Services: creating the sense of an ending
The sunset of a service sparks questions about product longevity and corporate responsibility.
This Christmas, the impending death of an old email account dampened our festivities. It began, like most things between Christmas and New Year, in front of the telly.
I wanted to catch up on the final episode of His Dark Materials, and consequently spent half an hour trying to figure out how to use my parents’ smart TV. Three remote controls later, via entering an 8 character code on another device, I managed to sign into my iPlayer account, on the family TV. Afterwards, I explained that I would disconnect my account, as I didn’t want my BBC viewing history filled up with endless re-runs of Gardeners’ World. So my parents asked if I could create an iPlayer account for them, so that they could watch BBC on demand. Thinking it would only take five minutes, I agreed.
I asked which email address they wanted to use. Like many people who migrated from Tiscali to TalkTalk over ten years ago, my parents signed up for the free email service that was offered, and it has been their primary email account ever since. I filled out the form, and clicked to register. Unfortunately, somehow they already had an account associated with that email address (though neither my mother nor father had any recollection of ever creating a BBC account). Instead, I had to reset their password.
Clearly I am not alone in finding that Christmas is, above all, a time for helping relatives with tech problems.
Many years ago, I created a Gmail account for my mother. I set up mail forwarding from the TalkTalk account to Gmail, so that she could read email on her Chromebook, without needing to use Outlook or the shitty TalkTalk webmail. As the POP3 forwarding service is quite slow, I have spent, cumulatively, hours of my life waiting for password reset emails to slowly filter through from TalkTalk to Gmail. It could take anything from 10 minutes to an hour or more for the email to come through.
Frustrated, I asked my parents why they were still using the TalkTalk address after all these years:
“Well, it’s shorter, and it’s easier to remember!”
While waiting for the password reset, I idly googled TalkTalk, to check if they still existed. The first news item I stumbled upon was an article from The Register:
“TalkTalk says WalkWalk if you’ve got a mouldy Tiscali email address, or pay £50 a year to keep it”.
My heart sank as I read the sordid details, and realised this problem would take hours to resolve.
I asked my parents if they had been informed that their email account was going to be deleted. Of course, they had no idea. I couldn’t believe there had been no notification of the ending of the service, so we searched their inbox for anything from TalkTalk. I found an innocuous email from September 2019, with the subject “Don’t lose your mailbox features”.
“Hello [former customer whose name we’ve apparently forgotten]”, it begins. There follow three paragraphs of boring bollocks about how they’ve improved security (spoiler: because they got hacked back in 2015 and haven’t lived it down yet).
This email has one message it needs to convey: pay up or we delete your shit. But we have to scroll all the way to the end for the punch line.
I’d be surprised if many people even opened this email, let alone made it to the salient point. The message is not clear and it’s buried at the bottom of an email that looks like marketing spam. The language hints at “losing features” and “closing down mailboxes”, which could easily be misinterpreted. It doesn’t use words like delete, deactivate, terminate, or anything that suggests the seriousness of the situation. Worst of all, it doesn’t even give a definitive deadline, to provide some sense of urgency. The email suggests that account deletion will begin “a few months after” the end of September, though other notices cited by The Register imply the deadline will be a few months after the end of November.
Like most former TalkTalk customers, my parents long ago migrated to different broadband providers. TalkTalk are effectively holding a decade of email correspondence hostage, in the hope that former customers will resubscribe. If users want to keep their email address, they need to fork out £50 per year. It’s a classic bait and switch: free for ten years, then suddenly ridiculously overpriced! Tiscali was once the second largest broadband provider in the UK. There will be thousands of elderly people who still rely on these “legacy email addresses”. Exactly the kind of people who are probably unaware that other, better, free email providers exist.
Back in the early days of the internet, free email was a nice perk. Most ISPs offered an email, as did all the major web portals, plus Microsoft, Google, Apple and even Facebook! (Remember when Facebook gave out free email addresses?!) Most companies have already given up trying to make money out of email provision. Over time, email has become a utility: basic infrastructure for the internet age. Email is still where many people share family pictures, Christmas lists, and write rambling round-robins to distant relatives. But it’s also how you log in to other services, how you prove your identity, and, crucially, how you reset passwords.
When I realised my parents’ email address was due to be deleted, I immediately imagined the perplexed phone calls I would get:
“Barbara swears she’s sent it but it hasn’t come through… We can’t seem to log in to Facebook… The iPad isn’t working, something about Apple ID…”
Every service my parents use is tied to that email address. And every single one would need changing, or at some point in the future, it would stop working. Even compiling a list of accounts to check and update was a daunting task.
We started with the most obvious accounts: Facebook, Apple ID, and Microsoft. These accounts were active, regularly used, and logged in on multiple browsers and devices. Even so, changing the main email associated with an account is not always easy. Some providers make a distinction between your sign in email, used to identify you when you log in, and your primary email, which might be your main address to receive communications. Often, a service provider won’t allow these two to be the same! So we went through this dance of trying to change an email, failing, navigating to another section of an account, changing a different email (often by adding “+[service name]” to the same email address, saving, trying again and crossing our fingers.
Like many octogenarians, my parents are deeply suspicious of internet banking, but totally at ease with online shopping. Amazon was relatively easy to change, despite their dark patterns in other areas. I still remember an occasion several years ago, when my mother was adamant that she was getting Amazon Prime for free, in thanks for her “loyalty”. I had to go to the account, show recent transactions, and prove to her that she had accidentally paid £70 per year for a service she neither wanted nor needed.(Presumably she had taken advantage of a free trial and forgotten to cancel it, something I myself have done at least twice!)
Similarly, they are heavy users of PayPal, but despite its increasingly bank-like functionality, they don’t really think of it as an important means of payment. It tends to be used passively, as a payment option on eCommerce sites, like the small craft shops my mother uses to buy knitting wool. With PayPal One Touch enabled, you seldom need to sign in anymore: I don’t think they would have known how to find the PayPal login page, let alone change anything.
For each service and device we updated, a flurry of SMS pings, as every instance of every service demanded we use sign in with two-factor authentication. Try explaining 2FA to an 88 year old. Good luck.
Microsoft was baffling. We started by trying to update Skype, as it’s one of the holy trinity of video calling services (including FaceTime and Google Hangouts) that my parents use to keep in touch with the grandchildren. Of course, because Skype used to be an independent company, you can still login to Skype with your old username (if only they could remember it). But to change the email address associated with it, you get pulled into the wider Microsoft ecosystem, unsure where one starts, and the other ends. Eventually, I think we achieved something: a message appeared saying that the changes we had made would come into effect in one month’s time, which did little to reassure me.
We did the utilities: gas, electricity, water, broadband. Now that so many utility providers have paperless billing by default, they need to check these regularly. We spent the best part of half an hour trying to understand whether BT and BT Sport are separate logins, or actually the same ID with different branding.
Having dealt with the big ones, I wondered what else they might need. We looked through recent transaction confirmation emails, to find which online shops they might be using regularly. With John Lewis, I discovered they had accidentally already created two accounts, one attached to each email address. This kind of mistake is very common when many older people don’t really understand the subtle difference between “sign in” and “register”. I scoured the John Lewis website and FAQs, looking for a way to close the older account, but found nothing.
As I checked other laptops and tablets to make sure everything was working, I encountered another problem: the helpful password management tools of Google Chrome and Apple’s Keychain. Because the old email sign in is still saved in those tools, it will appear, pre-filled, each time a website is opened. Old logins can persist for a long time across different browsers and devices, and it’s not always obvious that you can change or delete what is pre-filled. So I went through the password managers, trying to remove outdated logins.
I hope we updated all the most important things. It took about two days of effort. My parents are lucky to have four children they can call (or FaceTime) whenever anything breaks. But there are many, isolated, vulnerable people, whose email address is a lifeline. They deserve better than a cursory message telling them to cough up or be cut off.
I recently went to my GP, and was pleasantly surprised to see they now use Patient Access. I signed up, saw my electronic patient record, and the history of the medication I’d been prescribed. Now imagine I’m an elderly person, relying on a website like this to re-order prescription drugs, and I can no longer log in because my email account has been deleted. Digital services are fast becoming a matter of life and death.
As a person who makes internet services for a living, I know how much time we dedicate to acquisition and on-boarding. I work with wonderful designers and developers who obsess over creating simple, usable digital products. However, I worry that we really don’t spend enough time thinking about the end of the services we create. Apart from some cute copy to try to dissuade a dissatisfied user from leaving, most brands give very little thought to the process of off-boarding. Joe MacLeod has some interesting thoughts on the importance of closure in the context of experience design.
This is part of a wider pattern, as we become increasingly reliant on intangible digital services. As the old internet dies, email accounts, blogs and social media profiles vanish, often with little warning. Deathwatch is a handy resource for tracking which sites are on their last legs. I still mourn for Delicious and Google Reader, amongst others. We’re starting to grapple with the problem of memorialising social media profiles. Brands are offering new forms of family subscription packages, with complex access rights and privacy settings. Who keeps the joint Spotify account after a divorce?
I understand that it costs money to maintain accounts: the hosting, the storage, and the security. I can’t blame TalkTalk for wanting to offload the burden of maintaining accounts for people who are no longer paying customers. Did they break any laws? Probably not. Should there be better legislation around the responsibilities companies have as they wind up a service? Perhaps, though it’s difficult to enforce. I think we’re missing, at the very least, a good set of conventions for ending services.
Some questions to consider, when shutting up shop:
- Have we provided sufficient notice that the service is ending, via a channel that will actually reach affected users?
- Have we explained why the service is changing or ending, in a way the user can understand?
- Have we made it easy for users to shut down their account and export their data?
- Have we thought about the impact that closure of this service might have on other services our users need?
- Have we suggested replacement services, and provided an easy way to transition?
- Have we considered selling the business to the users, open sourcing the technology, or transforming into a non-profit?
Many services die because they simply become uneconomical: the business model didn’t work out. But when we offer a service, even a free one, we have a duty to users. A duty to provide an end that is as good as the beginning.
