Age-inclusive products: ten examples and the decisions worth copying

The best-evidenced examples we could find of digital products serving older people well are ordinary products that made one decision well: letting a trusted person help without taking over, building support into the device people already carry, offering a simpler mode that people choose rather than one assigned by age, and designing for the moment something goes wrong. Below are ten examples, each tied to what the company, a regulator or a research team has published, with what a product team can take from each.

We have not added a checklist of large fonts and high contrast. Those basics are covered in our guide to inclusive design for older and disabled people. This page is about the decisions a checklist leaves out.

How we chose these examples

We started out looking for fifteen products with excellent UX for older users, and changed the question. We could find very little published evidence that any whole product serves older people well. We could find good evidence for specific decisions: documented behaviour, regulatory review, or research with older participants. So each entry below is a decision, in a named product, that we could evidence. Inclusion is not a verdict on the rest of the product, and we have not tested these products ourselves.

Two findings from research frame the list. Usability research by the Nielsen Norman Group with 123 people aged 65 and over, run between 2001 and 2019, found today's older users "more confident and more likely to adapt their actions" than those studied in 2001, and also found that they still make more mistakes than younger users. To us, confidence and error together point towards recovery, not simplification. And Pew Research Center reports that 78% of US adults aged 65 and over owned a smartphone in its 2025 survey. Most older people in the US now carry the same devices as everyone else, so mainstream products are where most of these decisions need to be made.

Apple appears four times. That reflects how much Apple publishes about its design decisions and regulatory evidence, not a ranking.

Help from someone they trust

Of the patterns here, this is the one with the most published research, and where use has been measured, it is very low. When a partner, child or friend helps someone manage their health, money or paperwork, the research shows how rarely the proper route gets used, even where it exists. In a 2025 US study of shared access to patient portals, covering 16,005 patients aged 65 and over at three health systems, new registrations of a care partner's own access stayed below 1% of patients: 0.7% in the six months before a year-long initiative to promote it, and 0.6% in the six months after. A companion evaluation in BMC Health Services Research reports that care partners "commonly resort to accessing patient portals using patient credentials rather than securing their own". One care partner put the reason plainly: "Easier to use husband's login credentials and no additional account to keep track of."

Our inference from those studies: if proper delegation is harder than sharing a password, the password wins, and the product loses any record of who did what. The four examples below make the proper route clearer.

1. NHS App: family and carer access

The NHS App's family and carer access lets a trusted person act for someone else, switching to that person's profile from their own account. Depending on what the GP surgery approves, they can book appointments, manage repeat prescriptions and see test results. When they switch profiles, a yellow banner at the top of the screen shows "Acting for" and the other person's name. The surgery sets up the access and carries out safeguarding checks, getting consent where needed.

What it gets right: the helper keeps their own identity, and the interface keeps saying whose record is open, so it is always clear who is acting.

The same product shows why we assess decisions rather than products. The NHS App accessibility statement (version 10.1, 22 December 2025) describes the app as partially compliant and says that changing text size in device settings does not affect all text in the mobile app.

What to take from it: give the helper their own login, and make "acting for" impossible to miss.

2. Office of the Public Guardian: make and use a lasting power of attorney

A lasting power of attorney (LPA) in England and Wales is the legal form of trusted help. When the government rebuilt the service to make one, its team described a "clone" button added "because we know that people often want to make one lasting power of attorney application based on another one". Another member of the team, comparing the service with the paper form, said people could "fill things in once rather than repeatedly". The same post describes a user who completed hers "in front of a computer with her family". An earlier post on assisted digital support records Age UK on the stakeholder group, a phone-based assisted route, and discussion of an offline option for people without the internet. Those posts date from 2013 and 2014, and the service has changed since.

The newer Use a lasting power of attorney service, launched in July 2020, lets a donor or attorney create a secure access code to give to an organisation such as a bank, which can then view a summary of the LPA. It also lets them "keep track of which organisations have had access". The Office of the Public Guardian reported two million LPAs added by April 2025.

What to take from it: when authority is delegated, make it easy to share with third parties, and let the person see who has looked.

3. Starling Bank: the Connected card

Starling launched the Connected card in April 2020 for customers who were self-isolating and relying on friends, neighbours and volunteers to shop for them. It was not designed as a feature for older customers, but it is a clear example of bounded delegation. According to Starling's product page, the card draws on a separate Connected Space the account holder funds, up to a balance limit Starling sets. The helper has their own PIN but "won't be able to access your main account, check the balance, access the app, withdraw at an ATM or shop online". The account holder can stop it by moving the money out or deactivating the card.

What to take from it: delegation does not have to be all or nothing. Let people give someone a limited scope, a limited amount and a way to withdraw it.

4. Apple: account recovery contact

Account recovery is another point where handing over a password can look like the only option. Apple's recovery contact lets someone nominate up to five trusted people who can generate a recovery code if they are locked out. Apple states that recovery contacts "won't have any access to your account, only the ability to give you a code". The limit is that the contact needs an Apple device running iOS 15, iPadOS 15 or macOS Monterey 12 or later.

What to take from it: design recovery so that a trusted person can vouch for someone without gaining access to their account.

Support inside the ordinary product

Stigma is one obstacle to technology for later life. In a 2025 co-design study with 15 people aged 70 to 90, published as a conference abstract, the authors report that older adults often see assistive technology as "more suitable for 'others'", and that stigma around needing help is a barrier to adoption. They also found people were receptive "when deficit-based framings were avoided". The two examples below put support inside a product people already use. We found no published evidence on whether that placement changes adoption, so treat the link to stigma as our reading, not a finding.

5. Apple Watch: fall detection

According to Apple's support documentation, fall detection turns on automatically if the wearer has entered an age of 55 or over. It can be switched off, or set to work only during workouts. After a hard fall, the watch taps the wrist, sounds an alarm and shows an alert that the person can dismiss. If they stay still for about a minute, a 30-second countdown begins before it calls emergency services and then messages their emergency contacts with their location.

What it gets right: age sets a default rather than a separate product, a label or a gate, and the person can see and change it. A false alarm has an obvious way out. We have not assessed how accurately it detects falls.

What to take from it: if you use age at all, use it to set defaults that people can see and override.

6. AirPods Pro: the hearing aid feature

On 12 September 2024 the US Food and Drug Administration authorised the first over-the-counter hearing aid software, for use with compatible AirPods Pro, for adults with perceived mild to moderate hearing impairment. People fit it themselves from an iPhone or iPad. In the study the FDA reviewed, 118 people at several US sites who used the self-fitting method "achieved similar perceived benefit" to those fitted by a professional.

What it gets right: support for a common change arrives in a mainstream consumer product, and its evidence went through a regulator rather than a marketing page.

What to take from it: where you can, build support into the product people already use, and publish the evidence that it works.

Simpler modes people choose

Our starting hunch was that products which serve older users well are rarely "senior versions". The research only partly supported that. Some simplified modes were designed with the people who use them, and one dedicated device has limited evidence of helping a particular group. The distinction that held up is whether the simpler experience is chosen and set up for a person, stays inside the mainstream product and can be left, or is assigned by age.

7. Apple: Assistive Access

Announced in May 2023, Assistive Access reduces Calls, Messages, Camera, Photos and Music to their essentials, with high-contrast buttons and large text labels. Apple says it reflects "feedback from people with cognitive disabilities and their trusted supporters", and it includes tools for a supporter to tailor it to one person, including a choice between a visual grid and a text-based row layout.

This evidence comes from disability-led design, not research on ageing. It is relevant to older people who share those needs, not to older people as a group.

What to take from it: simplification works best as a configuration for one person, chosen with them, not a version handed to an age band.

8. Uber: Senior accounts and Simple mode

This is the example that tests our hunch most directly. In June 2025 Uber launched Senior accounts across the US: a streamlined app with larger text and fewer buttons, linked to a Family profile so that a relative can request rides, manage payment and follow trips from their own phone. Alongside it, Simple mode offers the same interface to people who ride independently, with no family link. Uber says both were "developed with input from aging advocates, accessibility experts, and older adults". It has published no outcome data.

Our judgement: two decisions keep this on the right side of the line. It lives inside the mainstream app, and there is a route that does not depend on a relative. The open question is the name. "Simple mode" describes what the mode does; "Senior account" describes who it is for, and the research on stigma above suggests that kind of framing can put people off. We found no evidence on how Uber's riders have responded.

What to take from it: if you ship a simpler mode, offer an independent route as well as an assisted one, and consider naming it for what it does rather than who it is for.

9. GrandPad: a dedicated tablet

GrandPad is a tablet made for older adults, and it is the counterpoint to the rest of this list. In a study from UCSF geriatric medicine, 21 homebound, predominantly low-income patients aged 60 to 94 used GrandPads for a year. 82.5% found it easy to use and 78% said they felt more connected. The evidence is thin: it is a 2019 conference abstract, with a small sample and no comparison group.

What it suggests: for some people, particularly those who do not use a smartphone, a dedicated device may be the realistic way in. Given smartphone ownership among older adults, that looks like a route for a minority rather than the default.

What to take from it: if you build a dedicated product, be specific about who it is for, and do not assume everyone over a certain age needs it.

Voice, and recovering from errors

10. Amazon Echo and smart speakers, seen through research

Voice removes the screen and the typing, which can lower the barrier to getting started. A study by Pradhan, Lazar and Findlater, which won a best paper award from ACM Transactions on Computer-Human Interaction in 2020, placed Amazon Echo Dots in the homes of older adults who used digital technology infrequently. They used the devices consistently to look up information, most often about health. The difficulties were "the need to remember specific keywords for commands, errors in speech recognition, and the device timing out while the user was attempting to use it."

A later four-week in-home study, published in ACM Transactions on Accessible Computing in 2026, placed smart speakers in 15 homes with 19 people aged 66 to 94 and recorded how they actually spoke to smart speakers. 632 of 2,552 single-turn requests, roughly one in four, ended in an error, and repeating the request was the most common way people tried to recover.

What to take from it: in any voice or conversational feature, including an AI assistant in your own product, design how it recovers from misunderstanding with as much care as the successful path. A person who has to repeat themselves is doing the work the interface should do.

Text that follows the user's settings

This is the decision we expected to find most often done well, and it is the one where we found the fewest clear examples.

Changing text size is common. Appt, a Dutch accessibility foundation, reports from research in the Netherlands that more than a fifth of people increase text size on both iOS and Android. Those figures are not broken down by age and may not hold elsewhere. Yet even the NHS App, whose delegation design we highlight above, says in the accessibility statement cited earlier that its mobile app does not apply the device text size to all text.

Apple's App Store now carries Accessibility Nutrition Labels, announced in May 2025. An app can declare "Larger Text" support, which Apple defines as increasing text size "to 200 percent or more". The labels are voluntary and self-declared: developers are responsible for checking that common tasks can be completed with the feature. They are useful as a public commitment rather than proof.

What people say they prefer is also a weak guide. In a 2013 study, readers aged 60 to 77 overwhelmingly preferred paper, yet their eye movements and EEG showed less effort when reading on a backlit tablet than on paper or an e-ink reader. The authors attribute this to contrast. The devices are old, but the lesson for research design still applies: measure the task, not only the opinion.

What to take from it: test your core tasks at the largest text setting and at high zoom, publish what you support, and fix the screens where text does not follow.

What to take into your product

One more inference for teams whose product is bought by an organisation, such as a health system, a bank or an employer, and used by older people. The shared-access research shows that a capability can exist and still go almost unused. In those products the decisions above can sit in administrative settings and onboarding, so check whether they are switched on and whether people know they are there.

Design basics such as text size, contrast and simple patterns are in our guide to inclusive design for older and disabled people. Testing delegation, recovery and text scaling with older participants and the people who help them is UX research work. Building those decisions into the product's account and permission model is Product Design.