Survey response rates, or participation rates as we prefer to call them, are themselves often a measure of engagement; a measure of how engaged employees are with the survey and feedback process. We’ve also done some analyses and found a moderate relationship between participation rates and employee engagement levels. However, the relationship is nowhere near perfect and there does appear to be a trade off effect above the 85-90% mark.
For example, we recently found a company with the lowest engagement score (in the last 12 months) had a 100% participation rate. In fact, a 100% participation rate is often a bad sign and it is usually accompanied by comments about coercion and a large number of missing responses or unmeaningful responses (e.g. all ‘strongly agree’ or ‘strongly disagree’). So, having a 100% response rate is sometimes associated with poorer quality data – and it is quality of data that we should be focused on, not the quantity.
In employee surveys we are not trying to generalise from a relatively small sample to a much larger population like we often are in customer or political research. In employee surveys we are mostly talking about participation rates well above 50% (which is about the rate achieved in US elections) and generally in the 65-85% range. These are rates that political scientists and market researchers can only dream of.
So what is a good participation rate in employee research? In small companies or teams (<50) we should be aiming a little higher, perhaps 80-90% is a good minimum benchmark allowing us to hear from 4 out of 5 people on average. And as we move to larger companies we can scale our expectations down – with 500 employees we will probably get a good sense of where we’re at with 70% of employees so 70-80% is a good benchmark. Moving up to companies of 1000+ we can probably aim around 65% as a lower bound – even though higher rates allow a stronger sense of involvement psychologically.
Sometimes there will be reasons for even the above participation rates to be difficult, and one can certainly still get good statistical estimation from smaller samples in large companies. However, putting participation rates into perspective is often one of the first steps towards a healthier attitude to People Intelligence data your organization can make. And finally, in the medium and longer term the best way to increase participation rates is to share results quickly and openly and demonstrate a genuine intent to make practical changes – those who didn’t respond this time may just be ready to join in next time.