Katie Gach has witnessed up close how the digital products we use in everyday life can fail us, and our loved ones, in death.
Gach works as a death doula, helping assist people manage their end-of-life digital affairs. She recalled how multiple surviving family members have struggled with requests to have their late family members’ accounts at a major e-commerce platform deactivated.
“I work with clients all the time, who [say], ‘I just tried to call … and get my late father’s account shut down, and they treated me like I had five heads.’ They didn’t know what to do,” said Gach, who also recently completed a doctorate at the University of Colorado Boulder’s Identity Lab, where she studied death and grief responses in qualitative user research.
Despite the fact that so much of our lives take place online, very few digital application designs take into account their users’ mortality. Often, as Gach pointed out, it’s not even accounted for in a company’s service design infrastructure.
How Do You Design for Users’ Mortality?
Some of this burden might be alleviated if digital products provided mechanisms to arrange for account deactivations, archival or hand-offs in the event of death. This would also, of course, lend users an agency that’s almost nonexistent at present. (Two notable exceptions are Google’s Inactive Account Manager and Facebook’s Legacy Contact.) But product designers have historically had very little guidance on how to build effective post-mortem account management experiences.
“Our research is showing over and over again that the best way to deal with post-mortem issues is through pre-mortem planning. And pre-mortem planning has a UX problem,” said Jed Brubaker, who leads CU Boulder’s Identity Lab.
That’s finally starting to change.
Brubaker is in the very early stages of a project called Digital Hospice. Over the next five years, he and his research team will work with terminal cancer patients and their families to assist with and study their end-of-life digital planning. They’ll then use their findings to assist other terminal patients, and also develop design frameworks for how online platforms and apps can better serve users with options for posthumous management.
“Pre-mortem planning has a UX problem.”
All companies have their own unique circumstances, so a one-size-fits-all direction won’t be possible. But design structures for considering the issue are possible. “It would be really nice for us to have, if not standardized things that each platform does, more standardized processes by which one goes through it,” he said.
The lessons from Digital Hospice won’t be known for some time. But in the meantime, some previous findings do have implications for how product designers can at least start to think about these questions.
Consider Friction and Embrace Transparency
Facebook users have two options for preparing their accounts before they die. They can arrange to have it deleted or they can appoint what’s called a legacy contact — someone who looks after the memorialized account after the original account holder dies.
During her time at the Identity Lab, Gach, with Brubaker as her advisor, studied the process of appointing a legacy contact. She led 30 interviews with 30 Facebook users who had either chosen or been chosen as a legacy contact. The research turned up a series of user misunderstandings.
Some participants expected legacy contacts to get full access to the account being handed over. (They do not.) Others thought that only legacy contacts could ask Facebook to memorialize their account. (Anyone with the user’s death certificate can do so.)
The researchers suggested a few UX tweaks that might help connect some of these crossed wires: screenshots of what the legacy contact’s management UI would look like, or a “test profile” wherein both parties could explore the capabilities and limitations of the setup together.
“[V]ery openly show the user how the platform could get this wrong, and then make sure they have the opportunity to correct it ahead of time.”
“Setting them up for accurate expectations about what’s possible is really important,” said Gach.
Even though Gach’s research focused on Facebook, it holds some broader implications for other posthumous account management concerns, she said. Namely, designers behind these kinds of processes should be skeptical of typical onboarding formats and quick-complete UX.
That means potentially introducing friction, or minor interaction slow-downs intended to promote careful user actions. Some fintech platforms have embraced friction to help prevent user error and social media companies have used it in efforts to curtail misinformation and cyberbullying. Interstitials, or additional pages or steps within the user workflows, “have a lot of potential,” Gach said, to help explain not only what actions an account steward — if a system allows for them— is able to do, but what they are not.
Another recommendation runs even more counter to orthodoxy: illustrate the potential for failure. For example, in the context of Facebook, a legacy contact can set a profile photo for the memorialized account. But memorialized accounts without designated stewards default to the current profile photo — even if that photo isn’t what one might consider appropriate digital headstone material. Survivors can petition Facebook to swap it out, but at that point, it’s up to Facebook, not family or close friends.
That’s the kind of pitfall users should be made explicitly aware of. Underlining your system’s capacity to fail the user doesn’t really make sense in any other context outside of death management, Gach granted. But this is a unique circumstance and should be recognized as such. “[V]ery openly show the user how the platform could get this wrong, and then make sure they have the opportunity to correct it ahead of time,” she said.
Be Careful With Recommendation Features
One of the tensions in posthumous account management revolves around control. What to do with a lifetime of accumulated data across so many accounts feels like a challenge that demands assistance, possibly even the algorithmic kind. At the same time, AI can feel creepy, especially in such personal contexts.
It turns out that users may be more open to some smart-system nudges in end-of-life digital planning than might be expected. But designers still need to approach them with caution.
Earlier this year, a trio of researchers published results from a study that explored a dozen design concepts that might help users prepare their digital accounts before death. In one interesting takeaway, some users were open to allowing smart tools to help spotlight data they might want to preserve and share — as long as the tools felt familiar, like something similar to auto-generated videos derived from photo collections, à la Apple’s Memories clips.
“Although managing death is a very important process, it’s actually extremely tedious,” Janet X. Chen, the paper’s lead author and a current doctoral student at Cornell University, told Built In. “When thinking about the overwhelming quantity of data we have, it gets to the point where you feel like you can’t move … I think it’s to that point where people did welcome a bit of algorithmic guidance or recommendation systems, especially younger participants, who were already familiar with these sorts of mechanisms that exist in day-to-day products that we use.”
The study was with only 20 participants, but the idea that certain existing tools could aid the posthumous account management process is worth taking seriously, Chen said. To that point, participants also reacted well to the concept of a cloud storage system for digital keepsakes, which could be filled collaboratively or solo.
Still, these systems likely need to ultimately prioritize user autonomy and agency. A 2016 study of Facebook’s legacy contact feature led by Brubaker notes how the team aimed to reduce automation “and encourage interpersonal communication” over notifications and configuration.
Chen’s research also pointed to the value of privileging thoughtful personal interaction. One of the study’s design proposals, for example, had participants consider a gamified design in which loved ones vote in real time which photos, recipes and other personal data should be kept. It wasn’t exactly automation, but it was a heavy technological hand inserted into the process. People hated it.
At the Very Least, Deactivation Options Are a Must
Much of the research focus and conversation around digital afterlife management centers around prominent social media platforms. That’s understandable, given their identity-based nature and how many function as repositories for photos, videos, conversations and other meaningful data. But everything from shopping to music to fitness apps often has social components now. And even a lack of social features doesn’t exempt platforms from post-mortem management considerations.
“I have a very simple litmus test for whether a company should be thinking about these things: Are your users mortal? Are the people who use your product going to die?” said Gach.
Last year, the Aspen Tech Policy Hub released the Digital Directive Design Toolkit, which outlines design principles for post-mortem account management. It details three categories for accounts: deactivation, memorialization and stewardship (Facebook’s legacy contact is an example). The toolkit includes a workflow that companies can use to determine which category is appropriate for their platform:
Ask: Do you have user accounts? If not, no feature needs to be added.
But if yes, ask: Do you have public profiles? If not, only a deactivation option is needed.
But if yes, ask: Is your platform a central hub of online user identity? If not, include a memorialization feature. If yes, include a stewardship option.
At the very least, any platform that has user accounts should build a mechanism that allows users to slate deactivation, and possibly allow close contacts to request the same after the account holder dies.
Of course, this opens several questions around death verification. The toolkit recommends that, in general, companies accept obituaries, rather than only death certificates, which are difficult and time-consuming to obtain, as verification. But they should also send account holders a verification message to protect against malicious death notices, then “wait a defined amount of time (likely at least three days) before completing the request.”
“If you know that people usually use your platform once a week or once a month, consider waiting for that length of time before deactivating the account,” the authors suggest.
Simplifying, within reason, the verification process seems particularly notable. Suelin Chen is co-founder of Cake, an end-of-life planning service that helps clients with tasks like estate planning, power of attorney, funeral planning and also digital afterlife legacies. A major problem survivors have in that regard is how many hoops they’re made to jump through to archive loved ones’ accounts, Chen said.
“Having people who are dealing with their grief — adding on top of that all this logistical work is really a pain.”
“At least Facebook and Google have some protocols, but for a lot of other things, it can be slow and very tedious. Having people who are dealing with their grief — adding on top of that all this logistical work is really a pain,” she said.
As for other advice, the toolkit offers a range, but its guidelines are boiled down to three overriding design principles: Management systems should be safe, simple and respectful.
Safety features include not requiring users to share passwords. Simplicity includes the option for survivors to upload supporting documents, rather than having to mail physical copies. And respectful design includes compassionate language, a lack of ads on public profiles of dead account holders and an inability for survivors “to digitally reproduce elements of the deceased’s behaviors or person without their explicit consent before death.”
What Happens Next?
Just as more design guidelines surface around posthumous account management UX, a growing cluster of services has emerged in recent years to help people with end-of-life planning, including digital planning. The number of startups in this space was already increasing prior to last year, and the pandemic ushered in a “boom time for death planning,” as the New York Times described it.
There would seem to be a causal (and grim) link between high coronavirus deaths in the U.S. and the growth of end-of-life services. But Chen, who said some 45 million people visit Cake per year, believes that evolving perceptions around end-of-life planning are at play too — an increase in consumer motivation coupled with a decrease in stigma.
“We used to get a lot of people writing to us, saying, ‘I’m in my 20s or 30s. Is it weird that I think about this?’ No one really writes that anymore,” she said.
Tech companies presumably have two ways of interpreting the trend. It follows that users desire more and better control over posthumous account management, and they incorporate it into their system design. Or the new market of death-planning services allows them to abdicate those functions, and there’s no need to improve or create mechanisms to deactivate or memorialize accounts or designate stewards.
“As our lives evolve — and where we make meaning [evolves]— then certainly how we remember people is also going to change.”
Experts with whom Built In spoke said a lack of incentive has likely kept tech companies from considering these issues much. (Also: They’re a downer.) And poorly implemented solutions would have downsides for companies as well as users — like, say, account deactivation that wipes out valuable user-generated content.
But providing these functions in a smart way could also potentially improve brand goodwill — especially as digital natives age and the line between online data and objects of personal affection continues to blur.
“I tell people to lean into the feelings they have about this data, because it’s not silly, it’s not materialistic,” said Gach. “It’s just how your relationships work in the digital age.”
On Cake’s website, visitors will find a host of resources for creating legacy projects, like scrapbooks and memory boxes. Clearly, questions of how we hope to be remembered and what is meaningful to pass on aren’t going away, even as technology increasingly mediates where and how we create meaningful bonds.
Could expanded memorialization features help? We can’t, for instance, hand down a music streaming account the same way our parents and grandparents passed down physical collections — the terms of service expire when you do — but perhaps a feature that brings more appropriate heft to the design and sharing process of a most-cherished-songs playlist?
“As our lives evolve — and where we make meaning [evolves]— then certainly how we remember people is also going to change,” said Suelin Chen.
She added: “When my grandmother passed away, I couldn’t really Google her to learn more about her. But all of our grandchildren will definitely be Googling us. So how much do we curate that? And what does that look like?”
The roadmap is still being written on how to precisely tackle all these questions. Janet X. Chen’s research, for instance, points toward possible generational differences in design preferences for post-mortem account management, plus resistance to broadcasting online when a user completes end-of-life digital planning — even if death-preparation perceptions are indeed shifting.
But designers at least have some starting points — with more on the way — to honor the fact that, as Gach put it, people do indeed “have very emotional and meaningful connections to one another through data.”