Cell Phones in Telephone Surveys: The New Imperative

telephone surveys sampleWhile the introduction of the online survey met with breath-taking predictions about the demise of telephone surveys, time has shown those predictions to be false. There are still times when a phone survey is the best methodology. Whenever you have a population that is relatively low-incidence, in a small geographic are, not computer savvy, or low literacy, phone surveys have the clear advantage over other methodologies. So don’t kid yourselves, there are still plenty of phone surveys conducted in our industry.

However, there is also no question that mobile phones are our way of life. Over 90% of American adults have a cell phone, and 68% have a smartphone. Americans spend between 4 and 5 hours per day on their phones, checking them an average of 46 times per day – nearly three times each waking hour. (If that’s the average American, just think what younger consumers are doing.) According to the latest estimates, nearly half (47%) of all US adults live in cellphone-only households, and that is increasing.

We are so comfortable with our phones we see our phones as a natural extension of ourselves. Do you feel anxious when you are away from your phone? Does the thought of losing your phone strike fear into your soul? Treating mobile phone dependency is a new area for mental health professionals (check out “nomophobia”).

While as recently as five years ago, it was acceptable to conduct a telephone survey without including cell phones, that debate is over. If you’re doing a telephone survey, you have to include cell phones in your sample.

Unfortunately, cell phone sample presents many challenges and costs for marketing researchers:

  • In the past, due to the prohibition on using automated dialing for marketing research surveys, cell phone respondents had to be manually dialed, making that sample relatively expensive compared to landline telephone respondents.
  • Additionally, there were concerns about the ethics of calling cell phones for marketing research, given that many individuals had to pay for incoming calls.
  • The fact that cellphones are – well, mobile – and not tied to a physical geographic location contributes to making sampling cellphone respondents more expensive.
  • Finally, the incidence of pre-paid cell phones (“burner” phones) presents a mixed blessing: while they are typically used by under-sampled populations, they also have a very high non-working rate.

Many of these obstacles are being mitigated by advanced by sample providers, but the fact remains that you will still pay a premium for cell phone respondents. None the less, leaving cell phone respondents out of your sample results in significant demographic biases, as cell-only households tend to be younger and more urban. So the only way to control costs is to limit the proportion of cellphones in your sample. In other words, how low can you go?

Pew Research Center recently announced that it would conduct its 2016 telephone surveys with 65% cell phone respondents. That is probably a gold standard that most commercial research budgets won’t allow. But you should probably target between 35% and 65% cell phone sample, to be on the safe side.

Three considerations (other than cost) in determining the correct proportion include:

  1. Who is your qualified respondent? If you are targeting Millennials, you need to be generous in accounting for cell phones and, indeed, not including cell phones could prove to be a false economy. On the other hand, if you are looking for more senior respondents, you will probably be OK at the lower end of the range.
  2. Where is your qualified respondent? If you are surveying in a narrow geography, find out the percentage of cell penetration in your specific area and adjust your cell sample accordingly.
  3. How much risk is there if you under-represent cell-only respondents? While this is much harder to answer, consider the type of survey you are conducting, and the precision and accuracy you need in your data. This might be a risk you worth taking, given the cost difference in sample.

All marketing researchers must keep in mind that this is a moving target. Just as broadband penetration made online surveys viable, increasing cell phone usage and cell-only households are making including cell respondents in phone surveys a requirement. There probably will be a day when all phone research is conducted via cell phone. It is never easy to accommodate moving targets, but ignore them at your peril!

Read our blog article on mobile surveys

Leave a Reply

Your email address will not be published. Required fields are marked *