AnalysisParalysis_Congratulations! You have completed the task of designing and rolling out the survey. It has been a hectic few days while you kept things going smoothly. Now the posters and mailers exhorting people to tell what they really feel are pulled down. The entire organization has stepped up and answered the questions diligently over the last few days.

You whistle a preppy tune as you walk into office and settle down in your chair to take a look at the data. But before that you see that you have a mail from the HR Head: ‘Congrats on getting the survey done, can’t wait to get our long pending Engagement issues fixed based on your insights. Meet me tomorrow.’

Your heart skips a beat.

Tomorrow? Actionable insights? In one day?

This is a common enough situation that once the first hill of designing and conducting a survey is done; senior executives want things to move fast. By the very act of conducting a survey, you have raised the expectations of both senior management and the employees that ‘something’ will be done – and fast.

And that is where you, the survey and data expert, have to separate the wheat from the chaff. (Don’t be perturbed; this is exactly how experts are crowned.)


Sidestepping Data Pukes:

It is fairly straightforward to check the statistical validation of your testable questions. Stats101 at your B-school (you were paying attention weren’t you?) or even some high-school background will suffice. (or dust off your old statistics text book and brush up on Mean, Median, Standard Deviations and Confidence Intervals)

The time consuming part is analyzing the responses to your open-ended questions. Some of the sites like Surveymonkey provide you with a decent free-text analysis in their paid version. There are tools available online that will easily crunch through a few MB of data and spit out what seems to be on top of people’s minds. At the extreme end if you are really on a shoestring budget, import free-text into an Excel sheet. Split Text to Columns, remove the ‘noisy’ words and do a frequency sort. Voila! You have a (pretty imperfect) list of top concerns – but at least you know where to start focusing your attention.

But that is hardly enough. Quite often this is where executives (and consultants) stumble. An elaborate report with responses to every question represented in fancy looking frequency distributions is dumped onto a bewildered HR Executive.

At the end of the day this report has little or no value. It is just a “data-puke” which represents text in graphical format and conveys little in terms of insights or actionable information.


Slice and Dice:

The real insights won’t come from presenting data, but from looking at it in intelligent ways. In a way your data’s ability to actually provide you with intelligent insights has a lot to do with how well you have thought through your ‘research questions’. (No shortcuts here, sorry!)

Some of the intelligent ways to look for insights is to segment the responses based on criterion that have bearing on engagement drivers.

  • A Generational Slice: Todays workforce is like a melting pot of seemingly entirely different eras. The ‘independence’ babies are retiring. Gen-X is moving into senior management roles. Gen-Y is starting to constitute a substantial chunk of the workforce in companies. Chances are there are substantial differences in the responses given by people in each of these cohorts – not just because of their past experiences, their understanding of how things are or should be, but also because of what motivates them, and their perception towards concepts like collaboration, team-work, transparency and organizational loyalty.

Any survey dashboard, which looks at the workforce as one homogenous entity, is going to get it wrong. Each generation needs a different shade of engagement actions for it to resonate and extracting that insight from your data can be a very useful first step.

  • Department/Geographical Dice: Senior management likes managers who deliver business results. But they may not be the ones who are good at engaging their teams. It might be possible to obtain short-term gains derived by driving teams hard and establishing an authoritarian management style, but it is sure to hurt in the long term. Your survey might just throw up such ‘surprises’ and a deep dive into data will uncover such managers. Such insights are priceless in helping top management correct employee engagement snafus brought about by perceptions.
  • A Gender cut: Very useful when used in the right way. Very easy to get it all wrong and spoil your entire analysis. Differences in Gender attitude towards the research questions have to be dealt with correctly. Validate the data and stick to what the data is actually saying. It is easy to let popular perceptions or personal beliefs cloud your judgment while presenting the analysis. But having a proper understanding of what each genders perception of issues are will go a long way in establishing a truly equal opportunity workplace – with facilities and processes that help employee rise to their true potential.

Beware of Pyrite:

These just three of the possible ‘cuts’ you can make of the data. I must warn you that this can be pretty addictive and there is a risk of getting carried away and starting to look at the data from all kind of angles. Yes, those angles exist, but always bear in mind that at the end of it, any data/insight you present should be actionable within a realistic time frame. Don’t let this become an exercise to show number crunching or graphing abilities.

Presenting an ‘insight’ on the lines ‘67% of Gen-Y employees are disengaged and more than 80% of those disengaged feel disconnected with their manager, feel that the company is not working on correct technologies, is not competitive in pay and doesn’t have adequate work-life balance’ is liable to get you strange looks.

The real nuggets of insight are what matter. Fool’s gold is easy to find, but can send everyone on a wild-goose chase with nothing to show for all the effort.


Walk the Talk:

You have done a survey; you have analyzed the data and now its time to commit to action. The absolute worst thing you could do is taking no action. The trust deficit inaction would create will be nearly impossible to overcome.

  • Communicate: Before setting up teams to plan and monitor actions, communicate the results and insights from the survey – and ensure that top management gets involved in talking about it. Mailers, Intranet blog posts, Town-hall meetings the options are endless. The channel does not matter, the message does. When senior leadership is seen standing up talking about issues that are affecting engagement, insights they have gathered and what they plan to do about it, employees are reassured that the exercise was not done to tick off some checkbox in a ‘Best Employer’ participation requirement.
  •  Plan and Communicate: One way to make sure most (if not all) bases are covered is to set up a team that will plan the action calendar and monitor progress. Ideally this team should have representatives from all relevant cohorts (Generational, Geographical, Departmental, Gender) and be lead by some one in senior management. (You will need some firepower to get things moving at times when the usual priorities of sales, costs, projects etc. take over.) To keep the motivation within the organization, establish a transparent communication process, which shows what is planned, when and how much has been achieved.
  •  Check the Pulse: The follow-up detailed survey should not be done for at-least a year because the impact of most initiatives will take a long time to be felt. But pulse surveys – short surveys covering a few key questions (essentially a subset of your larger survey) are very useful in getting a sense of what people are feeling about the actions being taken and also providing positive reinforcement that the organization is serious about making change. To the team in overseeing the changes, these surveys provide valuable input to do any necessary midcourse-correction. Pulse surveys should avoid new or differently worded questions or it will be difficult to compare with previously obtained results.


These guidelines should help you design, conduct and analyze an insightful survey. But remember that a survey is a tool, like the thermometer. You might use an old mercury based one to get an approximate feel or a new digital one to get double-digit accuracy.

What really matter are the diagnosis and the treatment that follows.



Image courtesy of

We love discussions! Please do leave a comment on what you think about this post.

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.