There's an ironic paradox in new research released today that speaks volumes about the future of technological adoption, and especially how we research it.
The Pew Research Center finds that smartphone penetration has exceeded household broadband penetration. That's not the paradox, but it's one part of it.
The paradox is that Pew also announced it is changing the methodology it uses to research consumer adoption of technology for reasons related to it.
Instead of using telephone-based surveys, it has begun using a web and snail mail-based surveying method managed by consumer research giant Ipsos. The reason: few people answer their phones anymore.
And while that is an insight worthy of a 3.0 column in itself, it speaks to a fundamental shift not just in consumer technology, but in how consumers use the technology they adopt. Or don't.
Ever since landline telephones reached a critical mass of U.S. households in the 1950s, consumer researchers utilized the so-called "random digital dial" method to survey Americans about all sorts of things, including their media usage.
In fact, that phone methodology was more or less Madison Avenue's gold standard until about 15 years ago, when it started weaning itself from it due to the erosion of its efficacy.
Between do-not-call list registries, the acceleration of mobile phone adoption, and underlying shifts in consumer media usage behaviors, random digit dialing was becoming less and less effective.
This may seem quaint by today's "Big Data-plus" approach to consumer media and technology research, but it's worth reading a MediaPost op-ed written by the esteemed Josh Chasin back in 2009 to remember that perspective.
Back then, Chasin was Chief Research Officer of Comscore, and until recently he was Chief Measurability Officer of VideoAmp, but he's always been a forward-thinking media researcher.
I'm not sure why Pew's methodological shift took so long -- or why it coincides, paradoxically with its finding that smartphones are now the dominant way Americans access digital media -- but I'm reprinting part of a Q&A that Pew provided along with today's release so you can understand the rationale in their own words.
As important as Pew's finding as well as its survey methodology shift are for the next-generation of media technology, I'd like to focus the rest of today's weigh-in on an attendant part of it: how technology changes consumer behavior, not just because of what it can do, but because of the unintended ways people end up using them.
Think about the landline phone, as an example. When American households first got access to them, people were thrilled to answer them anytime they rang. Over time, that novelty wore out, and Madison Avenue played its own significant role in that, because, you know -- telemarketing.
So what once was a ubiquitous lifeline connection to American households quickly eroded into both a personalized media connector/player as well as a means of screening unwanted inbound communications.
The truth is that the smartphone has evolved to become many, many other things -- and certainly anything that "has an app for that" -- but an equally significant game-changer is its increasing role as a media hub.
We have evolved from the landline phone to the cable/MVPD "triple play" to the broadband-only one, and soon, to a smartphone-only one.
That last part is my prediction, but it is informed by a conversation I had with my nephew Josh Lovison about 15 years ago when he was then leading the mobile practice in the IPG Media Lab.
We were having a conversation about smartphone penetration and I was gushing about the explosion of apps, the app economy, and how apps were altering the behaviors of how consumers used their phones.
"Those things are important," Josh said holding his own smartphone up in front of me, adding: "But the most important thing is how the microprocessing power is expanding inside this device."
Josh went on to explain that smartphones would soon become more powerful -- and have greater bandwidth -- than the gateways we were then using to connect to digital media inside our homes: cable broadband modems.
That was one of a handful of lightbulb moments that have gone off in my own head during my years covering media, several of which came from Josh, but ever since then I've been waiting for the day when I could swap my wired household connection for a mobile wireless one.
I'm still on the "contact me when it becomes available" waitlist for Verizon's wireless mobile household modem as soon as it comes into my rural Connecticut neighborhood, and I can't wait to write a column about that once it gets "installed."
In the meantime, my Verizon-powered handheld devices are getting pretty close, and I find myself increasingly streaming large video files that I otherwise would have watched on my WiFi-connected TV screen.
To explain the thinking behind the change in its consumer survey methodology, the Pew Research Center provided the following Q&A between Managing Director of Internet and Technology Research Monica Anderson and Research Associate Colleen McClain.
Monica Anderson: We see this research as foundational to understanding the broader impact that the internet, mobile technology and social media have on our society.
Americans have an array of digital tools that help them with everything from getting news to shopping to finding jobs. Studying how people are going online, which devices they own and which social media sites they use is crucial for understanding how they experience the world around them.
This research also anchors our ongoing work on the digital divide: the gap between those who have access to certain technologies and those who don’t. It shows us where demographic differences exist, if they’ve changed over time, and how factors like age, race and income may contribute.
Our surveys are an important reminder that some technologies, like high-speed internet, remain out of reach for some Americans, particularly those who are less affluent. In fact, our latest survey shows that about four-in-ten Americans living in lower-income households do not subscribe to home broadband.
Why is your team making the switch from phone surveys to the National Public Opinion Reference Survey (NPORS)?
Colleen McClain: The internet hasn’t just transformed Americans’ everyday lives – it has also transformed the way researchers study its impact. The changes we’ve made this year set us up to continue studying tech adoption long into the future.
We began tracking Americans’ tech use back in 2000. At that point, about half of Americans were online, and just 1% had broadband at home. Like much of the survey research world, we relied on telephone polling for these studies, and this approach served us well for decades.
But in more recent years, the share of people who respond to phone polls has plummeted, and these types of polls have become more costly. At the same time, online surveys have become more popular and pollsters’ methods have become more diverse. This transformation in polling is reflected in our online American Trends Panel, which works well for the vast majority of the Center’s U.S. survey work.
But there’s a caveat: Online-only surveys aren’t always the best approach when it comes to measuring certain types of data points. That includes measuring how many people don’t use technology in the first place.
Enter the National Public Opinion Reference Survey, which the Center launched in 2020 to meet these kinds of challenges.
By giving people the choice to take our survey on paper or online, it is especially well-suited for hearing from Americans who don’t use the internet, aren’t comfortable with technology or just don’t want to respond online. That makes it a good fit for studying the digital divide. And NPORS achieves a higher response rate than phone polls.
Shifting our tech adoption studies to NPORS ensures we’re keeping up with the latest advances in the Center’s methods toolkit, with quality at the forefront of this important work.
Anderson: Are the old and new approaches comparable?
McClain: We took several steps to make our NPORS findings as comparable as possible with our earlier phone surveys. We knew that it can be tricky, and sometimes impossible, to directly compare the results of surveys that use different modes – that is, methods of interviewing.
How a survey is conducted can affect how people answer questions and who responds in the first place. These are known as “mode effects.”
To try to minimize the impact of this change, we started by doing what we do best: Gathering data.
Around the same time that we fielded our phone polls about tech adoption in 2019 and 2021, we also fielded some surveys using alternate approaches. We didn’t want to change the mode right away, but rather understand how any changes in our approach might affect the data we were collecting about how Americans use technology.
These test runs helped narrow our options and tweak the NPORS design. Using the 2019 and 2021 phone data we collected as a comparison point, we worked over the next few years to make the respondent experience as similar as possible across modes.
Anderson: What does your new approach mean for your ability to talk about changes over time?
McClain: We carefully considered the potential for mode effects as we decided how to talk about the changes we saw in our findings this year. Even with all the work we did to make the approaches as comparable as possible, we wanted to be cautious.
For instance, we paid close attention to the size of any changes we observed. In some cases, the figures were fairly similar between 2021 and 2023, and even without the mode shift, we wouldn’t make too much of them.
We gave a thorough look at more striking differences. For example, 21% of Americans said they used TikTok in our 2021 phone survey, and that’s risen to 33% now in our paper/online survey.
Going back to our test runs from earlier years helped us conclude it’s unlikely this change was all due to mode. We believe it also reflects real change over time.
While the mode shift makes it tricker than usual to talk about trends, we believe the change in approach is a net positive for the quality of our work. NPORS sets us up well for the future.