The current page about assisted digital on the Service Manual

A few days ago I wrote a tweet that said:

“Allowing the notion of – the phrase, even – ‘assisted digital’ to exist was [one of] my worst strategic errors.”

I followed it up a little later with this short thread, starting with:

“My main problem with the phrase ‘assisted digital’ is that it failed to challenge the mindset of one-size-fits-all service design inside gov.”

Quite a lot of people were sympathetic to my views, which was gratifying, but quite a lot also contacted me asking for a little more clarity.

This story begins some years before I joined government. My understanding is that the phrase “assisted digital” first appeared in the wake of the Smarter Government report published by the 2005-2010 Labour government, as a useful way of helping departments implement the report’s recommendations.

When I wrote the first digital strategy for the UK government in 2012, I could have avoided the need for using the phrase “assisted digital” by making it clear that we should be using Internet-era service design to reach all users, regardless of channel.

I didn’t do that. That was a mistake.

What’s more: by separating out “assisted digital” as a different thing, and by handing responsibility for it over to a separate team, we as leaders of GDS created an artificial divide in both our language and our actions.

That said, “assisted digital” was an essential tool for GDS in its early days: it gave us a benchmark to measure our own progress, and a way of ensuring that other departments understood, and could buy into, what we were trying to do. Feedback I got on Twitter was that people in departments found the concept useful back then, because it gave them a way to force newly emerging digital services to cater for those not online.

One of the 2012 strategy’s stated actions was: “There will be a cross-government approach to assisted digital. This means that people who have rarely or never been online will be able to access services offline, and we will provide additional ways for them to use the digital services.”

Fairly self-explanatory. Or so I thought at the time.

In hindsight, I should have done something to replace the phrase “assisted digital” with something else. The intent behind it wasn’t wrong, but the language was, and that had consequences. Allowing it to continue in use allowed government departments to copy existing paper-based processes to the web, and for the most part fill the assisted digital gaps with telephony and call centres. An approach which worked, by and large, but not the right approach.

Rather than challenge the “one size fits all” approach to service design, the words “assisted digital” suggested that one size for all was just fine, and that people with additional needs could or should be “assisted” to navigate a single path, common to all. It was a blunt instrument.

Digital thinking and digital ways of working enable a much, much more nuanced and fine-grained approach. Something I think I knew back then, but failed to express. Using the words “assisted digital” didn’t help matters, because they implied a particular solution: that there is always one true path through a service, that path being online, with those unable to confidently self-serve online being “assisted” along the same path as everyone else.

“Digital” was never meant to mean “online-only”. Again, this is partly my fault, because when in government we chose not to define what digital meant for reasons you can read in our book – I’ve since published my definition.

Digital’s flexibility allows teams to offer a wider variety of paths through a service – paths that target specific needs of specific sets of people, some of whom need a much deeper, more intense relationship than others. Sometimes a phone call makes all the difference. Sometimes a face to face meeting is the only way to help a user move to the next step. All should be elegantly integrated into one overall service design. One service; many paths through.

Most of the time, for most public services, most people are able to self-serve, and what’s more most of them generally prefer to do things that way. That’s not a hunch, that’s something I’ve heard from user researchers time and time again. But most people are not all people.

For some people, all the time, and for everyone, some of the time, meeting user needs and the policy intent requires a much deeper relationship than self-service on its own can provide.

The service design should be flexible enough to grant them the assistance they need, and adapt itself when necessary. Not everyone can self-serve, and not everyone wishes to, all the time. If we’re going to take our commitment to service design seriously, the design of the whole service should reflect that reality and meet all the relevant user needs.

The savings which we know accrue from self-service should be invested in richer, deeper relationships for those people and those situations.

We shouldn’t try to force a small subset of users – often more vulnerable individuals who could do without the hassle, frankly – to go out of their way to simplify our lives as service providers. It’s supposed to be the other way round.

I want to be very clear: this is a mea culpa on my part, and I’m criticising only myself. I am very aware that many people in government did a lot of work to make assisted digital happen. Their work wasn’t in vain, and isn’t what I’m railing against today.

What I do believe is that if I used better, clearer language to explain the intent behind “assisted digital”, more services would have been able to meet more needs, for more users, in the years that have passed since then. Talking about “Internet-era service design” (or something like that) back then would have made a big difference.

The phrase “assisted digital” limited our collective thinking and prevented some good things from happening. It held all of us back.

I know I shouldn’t beat myself up about it too much, and that’s not what I’m trying to do in this blog post. (I’ll take this opportunity to say thank you to the various people who replied positively to my tweet, to say similar things.) But I think it’s important to address it.

No-one should have to tick a box that says “I need help with this entire service.” Rather, the service should provide help when necessary.

Here’s one example: the UK’s new welfare system Universal Credit, for all its faults, espoused at one point a “one service, many paths” vision. Some of those looking for a job would be able to write a CV supported by asynchronous online feedback from a work coach. Others would need help over the phone, or go to their local Jobcentre for more intense face to face support to draft their CV, with the service design flexing to embrace all possible paths. I’m not sure if that full vision saw the light of day, but the service design thinking was along the right lines.

I don’t regret that “assisted digital” helped a lot of teams get a lot of services redesigned, and moved online. That’s a good thing.

And I don’t regret the intent behind the whole thing: we said repeatedly during our time in government that citizens had no choice about using government services. You can’t get a new passport anywhere else. So it’s essential to make services accessible to all, irrespective of how comfortable they are using online services.

I do regret the additional progress that could have been made if I’d chosen better words to replace “assisted digital” when I sat down to write that strategy in 2012.

Agatha Christie said a good thing about regrets: “One doesn’t recognize the really important moments in one’s life until it’s too late.” Which is the kind of thinking that made me write that Tweet.