The Cellphone Challenge

Until the 1970s, face-to-face was the preferred method to conduct quantitative market research studies. But as telephones became so prevalent, it was more efficient for most studies to be conducted by phone.

In the late 1990s and early 2000s, the Internet emerged as the newest opportunity. Some studies that would have previously been conducted by phone were done online. More and more experts started (and have not stopped) talking about the opportunities and dangers posed by this method.

Now another opportunity -- and challenge -- has arrived: the emergence of cellphone-only households. Almost 15% of American households do not have a landline phone and rely solely on a cellphone.

So what is the solution?

For random-sample studies, we can add a sample portion where we call people on their cellphones. For cost-related reasons, we typically screen out people who have a landline and only include those people in the cellphone portion who cannot be reached on a landline. Afterwards, we weight the data accordingly. These cellphone portions are substantively more expensive than landline portions, for three reasons:

(1) We typically suggest an incentive to the respondent since we cut into their everyday minutes and most people are on a limited-minute plan. (This also lengthens the survey since their address needs to be taken down.)

(2) This is a population that is hard to reach: young people are less likely to take time to respond to surveys than older people, so the response rate is typically lower.

(3) Once we reach a person, we need to make sure he/she is not distracted by traffic or other things people do when talking on their cellphone.

If we do not adjust to this development, we are in danger of missing a crucial segment of the market: the cellphone-only segment tends to be younger and more mobile. If we do not include them in our samples, our data would be incomplete.