As each was completed, the results were stored and the email reminders stopped. We showed the database record which clearly had no email address contained, just a unique number.

Concern number 1 - that the the number could be associated with the end user.....it couldn't, but people genuinely thought that a single response would be worth tracking back through the system and decoding for the originator.

The results were then all summarised, and presented in a report to each manager with more than 6 respondents in their organisation.

Concern number 2 - how can this be done without the individual names in the system? Easily, as the original submission came back, the managers name was added to each. At the end of the process, all managers with less than ten responses had their name removed.

Then reports were distributed by organisational hierarchy to each manager to share with their team.

Concern number 3 - my manager will be able to work out from my response who it is. Well, they may guess, but experience has shown me that more often than not they guess wrong.

Concern number 4 - I know that the CEO/Top Managment/My boss get to see the names...... No they don't.


Concern number 5 - The system records date and time of entry, so checking that against the office door entry log will help narrow it down. What? Really - who has the time and how would that help?


You have to accept that there will always be a few who don't trust the system. Especially if the delivery of the end results is flawed (managers yelling at teams who all gave negative opinions, managers singling out people the think left the only negative opinion, managers not bothering to communicate at all, and so on). 

But lack of action is the fastest way to kill any kind of value, trust and participation.

Think about it. People went to a lot of time and trouble to enter information into the survey. This may have been five minutes of time, it may have been thirty minutes - but it was a choice to commit.

Then nothing happens. No visible decisions are made, no changes occur. 

So here's my advice. Please do not run an employee opinion survey if you don't intend to take action at the end (even if it's difficult or uncomfortable to do).

Then when you take action, make sure you tie it back to the survey and let people know.

If you don't take action on a particular area, make sure you communicate your reasons.

Because you only get one shot. Next year participation will be lower, the content will be bland, and you will be wasting your time. It will cost you a lot for a worthless document that will further damage trust.



Side note: 

I  wish we could do away with anonymity completely in surveys, but I have only worked with a  handful of companies who can do this effectively - and they work really hard to ensure trust and comfort with direct communication. 

At one such company I  saw a comment that said "The CEO should be fired, because he is a liar - as evidenced by these three things..." The CEO looked at the comments, agreed with what was said and picked up the phone to the employee to explain. 

What a great company to work for.

1 comment:

  1. This comment has been removed by a blog administrator.

    ReplyDelete