The research industry has a quiet under-35 problem.
Walk into any methodology review at a top brand, and somewhere in the appendix you’ll find a footnote conceding that 18-34s were “harder to recruit” or “weighted up.” Translation: the panel didn’t have enough of them, so the model is doing the heavy lifting. That’s not a sampling issue. It’s a delivery-channel issue.
The under-35 audience has migrated. They live on mobile. They don’t open desktop email recruitment links. They don’t sign up for landline panels. Many of them haven’t owned a desktop computer since they left university — or never owned one at all. The traditional research industry was built for a generation that did. We were built for the one that didn’t.
Where the under-35s actually are
Smartphones are the dominant computing device for everyone under 35 worldwide, and the only computing device for a fast-growing share of them. In emerging markets the share is even higher: the first device a Brazilian, Nigerian, or Indonesian Gen-Z respondent ever owned is, in nearly every case, a phone.
Desktop-era panels were optimised around the household — one PC, one shared email address, one verifiable identity per household. Mobile is the opposite: one device per person, used hundreds of times a day, in fragments of attention. That changes everything about how you recruit, how you screen, how you engineer the survey UX.
If your sample partner is still routing to recruited panellists on desktop, you’re under-sampling the audience whose opinions decide market launches and elections. Weighting up survey respondents you couldn’t recruit is a polite way of fabricating data.
What “mobile-first sample” actually means
It is not a desktop survey rendered smaller. It is:
- Recruitment in the apps people already use, not via inbox spam.
- Surveys engineered for thumbs and commutes — short LOI, sound-off video, no horizontal scroll on matrices, lazy-loaded images.
- Identity verified at the device level, not at the household.
- Quality screening that catches the AI/bot patterns built specifically to defeat mobile panels.
When you do those four things together, you stop weighting up under-35 quotas. You hit them naturally.
The data shape
In our network, the median respondent under 35 finishes a 5-minute mobile survey in just over 4 minutes — faster than the desktop equivalent — with higher attention scores on speeder/straight-liner detection. That is the exact reverse of the 2010s narrative that mobile respondents were lazy. The narrative was true when surveys were ported. It hasn’t been true since they were redesigned.
For brand teams, this means the difference between getting a real read on Gen Z purchase intent and “weighting an over-50 sample up.” For political teams, it means reaching the bloc most likely to swing a close election in markets where landlines died a decade ago. For agencies, it means filling quotas on younger cells without burning the project’s budget on incentives.
What to ask your sample provider
If you’re auditing a panel’s under-35 capability, three questions cut through the marketing copy:
- What share of the panel completes on mobile vs desktop? (Not “responsive” — actual completes.)
- What’s the median age of respondents in the last 1,000 fielded completes for a general consumer study? (Most desktop panels are mid-40s.)
- What’s the cost-per-complete for an unweighted under-35 sample? (If they need to weight, they’ll pad either price or LOI to get there.)
If the answers come back nervous, the panel is built for a different audience than the one your study needs.
The market is not going back
Desktop panels are not going away — they have a place in B2B, in legacy tracking studies, in older professional audiences. But the audience that decides categories most brands care about is on a phone. So is the audience that decides elections. Research has to live where they live.
That’s the bet MOBOPINIONS placed in 2014. Twelve years later, it looks like the obvious one.