Lessons Learned: A Comparison of RDD Sample to Panel Sample

Case Study

Recently, an Ipsos Insight client asked us to substantially reduce the cost of their tracking study while maintaining the quality of their data. Our goal was a total cost reduction of almost 25%. We looked at all the price drivers involved, and determined that simply saving money by reducing the sample size would not work. It was an initially attractive option, but we believed that the ending sample would be too small, and would consequently compromise quality beyond acceptable standards.

The study was a telephone tracker that had been using RDD sample and we believed that by using our Ipsos Insight HTI Panel, we could achieve similar results, while saving the client nearly 10%. There were many inherent advantages to using the panel in the type of tracking study in question, but we wanted to make sure that the results would be nearly identical, and lead to the same business decisions.

In particular, we were interested in seeing if the attributes or brand characteristics were ranked differently between the two groups. We anticipated that there would be some variance in level, but if the same rank existed between the RDD group and the HTI Panel group, the same business conclusions would be made.

The survey introduction was identical for both groups except panel members received their call from Janet Hall, which is the customary procedure for panel members. Completion rates, as expected, were significantly higher for panel members (19.7% vs. 2.4%).

Our Findings

  • As expected, the HTI Panel had a higher cooperation rate, which improved productivity significantly.
  • The effects on the data are minimal - only minor shifts in the data are evident.
  • Unaided awareness was slightly down for HTI Panel members when compared to the RDD group. After that initial unaided awareness, the numbers lined up
  • The attributes ranked the same; brand characteristics were also ranked the same.

Conclusions

  • Changing a sample frame from RDD to an Ipsos-managed panel can result in significant cost savings, due to higher respondent cooperation rates, which improve interviewing productivity (and can reduce timing). Moreover, the effects on data are minimal, and the business decisions made based on the data are not altered.
  • For tracking studies, a switch from RDD sample to panel sample is attractive since the purpose of a tracking survey is more to monitor change in market share and usage than to read absolutes.
  • If available sample allows, it's possible to control such factors as age, income or gender, and even "technographics" and "Webographics."

What Our Client said:

"The cost saving was significant enough to allow the addition of survey material that had previously been handled in another tracking study. The combination allowed a richer analytic framework providing a more holistic set of business decisions."