Understanding the people
I'm not sure how many employee opinion survey presentations I have sat through or been shown. They tend to say things like '25% of people in our retail arm are engaged'. So what?
To get value out of these things we need to be a bit more sophisticated with the data. The way that the data is usually segmented (by division, sex etc) is interesting, but generally not that insightful. To get more information you need to look at creating your own segments, based on finding interesting clusters.
What you generally want to understand is the groups, the likeliness of somebody having an certain set of opinions. You probably want to drive this from a result that you are interested in - for example 'what do our high performers typically think?' or 'what behaviours do engaged staff show?' Basically, you're interested in segmenting people not by their demographics (gen X, Y etc) or division within the business but by attitude. This will give you an understanding of which organisational levers you can pull which will be most influential to your chosen group.
There are a few techniques that you can use for this, one of which is CHAID (Chi Squared Automatic Interaction Detector). SPSS's AnswerTree product is widely used here.
What have I seen? Well we've seen that high performers are generally most interested in organisational issues (leadership etc), where the average people are more focussed on their role. We have seen that people who feel underpaid generally also are keen for more work / life balance and have spent longer in their current job. What do your people think?