News

Determining Panel Size

When bigger isn't better...

One of the most common misconceptions in the panel building business is that a panel must be as big as possible. When initially asked, most clients will quote a desired panel size that is easily twice as large as it needs to be. Every client and circumstance is unique, but by examining a few key factors the optimal panel size, can be determined

Frequency of Contact
The first, and most important, factor you should consider is frequency of contact. Ten years of online research has taught us that while panelists do not like to be overburdened with research requests, there is also such a thing as too little contact. A high quality panel is about creating a relationship with the panelist and if they are only contacted two or three times a year it is hard to build that relationship. Our experience has shown us that panelists should be invited to participate in research, or contacted in some other way (i.e. newsletters), between 12 and 15 times per year. If you are planning on rolling out 60 studies over the course of the year, you will need to ensure that you have ample panelists available to include only a quarter of your sample in each study. Conversely, if you only have a handful of studies planned you will be able to return to the same sample group for each subsequent study.

Sample Size
The second factor to examine is the acceptable sample size per project, and your anticipated response rate. You will need to consider the level of analysis you would like to do on the results of specific research studies, which will lead to a decision about your acceptable sample size per project. From there, response rates are anticipated and factored in to determine what size of panel you require.

Target Audience
The third factor to explore is the necessity for, and composition of, desired sub-groups. Certain demographics will respond better than others (females over males for instance) and, depending on the recruitment method you use, low-incidence sub-groups can make larger panels necessary.

Refreshing Panelists
Fourth is your willingness to refresh panelist throughout the year by inviting new people to join. This is mainly a factor when longitudinal research is planned. If it is important to survey the exact same people at the beginning of the year as at the end then a larger panel size is required at the recruitment stage.

Sample Source
The fifth and final factor is access to, and cost of, sample for recruitment. This factor is most important where sample for your recruit is expensive. In this case it is a wise decision to start with a smaller panel size in order to determine more accurate response rates and attrition estimates before recruiting a larger panel.

SAMPLE SCENARIOS

Example One- Season Ticket Holder Panel
A sports team wants to build a season ticket-holder panel. They want to conduct four large studies during the year, but also plan to use the panel to poll their biggest fans about team issues as they arise- maybe one per month. Their analytic needs are pretty straight-forward, they need 300 completes per study, and they anticipate a high response rate (55%). They do want to have a number of corporate ticket holders, but can target them directly through the recruitment process. They are willing to boost the panel size throughout the year and access to the sample for the recruitment is free. Considering these factors we would suggest a panel size of 800 to 1,000 people.



Example Two- Customer Satisfaction Panel
A telecommunications company wants to build a panel to monitor the satisfaction of their various customer groups. They have six departments who are interested in using the panel as well as the management team of the company as a whole. There are about 30 research projects scheduled for the year. They do have several variables they will want to use for analysis and therefore feel they will need a minimum of 600 completes per study. They anticipate response rates to be slightly below average. The have 5 key sub-groups that must be represented, though they will be able to target them directly through the recruitment process. They are willing to boost panel size during the year and access to the sample is free. Considering these factors a panel size of 3,000 to 5,000 would be recommended.



Efficient panel size is one of the keys to a healthy and productive panel. Vision Critical's panel experts can guide you through the process of determining the best panel size that is right for you.

Posted on September 23, 2004
Permalink