Why Calling Landlines for Market Research is Inaccurate

The Upfront Analytics TeamTrends

Why Calling Landlines for Market Research is Inaccurate

The Death Spiral of Landline Continues

Over four billion dollars was spent on collecting market research through telephone surveys during 2013[1], 70% of which is estimated to come from landline calls[2]. The market research community continues to rely heavily on RDD landline surveys, even as reduced landline coverage rates call the validity of the method into question.

How Bad Is It?

According to the latest National Health Interview Survey from the CDC, the number of households in the United States with a landline telephone has been decreasing steadily over the past 10 years-from 68.2% in 2010 to 56.4% in 2013. As of December 2013, 39% of adults and 47% of children were living in wireless-only households:


A whole generation of Americans who’ve grown up in wireless-only households will soon be buying their first cars, renting their first apartments and exercising their right to vote. At least we think they will. The fact is, the majority of telephone surveys conducted today are insufficient to capture the attitudes and predict the behaviors of the wireless-only population.

And it isn’t just children. According to the CDC, all of the below groups are more likely to live in wireless-only households:

  • Younger adults: 53% of 18-24 year olds, 65.7% of 25-29 year olds and 59.7% of 30-34 year olds live in wireless-only households.
  • The less affluent: 56.2% of those classified as ‘poor’ and 46.1% of those classified as ‘near-poor’ live in wireless-only households.
  • Ethnic minorities: 53.1% of Hispanics and 42.7% of African Americans live in wireless-only households.
  • Renters: 61.7% of renters live in wireless-only households.

Why Is It Happening?

In an effort to understand the underlying reasons for landline abandonment, we asked 270 of our wireless-only participants why they don’t have a landline telephone. The most common reasons provided were that they canceled because it wasn’t being used (37%), they never had one at their current address (33%) or they canceled it because it was too expensive (11%).

The reasons provided varied significantly with age. Participants aged 18-34 were most likely to answer that they never had one at their current address, while participants aged 35-64 were most likely to answer that it was canceled because it wasn’t being used.

What Does It Mean for Researchers?

The accelerated rate at which some demographic groups are becoming wireless-only forces researchers to rely more heavily on weighting techniques and reduces the likelihood of picking up important attitudinal and behavioral changes in the population.

According to Pew Research, results from their landline-only samples in 2010 were already significantly different from their combined landline-wireless samples. In particular, they found that landline-only samples underrepresented presidential approval and unemployment rates, and overrepresented desktop computer ownership and internet usage. The latest CDC report also found significant differences between landline and wireless-only households across a number of important health factors, including alcohol consumption, cigarette consumption and health insurance coverage.

What Can We Do?

A number of leading innovative market research companies have switched from landline-only to combined landline-wireless samples. Unfortunately for researchers operating in the United States, federal law prohibits using an auto-dialer to call wireless phones, making it 1.5 to 2 times more expensive to gather wireless responses[3]. This additional cost can be prohibitive for smaller companies.

Another option is to supplement or replace landline-only surveys with methods that capture the missing demographic segments. A number of firms are now combining multiple modes-in-person, telephone, internet, mobile app-to allow hard to reach demographics to participate using whatever medium they have access to and whichever methods they are most willing to engage with.

[1] http://www.esomar.org/web/research_papers/book.php?id=2635

[2] http://www.esomar.org/web/research_papers/book.php?id=2589

[3] http://www.people-press.org/methodology/collecting-survey-data/cell-phone-surveys/