When Small Samples Work

Sample size is one of the most important parts of good research. The larger the size of your sample, the more accurate your data will be. Small sample size can yield wild inaccuracies in the data, and those variations can lead to bad research, bad conclusions, and ultimately mistakes in your business plan.

When you perform quantitative market research, it is important to try to find as large a representative sample as possible. You want to avoid the chance of statistical noise interfering with the success of your research. Yet there are a few times that a small sample is acceptable when it comes to collecting your data.

Here are a few examples.

Customer/Client Satisfaction Surveys Polling your current customers and clients is an important part of market research. You want to find out how those that use your business currently view your products, services, customer service, etc. Although a larger sample is useful, an anomaly on these types of surveys – mainly, if your customers are unhappy – is still important.

Depending on the reach of your business, it may be relevant to find that any of your customers are unhappy so you can do what you need to change them. Lost Clients When you lose a client, their surveys are relevant no matter how large the sample. Every lost client or customer has important information to share.

It doesn’t matter if you only lose one client over the course of the year – why you lost the client is interesting, and may provide you with data that helps you improve your business further. Feedback Finally, small samples are useful when you are simply getting feedback on an idea that you already plan on running with.

For example, if you have a marketing plan and you want to find out how others view that plan, you can send them what you have created, and then poll them on their reactions. Although the insight you get may not reflect the feelings of the public at large, it is still relevant insight that may help you change your strategy.

In general, you will want to use as large a sample as possible – and try your best to make that sample represent your client/customer base. However, in some cases a small sample is still relevant. The above list represents a few of the reasons that a small sample may provide you with interesting information to support your research.