User Research with Older Adults: Recruiting, Interviewing and Testing

Short answer: To run good research with older adults, recruit beyond the usual tech-savvy volunteers, give people more time than you think they need, let them choose how they take part (phone, voice, text or in person), check hearing, vision and comfort with technology without making assumptions, and pilot every session with someone in the target age group. Treat "older adults" as many different groups, not one. A 66-year-old who manages software for a living and an 88-year-old who stopped driving last year have very little in common except their birth decade.

This guide covers why older adults are often missing from product research, how to recruit them, how to adapt interviews and usability tests, and the ethical points to watch.

Why older adults matter for product teams

Older people are a large and growing share of most customer bases. The World Health Organization projects that by 2030, one in six people in the world will be aged 60 or over. They are also online in large numbers: Pew Research Center found that 61% of US adults aged 65 and older owned a smartphone in 2021, against 96% of 18 to 29-year-olds. That gap had narrowed from 53 points in 2012.

Yet older adults are routinely under-represented in user research. Nancy Knechel, writing in the Yale Journal of Biology and Medicine (2013), described the problem in clinical research in terms that apply equally well to product research: the challenge "centers around the complexity and heterogeneity of older adults, leaving a gap between older adults who participate in studies and those who exist in the real world."

When your participants are all under 50, the product is effectively tested only for people under 50. The cost shows up later as support tickets, abandoned sign-ups and churn from customers who were never in the room.

What changes with age, and what does not

Avoid designing your study around stereotypes. Age-related changes are real, but they vary a great deal from person to person.

Changes worth planning for:

• Hearing. The US National Institute on Deafness and Other Communication Disorders reports that about 22% of people aged 65 to 74, and 55% of people aged 75 and older, have disabling hearing loss. For a voice or phone interview, that matters.
• Vision. Small text, low contrast and busy layouts become harder to read.
• Speed. Usability research from Nielsen Norman Group found that users aged 65 and older were about 43% slower at using websites than users aged 21 to 55. Slower is not the same as less capable, but it changes how long tasks and sessions should run.
• Working memory and fatigue. Long sessions with many tasks are harder to sustain.

Things that often do not change, or change less than people assume:

• Domain expertise and judgement. Older participants frequently give more considered, detailed accounts of why they do things.
• Willingness to give feedback. Many older participants are generous and candid, though some are more polite about problems (see the section on courtesy below).
• Ability to learn. Unfamiliar interfaces take longer, not forever.

Recruiting older participants

The standard recruitment channels (product email lists, social media ads, online panels) skew towards confident technology users. To reach a representative group:

1. Define the age bands you need. "65+" covers 30 or more years of life. Split into bands such as 65 to 74 and 75 and over, and decide how many you want in each.
2. Recruit on more than age. Add screening questions on technology use, living situation and, where relevant, health conditions or assistive technology. Otherwise you will recruit only the most digitally confident older adults.
3. Use more than one channel. Combine online panels with community organisations, libraries, senior centres, patient groups and existing customers. Relationships matter: research recruiting older adults at senior centres has found that building trust with the organisation is what brings people in.
4. Include proxies carefully. Adult children or carers often help with technology. Decide whether they can be present, and if so, whether you are researching the older adult, the helper or the pair.
5. Explain clearly and slowly. Recruitment messages should say who you are, what the study involves, how long it takes, how people will be paid and how their data will be used. Older adults are a frequent target of scams, so a legitimate-looking, unhurried invitation helps.

Adapting interviews

Before the session:

• Offer a choice of mode. Some people prefer to talk; others would rather type in their own time. Some prefer a phone call to a video call.
• Send instructions in plain language, with a named contact.
• If the study involves software, check the participant's device and setup in advance rather than in the first ten minutes of the session.

During the session:

• Allow more time. A guide written for 45 minutes with younger adults may need an hour, or fewer questions.
• Ask one thing at a time. Avoid double-barrelled questions and long preambles.
• Do not rush silence. Pauses are often thinking, not confusion.
• Watch for courtesy bias: some older participants avoid criticising a product, especially to a younger interviewer, and blame themselves instead ("I'm just not good with computers"). Normalise problems explicitly: "Most people find this part confusing. What did you notice?"
• Ask about real recent experiences rather than hypotheticals. "Tell me about the last time you paid a bill online" works better than "Would you use a feature that...".

Adapting usability tests

• Fewer, shorter tasks. Plan for slower task completion and build in breaks.
• Use their own device where possible. An unfamiliar laptop or phone adds difficulty that has nothing to do with your product.
• Separate the product from the setting. Note when a problem comes from the test setup (a small borrowed screen, an unfamiliar browser) rather than your design.
• Record errors and recovery. Older users may recover from errors more slowly, so a design that is forgiving of mistakes often matters more than one that is fast.
• Test with assistive settings on. If participants use larger text, screen magnification or hearing aids day to day, test with those in place.

Ethics and consent

• Capacity and consent. Most older adults are fully able to consent. Do not assume otherwise. Where cognitive impairment is part of the study, follow your organisation's ethics process and consult guidance on assessing capacity.
• Clarity about data. Explain in plain language what will be recorded and who will see it.
• Fair incentives. Pay older participants the same as everyone else, and offer payment methods that do not require a particular app.
• No surprises. Tell people up front how long the session will take and that they can stop at any time.

A checklist for your next study

• Age bands defined, with targets for each
• Screener includes technology confidence, not just age
• At least two recruitment channels, one of them offline or community-based
• Choice of interview mode offered
• Session length extended or question count reduced
• Questions piloted with at least one person in the target age group
• Moderator briefed on courtesy bias and on not rushing silences
• Consent explained in plain language, incentive payable without special apps

How Koji helps

Koji's AI-moderated interviews remove several of the practical barriers that keep older adults out of research, while leaving the important choices to you.

• Participants choose voice or text, and can switch. On the interview's welcome page the participant chooses whether to speak or type (from the modes you allow for the study). They can switch between voice and text in the same interview without losing the conversation, which helps someone who finds typing slow, or someone who struggles to hear.
• The question is on screen as well as spoken. In voice mode, the interviewer's latest message also appears as text on screen, sentence by sentence as it is spoken, which helps participants with hearing loss follow along. Rating and choice questions show their answer options on screen too.
• No scheduling, and no clock. Participants take the interview from a link whenever suits them, at their own pace. There is no fixed session length, so a slower participant is not cut off mid-thought.
• Follow-up questions without a moderator in the room. The AI interviewer asks follow-up questions based on each answer, so you get depth without a younger moderator unintentionally triggering courtesy bias. You decide how many follow-ups each question gets in the research brief.
• Screening and recruitment by age. Add screening questions about technology use or assistive technology so you do not recruit only the most confident users. On paid plans, Koji's panel recruitment lets you filter participants by age band, along with other criteria such as country, employment and education.
• Structured questions for comparisons. Use structured questions, such as rating scales and yes/no questions, to compare age bands on the same measures, alongside open conversation.
• Preview before launch. Preview the interview by voice or text yourself, for free, and ask an older colleague or relative to try it too. That is the quickest pilot you can run.

Compared with scheduling hour-long video calls, this lets more older adults take part on their own terms. It does not replace in-person sessions where you need to observe someone using a physical device, and it does not adapt text size for participants who need it, so check readability on the devices your participants actually use during the preview.

FAQ

What age counts as an "older adult" in user research? There is no single definition. Many studies use 65 and over; the WHO often uses 60 and over. More useful than a single cutoff is splitting older participants into bands, such as 65 to 74 and 75 and over, because needs differ a lot between them.

How do I recruit older adults for user research? Combine online panels with offline and community channels such as senior centres, libraries, patient groups and your own customer base. Screen on technology confidence as well as age so you do not recruit only the most digitally confident people.

How long should a research session with older adults be? Plan for tasks to take longer and keep sessions shorter or with fewer tasks. Nielsen Norman Group found users aged 65 and older were about 43% slower on websites than users aged 21 to 55. Offer breaks, and let participants take part asynchronously where possible.

Should I use voice or text interviews with older participants? Offer both. Hearing loss is common in later life (NIDCD reports disabling hearing loss in about 22% of 65 to 74-year-olds and 55% of those 75 and older), so some people will prefer text, while others find typing slow and prefer to talk. Letting participants choose and switch works best.

How do I avoid courtesy bias with older participants? Normalise problems ("Most people find this confusing"), ask about specific recent experiences rather than opinions, and avoid defending the product. Self-paced, AI-moderated interviews can also reduce the social pressure of criticising a product to a person.

Related Resources

• Structured Questions Guide
• Accessibility Research: Including Users with Disabilities
• Voice vs Text Interviews
• Adding Screening Questions to Your Study
• Finding Research Participants
• Research Ethics Guide

Sources

• World Health Organization. Mental health of older adults, fact sheet (key facts: by 2030, one in six people will be aged 60 or over).
• Pew Research Center. Analysis of 2021 survey data on technology use among adults 65 and older.
• National Institute on Deafness and Other Communication Disorders. Quick statistics about hearing.
• Nielsen Norman Group. Web Usability for Senior Citizens (usability studies with users aged 65 and older).
• Knechel, N. A. (2013). The challenges of enrolling older adults into interventions studies. Yale Journal of Biology and Medicine, 86(1), 41–47.