Don't trust those survey numbers
Let's say you have a company with a bunch of employees. Business is "super super awesome", and everyone is really happy with how you are running things. Some magazine comes along and polls a subset of your employees to see how things are going. It stands to reason that the resulting figures would paint a picture of an awesome place to work. They'll "peg the meter", so to speak, and come out on top.
Now let's say things change, and now they aren't so awesome. Pretty soon, those same surveys will start picking up the people who are disenchanted. The company's numbers will drop off, and it will fall out of the top spot. It might even fall out of the top ten.
Unhappy people don't stick around in bad situations unless they are also depressed and unable to take action. Maybe they stick around longer than they should because of "the economy", but even that excuse runs out of steam eventually. Sooner or later, they quit.
Another year passes, and another survey is held. This time, the company again gets high marks. Everyone pats themselves on the back and feels that everything is fine. Yay! We did it!
No, things are not fine. You merely chased out the people who soured when you started screwing things up a couple of years before. You have now filtered the system of all of those annoying people who point out uncomfortable things, and now you can soak in the rewards of a nice survey result yet again.
You also now have a really neat echo chamber.