By Claudia Perez | November 2024
Dear small and medium business owners: in this series of posts dedicated to marketing research, today I’ll discuss the survey process and sample selection—key elements in effective marketing research.
The first question we might ask is: Is More Always Better in Marketing Research?
On the surface, it might seem logical to think that more responses lead to better or more accurate results. However, in reality, this reasoning doesn’t quite work that way. First, limiting sample size in market research is essential for optimizing time and resources, making the cost-benefit equation more balanced.
Secondly, we must consider the law of diminishing returns, which states that, beyond a certain sample size, each additional response adds little value to precision but increases costs. Third, an appropriate sample size helps avoid data oversaturation and ensures the quality of the information obtained.
So how should we carry out a research process?
As we have discussed before, dear readers, surveying an entire population or a very large portion of it is impractical in terms of time, resources, and outcome optimization; that’s why we work with samples. Sample selection processes must be rigorous and include several steps, such as: Identifying the research population of interest, specifying a sampling frame, choosing a sampling method, determining the sample size, and implementing the sampling plan. A properly selected sample allows the results to adequately reflect the diversity of the target population and be applied to the population as a whole, which is a basic purpose of the research exercise. An optimal sample size reduces the margin of error, making the results reliable without needing to include the entire population.
Examples in Focus
Let’s briefly see how these concepts are applied in a real example through our “Examples in Focus” section in this weekly column.
Suppose, hypothetically, that PetDelight, a company specializing in pet food, launches a new premium dog food aimed at improving digestion and coat health. After launching the product, the company conducts a survey to gauge customer perceptions (clearly, the pet owners) to make future product improvements and refine marketing strategies.
Applying what we’ve learned, PetDelight will not survey everyone because it would be impractical in terms of cost and time, and the results might be poor or unreliable if the surveys aren’t conducted with the right audience. What impact would there be, for instance, in surveying someone who doesn’t own pets or hasn’t heard of the brand? Considering this, the company will survey individuals from its target population, which includes all customers who have purchased PetDelight’s new premium dog food within the last six months. This population represents dog owners with recent experience using the product who can provide valid opinions on its effectiveness and quality.
However, PetDelight will not survey its entire target population either. Instead, it will define a sampling frame (think of this process like a funnel). For this, the company might use customer purchase information, which includes both online sales records and in-store purchases registered in their loyalty program.
Next, the company will continue with the sampling process steps, establishing a sampling method that may be probabilistic or non-probabilistic. Suppose they use quota sampling within Non-Probabilistic Sampling to understand if satisfaction varies between owners of small and large breeds or between urban and rural dog owners. The company still has one more task: determining the sample size. PetDelight calculates that a sample of 350 customers is enough to achieve a margin of error of +/-5% with a confidence level of 95%. The company notes that as responses accumulate, precision improves. However, once they reach 350 responses, data precision varies little, but costs increase, marking the point of diminishing returns.
Finally, the company will implement the sampling plan and stop data collection once they’ve reached 350 responses.
Final Thoughts
In conclusion, dear small and medium business owners, limiting sample size not only optimizes resources but also ensures that the results obtained are precise, reliable, and representative of the target population. Through PetDelight’s example, we understand that a well-defined sample and a rigorous selection process allow us to obtain valuable insights without falling into the trap of seeking quantity over quality.
So I encourage you, dear readers, to apply this strategic sampling approach in your own marketing research. If you’d like to learn more about best practices in Marketing Research, you can refer to my post, “What If Starting with the End Could Make Your Research Plan and Business Outcomes Unstoppable?”
Remember always, dear readers, that a well-chosen sample can be the key to a deep understanding of the market, allowing you to respond to your consumers’ needs with precision. With that, I bid you farewell. I’ll see you next week as we continue exploring the fascinating world of marketing.