Are You Buying Skewed Panel Data?

In yet another fascinating case study, Mr Green’s blog shares a mobile marketing campaign aimed at recruiting panel members for a demographic research service.

What’s fascinating is the bit about optimizing the campaign by device, carrier etc. Some phone models convert better than others, and not always for reasons of usability. Shoemoney years ago shared a case study of targeting ringtone offers by carrier, and showed how he dug into the demographics of each carrier to find the demographic most likely to convert on his offer.

While that’s well and good for affiliate marketers, it’s another thing entirely for customers of panel measurement services. If you’re buying data, you want it to be representative of the audience you’re researching, for ex.: smart phone users in general.

But if the panel is disproportionately heavy on one particular model or carrier etc, relative to the general audience you’re researching, the data you’ll be buying will be skewed. Of course, you can then slice and dice the data to solve the problem, but that depends on the panel measurement service enabling that.

Another related issue is panel recruitment based on giving away freebies. That also attracts a skewed demographic that may exclude the wealthiest elements of the population.

If you liked this bit of critical thinking, then check out my advanced SEO book, which teaches Advanced SEOs’ 7 Rules of critical and creative thinking.

Leave a Reply