Why and How to Blend Sample Sourcing in Your Online Survey
Blog | January 30, 2024
By Meghan Ellrich
When writing a research paper, an author likely would not rely on one source for all the data. At OpinionRoute, we believe in this principle for online research. Every field plan we create is custom and includes blending vendors.
There are two main reasons we believe blending is the best approach.
- Blending samples improves data quality.
Let’s say you are running a survey targeting gen pop respondents with three sources. After the soft launch you do your first round of data cleaning and find that:
Vendor A: 10% QC rate
Vendor B: 40% QC rate
Vendor C: 5% QC rate.
Blending allows you to benchmark quality and ultimately take tangible action to improve data quality as the field progresses. In this case, you could remove vendor B and reallocate those completes to vendors A & C.
If you had just used vendor B on the study, you would be facing quality issues and lost labor from cleaning for the whole project. Timing will be at risk, and budget will be threatened as conversion rates drop from your removals.
- Blending can speed up the field, if necessary.
Especially at the end of the year, there is an influx of online studies fielding. Because the pool of available survey takers remains steady even in busier times, this generally means it takes longer to field studies in peak work months. If you tap into numerous sources on a given project, you can reach a broader net of people in less time. Also, working with a blend makes difficult quotas close easier.
How to do it the right way
It is important to note that while blending can prevent many issues in field, there is a right way to operationalize this.
- Deduplication – It is important to ensure you have the appropriate deduplication technology (such as CleanID) in place. It’s very common for individual vendors to use the same sub-sources on a given project.
- The Law of Diminishing Returns – Today’s online sample landscape is dominated by Programmatic sample technology. One of the key features of these platforms is API connections from hundreds of different sources. However, these sources are not exclusive to one platform. As a result, when blending too many sources for a specific target audience, the odds of overlap skyrocket after 4-5 sources.
Here’s a visual example from one recent study of ours.
*Data Source: OpinionRoute’s NavigatOR platform for Survey Operations.
So, if you’re a researcher who wants a smoother field experience with better quality, we encourage you to give blending a try. We can talk you through it!