top of page

(Sample) Size Matters. But, Maybe Not as Much as You Think (Part 2)

Part 2: Three times when a small sample isn’t a big deal

Common sense dictates that having a large sample size is always better for the outcome of any research. When it comes to scientific polling, trying to capture public sentiment on a political issue, for example, this is certainly true.

However, in Voice of the Customer research, it’s rarely your goal to extrapolate your findings to a population of several hundred million people. Because of that, you don’t need hundreds or thousands of research subjects to walk away with valuable information (and frankly, do you really want that if your goal is truly on the value-add you can provide to each and every individual customer?).

While more information is usually better, your sample size will often be limited by real-world constraints such as time, budget, and available participants. That’s why I want to point to three specific examples where you can still expect to get great results from your study, despite having a small sample size.

The Million Dollar Customer

If there’s an argument to be made for a sample size of one, it’s when your largest customer shares an opinion. Imagine if a very large client gave you some feedback and it was clear they would be shopping for another vendor if you didn’t adapt to meet their needs. Would you require validation from 100 other clients before implementing a change? I think not…

In this situation, you’d likely do what needed to be done IMMEDIATELY to keep the large client. Depending on the specific feedback, you may already know this is an issue other customers have and address it across the board. Or, it might be one that’s unique to this client and would not translate well on a larger scale. Regardless of whether or not the opinion or experience is shared by the majority of your customers, learning what matters to your very top ones can be a relationship-saver and may result in one or two accounts bringing in more revenue than all your smaller ones combined. If that’s the case, don’t let a small sample size stand in the way of your progress.

When you know your audience very well

Another instance of a small sample size not being a barrier to the success of your research is when you already have a lot of different data points on your customers. If you make an effort to ask for feedback on a regular basis and through a variety of channels, you may be able to generalize more broadly from a small group than a company who does not maintain that sort of connection to its customer base.

For example, if you send out satisfaction surveys after each interaction between your employees and customers and have a well-established trend of what’s “normal” for your business, even a small number of abnormal responses can be enough to raise red flags, then justifying a deeper-dive .

This can also be the case when the feedback is regarding a specific department or employee, not the company or its products as a whole. In the case of a customer service representative who normally gets four- or five-star ratings from customers, a cluster of lower ratings – even if it’s just a handful – can indicate a legitimate need for feedback and change without requiring hundreds of validating responses. Again, it’s a factor of how well you know your customers; how much data you already have about their opinions over time is more relevant.

When time and money are of the essence

If you’re a day late and a dollar short, you might not have the ability to realistically capture a large number of responses. When factors such as time or money limit your ability to gather data in quantity, you need to focus instead on quality. Not just “qualitative” versus “quantitative” data (although that can help too) but if you are only going to be able to get a small number of responses, make sure the ones you get are as representative of your customers as possible.

If you’re looking for feedback from customers of a specific product, focus your efforts on soliciting responses from them. The same goes for a geographic region or an audience demographic (age, gender, industry, etc.). When the timeline is tight, don’t spent your valuable time going after those who don’t meet your study’s needs. If you know you’ll only be able to capture a few responses, at least make sure the ones you get provide the best possible viewpoint for your objectives.


With all of that said, there’s rarely a downside to having a larger sample size. The more responses you get, the more patterns you can detect and the broader you can generalize your conclusions.

The point is, sample size isn’t everything. Ask yourself this: how many times do you need to hear the same feedback before you’re ready to make a change? If the answer is “once” then you don’t need to conduct 100 interviews. Said another way, clients often ask me how many study participants are “enough.” My answer is, “the number it would take for you to put their recommendations into action.”

To learn more about how we can design an individualized Voice of the Customer research program for your business, contact us now.

15 views0 comments


bottom of page