Why use your own research panel?
Setting up your own research panel has a number of advantages:
- You have access to a pre-qualified set of respondents from relevant backgrounds
- They’re ready and willing to help – providing genuine, considered responses at good response rates
- Fidelity over time – you can expect like-for-like results on repeated tests, since they’re from the same cohort of people
How to recruit panel members
Your first stop when looking for panel members should be contacts you already have, such as past customers or sign-ups to your mailing list. You can also advertise that you are looking for members on your website, social media channels or at the point of sale.
These are free or low-cost ways to recruit, but they will tend to be biased towards customers who have chosen your brand above the competition.
To go beyond your existing contacts, you can use ads, lead generation services, affiliate networks and advertising networks, but bear in mind that this could be costly.
Who should be on your panel?
After recruiting panellists, you will need to qualify them to make sure they’re suitable for your research. Put together a short, user-friendly profiling questionnaire to help you get to know your panellists better. Ideally this should be no more than 5-7 questions long and include basic demographic information such as age or income bracket.
Read more: How to write good survey questions
How big should your sample size be?
Determining your optimum sample size is a delicate art. You need to ask enough people to get a rounded picture and even out anomalous or outlying results that could skew things. And bear in mind that not everyone you survey is going to respond.
At the same time, you don’t want to make your sample too large, as this can be costly and produce diminishing returns.
To determine your sample size, begin by gauging the total number of customers you’re aiming to represent. Looking at your own sales figures and CRM database can help with this.
You’ll then decide on how accurate your results need to be – your margin for error and confidence level. Crunch the numbers to find your ideal sample size. Our sample size calculator might be helpful here.