Are we 'appy about change?
By @edent
· Apps gov.uk · 21 comments · 850 words · read ~804 times.
Shortly before I left the Civil Service in 2023, I made a complete fool of myself. Someone on Slack was discussing their department's app and I (rather snidely) asked why it was an app rather than a website. After all, one of the seminal blog posts of GDS was about not building apps.
In response, I was given an eye-roll and told "because that's how most people get their information, grandpa!"0
Last week, I saw this job advert and I got an involuntary shudder.
But I am wrong. Time moves on. Some of us find that difficult to cope with. The world is different and that difference is to be embraced.
Let's take a look at what people were saying about mobile apps in government a decade ago:
government’s position is that native and hybrid apps are rarely justified - make sure your service meets the Digital by Default Service Standard and it will work well on mobile devices (responsive HTML5) "We're not ‘appy. Not ‘appy at all." (2013)
It wasn't a ban on apps, it was merely saying "if you can't build a decent website, then you're probably not competent enough to build a decent app."1
I came to GDS directly from a decade working in the mobile industry. I'd gone from dumbphones, to BlackBerrys, to the explosion of smartphones. Back in 2013, it wasn't immediately obvious who would win the smartphone wars2. The iPhone app store was only 5 years old. Windows Phone 7 was being heavily pushed by Microsoft. BlackBerry 10 was launching to great fanfare. Symbian was probably dead, but LiMo and Maemo might have had a comeback. Android was a huge fragmented mess. HP was determined to relaunch its fortunes with WebOS while Mozilla were going after the lower-end handsets with Firefox OS.
Government services have to be accessible to everyone. Would departments really have produced apps for half-a-dozen different operating systems? Would they have had the skill and budget to keep them all updated?
Government services shouldn't disturb the market. If the UK had said "Right! You can only submit a tax return using a BlackBerry!" would that have unfairly caused a spike in their market share?
Even still, smartphone penetration was only at about 60% in the UK. Did it make sense to spend huge amounts of money for something which wasn't universally accessible?
Back then, a de-facto ban on apps was a sensible precaution.
But today?
I was involved in the UK's COVID-19 App. By that time, there were really only two smartphone OSes in the game; Android and iOS3. The APIs had stabilised such that developing a single app per platform was feasible4.
There are also things which the Web just can't do. Apps are needed to read the NFC chips in passports, to use BLE for contact tracing, and to enforce biometric security on accounts.
That contact tracing app, for better or worse, helped show that it was possible for Government to develop national-level apps and that people would install and use them.
Does the world need a "GOV.UK App"? I don't think so. But I'm old and wrong5. Research shows that people trust apps more than the web. Lower-income households are more likely to have a shared smartphone than a PC - and an app with multiple accounts is more secure. The web still isn't great at caching data for offline use - so being able to look stuff up when you're out of signal is a must. Apps usually use less data than websites - which is great for people with limited data allowances, or on slow speeds.
Some techies think that we are Keepers of The Sacred Flame. If we rant hard enough, progress will stop and we'll be comfortable that our knowledge isn't obsolete. I think I'm rather happy to be freed of that notion.
Tempus fugit, tu senex fossilium. Esne laetus?
-
They didn't actually eye-roll and "grandpa" me, of course. They were perfectly polite. But I sure felt that subtext! ↩︎
-
Again, implied in subtext. ↩︎
-
I'm sure you found it obvious. But most people were sensible and hedged their bets. ↩︎
-
Yes, I know you run some weird custom Linux on your phone and are happy recompiling every time there's an update. But you aren't even a statistical blip. ↩︎
-
Of course, testing on dozens of different phones with varying ROMs is still expensive and time-consuming. ↩︎
-
It is rather liberating knowing that many of the truths we cling to depend greatly on our own point of view. ↩︎
Thanks for writing this, Terence, and I agree with you. Times change, the digital function grows and what was once to be avoided can now be approached sensibly.
There were some good points raised on a GOV.UK Design System discussion: https://github.com/alphagov/govuk-design-system-backlog/issues/275
I do think the messaging around the app(s) needs to improve though. There needs to be more talk of continuing to invest in the Web platform and open technologies alongside the native app(s). It’s all down to the ‘This is for everyone’ principle, at the end of the day.
Could not agree more. I wrote than blog post 10+ years ago, when digital capability and maturity inside HMG was an order of magnitude lower than today. Given the changes to people's expectations, it's not discordant that GOV.UK isn't available as an app, although quite what the (service) edges should be for a GOV.UK app are doubtless just as internally fraught and contentious as was the case for the GOV.UK website.
But make things open, and you'll make things better...
I'd add messaging in as the other key benefit of an app over the web. Although it's far from perfected, getting hospital appointment details on the NHS App is definitely a gamechanger, compared to getting a letter that I then put down and forget (if it ever arrived in the post in the first place). And yes, I'd be happy to get it via email, but I can understand why others can't; both from a privacy issue and that email inboxes fill up pretty quickly.
It does seem a bit weird when HMRC email me and say "We've got some important information about your tax status... but we can't tell you what it is, and we can't link to where you can find out about it." Or the DVLA reminder emails when your tax/MOT is due, which inevitably get lost between the endless emails from DPD updating you that they're delighted to have received your parcel and it's now at their fulfilment centre.
Sir Matt Wakeman Esq III says:
WebBLE is quite widespread, WebNFC still pending real use cases and WebAuthn is starting to come to life - the challenges with all of these types of web technologies is the adoption and implementation at the browser level.
My hope is that rust complied into WASM used with WebGPU will bring a whole new platform.
@Edent I still believe that (notwithstanding the exceptional features you mention that the Web platform hasn't got yet) by linking to a Web site, an app can fulfil its main function, which is to be discoverable to people that don't trust or like finding things in their browser. But I'm an old fart too now I guess 😊
@Edent I was under the impression that the web is closing the “gap to native”, but your blog post makes it sound like we’re moving in the opposite direction.
Google has already brought all of these to Chrome on Android.
A good team can create a web app that works offline without any issues and that is several times smaller than the equivalent native app.
@simevidas @Edent
I believe Firefox and Safari have opposed supporting NFC and BLE.
Offline use of infrequently used web apps are complicated by Safari discarding stored information after 7 days of non-use. I don't know about other browsers.
@Edent There are good reasons for using native apps - such as the mentioned NFC/BLE access, or outright performance (gaming). For something like Gov.uk though, web should predominate - iOS/Android apps won’t help the young lady seeking benefits info on her PSP (/Steam Deck). It would need to be something highly specific (like the COVID tracking app) to justify the effort/expense. A shame PWAs are not more popular (albeit Apple have just killed them in the EU).
@Edent About most of the things you claim the web can't do, you're technically wrong, the very worst kind of wrong:
https://fugu-tracker.web.app/
The only reason they aren't pervasively available is Apple, but instead of arguing the side of open tech, you're folding. Deeply disappointed. Fugu API Tracker
I’m torn because I see why there is a push to apps but I get sad because to me it is a sign Apple’s strategy is working.
The open web should prevail but it will really really struggle to if Governments capitulate to the degree being suggested.
@Edent Obviously my field (box offices) is small and niche, but the kind of venue manager who says "wouldn't it be great if we had an app so that people can buy our tickets just as easily as they buy groceries on the Tesco app or tat on the Amazon app" have not even thought about answering the question "why the fuck would anyone install an app for something they do twice a year?"
@Edent last para is a bit of a provocation, by which I mean can you not poke me where it hurts?
Mike says:
Summary of a conversation I had with someone not long ago: Me: Here’s link to the thing on their website (send it via WhatsApp) Them: silence Me: Well? Them: I’m just downloading their app. Me: What? Why? Them: It said to to install the app Me: You don’t need the app, all the info is on the website at the URL I just sent you. Them: But it said to install the app. Me: No, it didn’t, the website told you an app is available, there is no reason for you to install it. Them: But app.
I think a lot of apps exist for the benefit of the people making them rather than the people using them. They exist as an attempt to drive engagement, and/or gather data, which hopefully can be turned in to money. Open website, provide only the information you want and possibly none at all, look at what you want, close website. But if you install the app, look look here's a notification about how horse_shagger_2001 just posted a new comment, open the app now, engage more. Why not give our app access to all your contacts so you can find your friends and psychiatrist and dentist and racist Uncle on here. Why not give our app access to your location so something something user experience. Please don't look too closely at what other permissions are being requested or what we'll do with that data just say yes because something best experience something personalisation something.
My car insurance provider has an app. To paraphrase someone above, why the fuck would anyone install an app for something they probably deal with once a year? The company has a website where I can look at my policy should I ever want to. What could anyone ever need to do regarding insurance that could possibly require use of a smartphone app because it cannot be done on a website? And yet, the Google Play store says their app has over 1 million downloads. And of course they have an iOS app too, I can't see in that App Store how many downloads that has. I'd love to see the data of how many times a year anyone opens that app.
I think some people, whose primary view of the Internet is through their phone, don't really get websites. They've had apps for everything, no matter how trivial or pointless, shoved at them, and they just accept it as the way things are. And that's a shame. A lot of people would probably benefit from fewer apps trying to engage them.
James Higgott says:
I proposed (unsuccessfully) a session on this topic at UKGovcamp XL!
I agree there is a place for apps for services you use regularly and where there is a genuine reason why it needs to be an app (offline use, biometric security). I've not come across the 'eyeroll, grandpa' argument before but I can see how some people might just prefer apps.
We're exploring options to make the most of native functionality on the NHS App right now: maps integrations, digital wallet, native accessibility features, etc.
I was bored yesterday on a train and challenged myself to rapid fire (it was only Bristol to London) jot down as many (hardly) 🔥hot takes🔥 that might make for interesting blogposts that might gener…
In our local government operation in the U.S. (Columbus, Ohio), we've effectively banned development of any new platform-native mobile apps and directed our (very small) team to focus on mobile-optimized web apps instead. But the reason for this is different than those stated so far in the comments.
My take is that we should eat our vegetables (build universally-accessible web apps) before we eat our dessert (craft platform-native mobile apps on both iOS and Android). We should only leap to mobile-native apps after we've repeatedly kicked ass with web-based mobile-optimized code.
A few years back our team started building apps via Xamarin (ensuring the results were lackluster on both platforms), then struggled to keep them current or grow them, amidst lots more work in .NET, JavaScript, and some low-code platforms, too.
We simply couldn't do native mobile apps justice, and even when we did create them, the functionality was no better than we could have achieved with an HTML5-based app.
So while it feels like we "banned" mobile app dev, in fact we banned it only until such time that we've "graduated" from base platforms to higher-order platforms. We're paying attention to our "capability and maturity" and building according to our capacity.
Sam says:
how long till a new SofS wants his face on the splash screen for the app?
@edent says:
What's to stop them asking for it on the departmental homepage?
I’ve been thinking a lot about public sector apps.
Recently a combination of Terence and James forced me to confront my prejudices about mobile apps for public service needs in a way that I haven’t done since the days of Tom’s seminal blogpost. I’m…
Terence, I cant believe you have FINALLY admitted apps are "sometimes" worth having. 🙂
My work here is done.
I think we need a drink. come say hello... it's been too long.
More comments on Mastodon.