How many responses do I need for my research?
In my last post, I mentioned a number of principles that you should keep in mind when considering market research in the tourism sector – actually, they’re good advice for any sector really.
Getting more specific, one of the things that clients are always concerned about is the number of people surveyed or, as it is known, the size of the sample. Put simply, the more responses you get, the more accurate and robust the survey. Except in some cases, it isn’t.
If you are considering doing research, the first question then that you should ask is, “why do I need this?” Your answer will immediately go some way to determining how many people you have to bother for your research.
If you want to know whether your customers think your establishment is looking a bit tired, then you just need to ask your customers.
In this case, the more of them you can get to tell you the better but generally speaking, I think you should be looking for a trend in the responses rather than an absolute score. It’s all very well scoring a hypothetical 3.9 (by whatever scale you chose to follow) on bathroom cleanliness but the real story comes if this has been constantly falling from 7.1 on your scale six months ago.
If, on the other hand, you are looking to develop into a new market, develop a new product or generally do something that isn’t what you are doing at the moment, you need to think about not only using your existing customer base but also looking beyond them and this is where the question of sample size becomes more important.
I’ll let you into a little secret at this point. As a rule of thumb (and this is a very broad rule of thumb) 30 responses will give you a rough indication of the way the wind is blowing. The more responses you get, the more accurate this will become but it might just be the case that a good pointer in the right direction is all you need.
BUT…if those responses include responses from your mum, dad, brothers, sisters, aunties, friends etc, then their validity goes down. Like it or not, most people tend to want to be kind to people they know well and tell them what they want to hear. If you get 100% of responses telling you that your fuchsia-mottled Disney wallpaper is lovely, it is a near statistical certainty that there is an element of lying going on. So, the people you speak to must be ‘ordinary people’ and not people close to you (although there is also a perfectly valid test dubbed ‘the mother-in-law test’ but I’ll leave that for another day…).
This shouldn’t be read as saying that market research with samples of more than 30 people is a waste of time – far from it. As well as increasing the accuracy of findings, larger samples generally allow greater depth of segmentation, meaning that you have more reliable data to confidently discover the weird and wonderful niches among your customers. However, if you just need to get a feel for the way the market is moving, this might well do the job for you – just don’t expect it to be that accurate.
If you want to get scientific about what a good sample size is, go to this link for some mathematical wizardry that gives you an idea of how big sample sizes should be.
The Internet has revolutionised the ability of companies to conduct surveys. My next post will deal with the pros and cons of the Internet as a survey tool and discuss some of the alternatives.
This entry was posted on Wednesday, August 22nd, 2007 at 5:35 am and is filed under Tourism market research. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.