Alex Blandford @alexblandford - Tumblr Blog

Scaling to meet demand.

A couple of years ago, I switched to a new type of blood glucose monitoring. It was a bit of a learning curve (sometimes literally). Trying to manage your diabetes with a lag time of 30-60 mins in the readings is difficult and makes you bump up and down (as your insulin also has a lag time) until you get the hang of it.

But there have been other problems.

Firstly, the app has been horrific. It was advertised as working with most android phones with a modern enough version and NFC turned on. However, it kept bricking sensors. And I mean, within half an hour of activating them, they would be bricked. Each sensor cost about £50. Gradually, the phones listed as supported on Google Play reduced to just a Samsung Galaxy (suggesting that the pool for testing had been limited) and as I was using either a Nexus or Pixel at the time, I thought it was weird that it wouldn’t even work with Google’s reference hardware.

Secondly, supply. When the product was first released, it was rare as hens’ teeth. In a sense, this was to be expected, it was an affordable way to get beyond fingerprick testing and it was from an established, good company (I always swore by freestyle over bayer as their test strips seemed to fail less often and needed a lot less blood).

However, the growth of the customer base hasn’t been matched with an increase in manufacturing capability. For the past six months, the website has shown problems with stocks and supplies.

There are big reasons why this is a problem:

1) It’s expensive to stockpile. At 50 quid a pop, I don’t like dropping hundreds of quid on medical supplies.

2) You can’t stockpile because stocks are too low and they’re restricting orders to two per person per 25 days

3) Each sensor lasts 14 days, but the delivery time quoted is longer than 14 days.

So, what they’re doing is making sure that ordering their product becomes something you have to think about. You can’t set up a subscription of two a month, you have to remember. I tend to remember when I put a sensor in my arm and notice that that was the last one.

I can get hold of other test strips, but they haven’t been on my prescription for years. It’s more faff. And it is more faff because a company has let their production ability fall behind their marketing.

I’m trying to say, I rely on this technology. It has improved my life a lot, but the supply troubles are really worrying. It is hard to imagine this getting any better as more areas of the UK allow Freestyle Libre on prescription and Brexit potentially screws up the supply chain.

As I mentioned in my first blog on the company, the most frustrating thing is that they don’t reply on social media. They don’t email you to say there are supply problems or iterate the product to prevent ebbs and flows in demand from people stockpiling (by allowing you to set up a monthly subscription).

I wish organisations like Diabetes UK and Input Diabetes would put a bit more pressure on Abbott to up their game in terms of dealing with customers. We pay a pretty penny for their product, and we are rewarded with fragile apps, fragile supply chains and more life admin for a disease that is already a X2 multiplier of life admin. I also wish another manufacturer would bring a flash glucose sensor to market so that Abbott might have to start behaving like they needed to compete to keep their customers.

#t1d #doc #diabetes #medtech

Lego laptop

A week or so ago, I left my job at Wunder. That’s a longer story and really, one that I’m not going to post on the internet. There are two side effects of leaving though: 1) I’m now available for contracting and permanent roles 2) I had to give my work MacBook back.

Me and that box had some good times. I dropped it on airport floors in three different countries, I completed the StarCraft 2 campaign on it and it had my best and most offensive collection of stickers to date on the lid.

I don’t own a lot of expensive tech on my own steam. I buy Nexus phones and cheap laptops. So, to deal with needing a business machine for a while until I get a big contract, I took two approaches. I got the cheapest named-brand laptop (a lenovo ideapad 100 since you ask) on amazon and an RPi3 and saw if either of them could be all business and no pleasure.

Let’s be honest, the Raspberry Pi was a bad idea. So the laptop got used instead. I took it apart to upgrade it from 2gb RAM to 8 and set about getting linux to work on it (one of the reasons for cheapness was no bundled operating system). This was harder than I thought - it has been a few years since I ran linux as my main machine and I’ve never had to make it play nicely with UEFI before. I had to make a few changes to kernelly things to get wifi to work and the lid being shut to suspend the laptop.

It also came with a German layout keyboard, so I’ve had to use some bright stickers to enforce the PROPER qwerty layout that god intended. It looks ridiculous though and has earned it the “lego laptop” nickname.

I’m using a mashup of Mint and Lubuntu, running Vivaldi as the main browser, with slack, spotify and skype all runnning at the same time. This has meant that I am able to do just about everything that a modern digital consultant type needs to do, even if the webcam makes me look like I’m being caught in the white flash of a nuclear blast.

Bad points:

I think it is using an Intel architecture from many years ago. The supposedly 2ghz dual core processor maxes out regularly and for very little reason (changing windows, visiting a page with javascript).

Plastic. Oh god, so plastic

Taking it apart broke quite a few things so now there is a bit more wobble than I’d like

The screen says HD, but has a 1366x768 resolution not much above my 800x600 laptop from 1999

Youtube is all clanky and the audio jumps around a fair bit

There is only one speaker and it is really quiet

Good points:

All in, it cost €225. Cheaper than a Chromebook

It actually, basically works.

USB3 port.

Reminds me of how not everyone is using a macbook and that cheap computers don’t do well with bells and whistles on webpages.

It can run Steam for Linux and some not graphically intensive puzzle games.

So. Hire me to do some stuff so I can at least buy a refurbished 2009 iBook or something.

Businessing the business. Business.

I’m starting a business for reasons that I’ll go into in another blogpost. I’m probably about the demographic for who the Treasury would like to start a business - I work in digital, I sell non-tangible though services and I know how to do business across Europe (Exporting is GREAT apparently).

Thing is, I don’t really know much about business. But for various reasons I need one. My lingo level is _low_. So I went to GOV.UK to try and fix my no-business-yet blues.

I’ll be honest, it wasn’t the best experience of my life and I was kind of surprised as to how fragmented it all was.

There are a couple of caveats to this; the first is that I’m not clicking on anything that looks old. I’ve rationalised it to myself in two ways, firstly that the UX change is so significant from modern GOV.UK to old things that I could be forgiven for thinking it was an XSS, secondly, I used the passport service that’s made of hammers and 2006 and I don’t want to go through that pain again. In fact, I still don’t understand why it was allowed to be called a beta. Even Civil Service Jobs looks and acts better.

So, with that in mind, let’s try and form the company. By googling, obviously.

Right. So, lots of sponsored links, we can infer that this is a pretty competitive keyword. But, I’m young-ish and digital native (ugh), so no ad clicking for me. Let’s look at the government options. I’m looking for a big green button. Register a business. Button. Big. Green.

Nope

Nope. And considering the page is called “Start your own business”, that is a shock.

HELP I’M TRAPPED IN A LIST OF WORDS I DON’T KNOW. Apart from farming. I’ve heard of that.

Is this it? Does the act of registering for tax actually start the company? I guess this page is mobile friendly by dint of only being possible to view on a normal screen if you’re using a Nokia 3310. Let’s try somewhere else.

Ah good. Information on not-GOV.UK that should only be in one place. Why is this here? It’s almost a direct duplicate of the GOV.UK information. How’s the analytics on this page going? Why would I even be on this site rather than GOV.UK?

The Companies House beta site doesn’t have it, and you’d think it would. It is called Companies House. It is where the companies live, right? Still no green button or links elsewhere.

More clicking around and finally getting off the Companies House beta site got me to here:

YES! A familiar “do this thing now” GOV.UK button. I remember this from such processes as registering to vote. I do want to point out though, for three sets of search terms, this was not on the first page of google results. That is bad. Let’s click though anyway.

The fuck is this? Accordions? Immediate account setup? At least there’s a list of things I’ll need before starting though.

FNGNgnfnfngnannngngnngng*grunt*

None of these pages provide a clear, consistent journey funneling towards starting a business. None of the “actually do the thing” pages was in the first 5 google results. The pages swaddle you in “guidance” without a call to action other than reading more guidance. I don’t want guidance notes, I want to see the forms, see what’s needed and then read up on the bits I don’t understand. They’re split over six separate websites, with four different look’n’feels.

The aims of the government (I’d say probably of whatever stripe) are to have more successful businesses. It’s a pretty key plank of capitalist market economies. The current journey doesn’t meet my needs, and even if there is some swanky future with Government as a Platform components, this is a need that is broken now. It doesn’t necessarily need development to get a bodge in, it needs design, content structuring a focus on this that matches its importance to the government’s policy aims. Hashtag long term economic plan (sorry).

One last caveat. This is something not done, but I am forever of the cockup over conspiracy school of government projects. This moment of setting up a business is a funny thing to research - you may not already be a small business-person or define yourself as such. It is hard to prioritise a group that doesn’t lobby or self-define.

Things may be in the pipeline, but they could (always) be clearer. Especially when it comes to timescales. GOV.UK replaced business link many years ago and going through this process, it is a bit of a shock that it is this complicated and fragmented.

Digipology

I saw a tweet from Tom Loosemoore this morning which unknowingly wandered into a landmine zone of an argument I’ve been having with myself and other people for a while, so I thought it’s worth going over the pros and cons of it all (quickly, I promise).

A great user researcher is an anthropologist of the internet-era

— Tom Loosemore (@tomskitomski) April 12, 2016

What is an anthropologist?

Anthropologists are mostly academics practicing a type of fieldwork called ethnography. It involves participant observation for between 12-18 months, broadly defined as “whatever the people you’re studying do, you do, and then you write it up”. Anthropologists have evolved from being the “handmaids of colonialism” as Talal Assad put it, emphasising their role in understanding (mostly colonial) populations better so that governing them would be easier, to a diverse field of study covering the entire human experience, with protracted arguments around methods of fieldwork, the exportability and comparability of human cultural objects and events, right up to the nature of reality.

What is a User Researcher?

A user researcher is what an anthropologist becomes when they realise that anthropology on its own pays precisely 0% of the bills (Hi!). By and large, they work improving a product or service so that selling it, or promoting its use is easier. The primary fieldwork tools are usually much shorter interventions in timescale and more focused on the product or service in hand.

Production of knowledge

The big difference however, is in what that knowledge is for. None of the methods of a user researcher are invalid. And I agree with Tom’s broader point, that a large, fragmentary ethnography of the internet is being carried out.

What is missing that anthropology tore itself to pieces over is more of the reflexive voice of the researcher. How do they deal with their prejudices, expectations and so on?

Where does the knowledge go? Academia is slowly getting better at opening access to the results of publicly funded research (Open Library of the Humanities being a sterling example). However, the knowledge that is gained doing research that is done for an organisation usually stays inside that organisation - the output becomes the product, not the knowledge.

Anthropologist in residence

Think of it like a Wikipedian in residence. Modern organisations conducting user research generate hundreds of hours of video, thousands of scirbbled or hurriedly typed notes and transcripts. Get someone to be a custodian of those with a remit to get deeper understanding out of them, and be open (but respectful to the participants).

What I’m saying is that user researchers aren’t anthropologists of the internet, yet. Just needs a bit more work, a bit more structure (no structuralism jokes, I promise), and a better interaction between the academy and business to prevent wheel re-invention.

tldr: You need to hire a user research librarian/meta-anthropologist

Nobody does it better

You can spend a lot of time comparing Parliament and Government. They are, as is regularly pointed out, not the same. And they have different ways of presenting information to the public and crucially, different ways of asking for a response.

Now. This is a difficult thing to get right. There are potentially hundreds of millions of respondents to a government or select committee consultation (as private citizens or businesses/learned societies(whatever they are)/sports clubs etc).

I’ve seen this from a number of standpoints - working in charities and responding to consultations, working in Parliament and seeing the effort in terms of thought and development and a genuine wish to be consultative and then seeing the user centred work in government as a contractor.

Everyone’s system here is broken. Because consultation exercises have to do the impossible:

Be short enough that it is somewhere between non-trivial and acceptable for someone to take time to fill one out. This means that you can’t do a line-by-line readings of draft legislation/policy as that would take far, far too long

Be useful enough that the responses can be analysed and used constructively. This means that a binary “do you like our thing please?” type question won’t be of use.

Uncharitably: not actually make any difference, merely rubber stamp the current direction of travel. Sometimes.

Get away from "just the usual suspects". This is a problematic line in that, the usual suspects are often the ones who give enough of a shit to fill out paperwork that satisfies clauses 1 & 2 in the knowledge of the potential for point 3.

But there’s the word eh, “paperwork”. Here is the current Land Registry consultation.

It follows similar patterns of consultations on politically charged subjects such as the NHS in that the consultation is a .DOCX file and an email address.

Here’s Parliament’s offering:

There’s also a page to help you structure your response.

Now. Yes, it is wordy. Yes, it is not “easy”, but actually, the task itself isn’t. It is about the same difficulty as a job application. Having guidance notes clearly linked is better than not. Setting users’ expectations early on as to the questions under consideration and the practicalities of that is good.

Government needs to improve consultation processes. Parliament does too. I know a lot of work has gone into these sorts of questions on the more fiddly, chin stroking end of the scale, looking more at the efficacy of the consultation approach as a whole. However, this feels like an area ripe for a joint parliament/government discovery phase with some principles and user needs that could be jointly applied.

But just whacking up a non-open standards document and an email address is not necessarily going to improve consultations. If anything, in these situations it makes the divide between those who know how to navigate government and its mental model, and those who are on the outside, interested but frustrated, even more acute.

Sugar subsidies

A few years ago I wrote about some diabetes tech that I got for free to try out (the reader is usually free unless super swanky, it is the consumables where they get you). Healthcare tech moves frustratingly slowly. Look where the NHS Beta team is up to compared to GDS - it’s not because they’re shit (they’re definitely not), but the implications of screwing up are often, literally, life or death. So things get more held up by compliance (or empire building shaped like compliance).

But I’m getting older and my youthful diet of cigarettes, beer and cake has meant that after giving up one, and moderating the other two, I want to get my shit together and look after myself properly. Now, to do that in this day and age, you need two things:

1) An insulin pump. This is a biggie and one I’ll come back to when my German health insurance pays for one.

2) Continuous glucose monitoring, or at least something that gives you better data than four snapshots of your blood glucose each day. Something with lines and charts. Design with data and all that.

Long story short, I got a Freestyle Libre. It’s the new hotness. I’m assured. It has a thingy that sits in your skin fluids and measures stuff and then you tap it with the reader and get the past 8 hours of data. MASSIVE IMPROVEMENT on fingerprick testing.

However, it wouldn’t be me if I didn’t gripe about it so:

1) The reader is crappy. It feels cheap. In an age of £20 android phones with RFID, why is it £160 for this capacitive screen catastro-fuck that is exactly as shit as its predecessor in 2012 in terms of build quality

2) Why are there only three languages? It’s not much data to add another one, unless your design has the same ROM limit as a tamagotchi or something. I don’t speak DE/FR/IT enough to be comfortable using them for medical decisions. Nowhere on the sales site did I see it say you couldn’t buy it in English in Germany. [UPDATE: Customer Service can’t give me an English one in Germany as “many people want one” and “insurance reasons” 👌]

3) No engaged twitter/social media presence. I can’t send this to the manufacturer for a dialogue because they don’t exist on social media. I have to ring them up.

4) No phone app. I can’t get around the shit reader by using my NFC enabled, brain-the-size-of-a-planet smart phone to do this. Apparently there is an app being trialled in Sweden and it will launch in “early 2016″ which is rapidly ceasing to exist. There are tonnes of github repos of people trying and failing to do this.

5) What is data? I want to get an insulin pump that can read what my sensor is saying. But each system creates proprietary data. Still. STILL. It only outputs data as a PDF, but hey now they’re in COLOUR.

Diabetes tech still feels like it comes from a committee of angry people who don’t trust you. It also doesn’t feel as though it was designed. It feels as though it went through some horrible waterfall UAT. It doesn’t feel as though it was tested with users for how real life situations work. It reacts to patterns of the past so my reader always feels like it is 10 years behind any of the other things I use.

Also the cat screensaver has gone. :-(

Elbow grease

Following on from the last post I was asked why I’m currently having a problem with “service design”.

I’m fine with service design. I’m not fine with “service design”.

A few years ago I heard a convincing argument for not describing yourself as a “feminist” as that gives you a shield against critique - ‘I can’t have been doing something sexist; I’m a feminist’. Instead feminism becomes something that you do. You turn a signalling self-identification into an active, performative state.

I feel the same about service design. I like designing services, I think there are specific skills, a canon of works around the field and a stupendously talented bunch of people trying to build a discipline.

However, I hate sentences that start “we are service designers, so we can fix this”. I went to Service Experience Camp in Berlin last year. A lot of the time I spent and people I met there were informative, interesting and funny, but there was a session at the end about government.

This is where it gets a bit “local politics”, but stick with me. Berlin is crap at digital government. No more or less crap than the rest of Germany, but basically, it ain’t Estonia.

So, this session was about creating a two week innovation lab for “redesigning Berlin’s digital services”. There are several problems with this approach.

Don’t ask (just) me what the problems are, do some research and prioritisation with actual users who aren’t internet wankers (like me).

Civil servants will not spend two weeks in an innovation lab without their boss’s boss’s boss’s sign off. “If we have it, they will come” is not a viable strategy. I say this having organised hack days where we only asked for 2 hours of time and received sod all interest.

Hacks don’t fix things. Doing a two week hackathon (or innovation lab, but be honest, they’re the same thing) doesn’t fix the problems, it creates an unrealistic expectation that problems can be fixed with a few weeks work by some “nerds”/people with round glasses/whoever is the cool kid this week.

Fixing big government digital problems takes elbow grease. It takes political campaigners to get the politicians buy in and the budget, it takes business planners and procurement people to get civil servants used to agile, it takes product managers and delivery teams to show value for money, it takes a range of disciplines, skills beyond sharpies and patience beyond saints.

Louise Downe from GDS was talking at this conference and made the most pertinent point - service design shouldn’t be a buzzword, it is just the design of services. Everyone needs to pull together. No rockstars, no unicorns.

Community

#community

Over Christmas, I was in Australia meeting my future in-laws. I circulated round a party, pint in hand, being asked again and again “I hear you’re in IT - what exactly do you do?”. The best answer I came up with after a few goes round was “You know when you have to do something governmenty online and it is shit? I try to make it less shit. Right now, that’s mostly on the internet, but a lot of the work is in the real world too and you’re going to notice the distinction less and less over time”. I’ve been struggling to put into words what I do for a while - partly because I used to self define as a user researcher, but so few clients pay for it regularly that I do other work to keep the books balanced. In the last year I’ve been a delivery manager, product manager, business consultant, unconference compere, agile coach and content analyst. I don’t like calling myself anything to do with service design at the moment as service designers are doing my head in.

A couple of years ago I said we should all just stop and go home. I’d been an organiser and it had been a hugely stressful experience and I felt that the day I’d worked towards hadn’t materialised. GDS had just exploded in size and suddenly everyone who had previously been Robert Brooked in a cupboard somewhere in their particular part of government had been given a Macbook Air and a mission to fix the government. Suddenly there were a lot of job titles, a lot of commercial confidentiality, a lot of us and them. That was three years ago and a lot has changed. Going back to Govcamp yesterday I was surprised how more animated, collegiate and brimming with ideas it was - especially when last year felt so much like a stalemate. There was a mission, a collection of people and a community again.

Fixy question/suggestions?

There are a couple of things we should make better. It was nice to welcome newcomers properly - that As a community, we need some rules to clarify the implicit “don’t be a dick” vibe. It would be good if next year there was something explicit about behaviour standards - it’s better to put the framework in place now before we have a problem. I noticed that loads of people were volunteering but doing things like having three people on reception after the main rush when you could really make do with one. The volunteering is great (having been one myself) but the value of quite a lot of the volunteers would be better spent in rooms, contributing to discussions (where possible and at risk of redistribution to other tasks at any time).

Sessions These were all absolute fried gold. After a couple of years of wanting something that I couldn’t put my finger on, it was wonderful to go to so many well considered discussions. I didn’t leave any of them, because I didn’t need to. Procurement I’m serious, this is one of the biggest issues facing good governance today - inexpert buyers.

@fpowrie I want to start a political party based on fixing procurement. We will have rousing songs.

— Alex (@blangry) January 22, 2016

Let’s fix this and skill up people to buy better. Otherwise the current process is all a bit

John Sheridan’s fireside lecture With John’s (ridiculously well-deserved) promotion to Director of Digital at the National Archives, this was billed as his last “niche” chat on legislation after years of introducing rapt audiences to the complexities of Norman French and computational law for taxation, we covered the problems with taxonomy categories for statute and how the Archives have been working to mash together the statute book into something useful, for citizens and scholars alike. I hope this session happens again next year too.

Review of the year 2015 was busy for a lot of attendees. Also our digital projects starting cocking up. We made the Daily Mail front page and Private Eye (as well as the New Camden Journal). We’ve all survived cockups and as a community, we need to learn to support and learn from each other’s failures as much as successes.

This is what I learned. We’re a community. We’re struggling with working on different projects, with different employers, but we’re all there for pretty similar reasons of making things better.

Let’s keep the momentum going, and find other tools than just slack to keep us in touch. We need to talk, share ideas, hold meetups and keep teaching and learning. Once a year is too little time to see so many people with so many ideas.

Tickets

Couple of points about what James has said on ticketing for Govcamp.

1) There are a clique of people who go to Govcamp. They are also the institutional memory. They are also the people who go to lots of these events and help ideas cross-pollenate. There is an advantage in having them there.

2) Scarcity of tickets is because the event is too small, not a problem with the ticketing system. If the demand is so great that they sell out in under thirty seconds then you need more tickets.

3) Live streaming doesn’t make something more accountable in and of itself. It just becomes another time to remember. And if the demand is outstripping supply by 3, 4 or 5 to one then it’s just another time to tune in and be disappointed.

4) As a sponsor, I like knowing that some people will be in the room. I don’t need a breakdown of job titles. But. It is helpful to know that the room will have groups of people that you know. In an ideal world, tickets would be easy enough to come by that you’d send all your clients, but I would like the ability to say that I’m not just putting money in to a community space, but something that brings everyone from academics, senior management, other agencies and so on. It would be harder to do that if the draw is random. Not impossible, but harder.

5) Logistics and social. What if you got a ticket in the lottery, but your colleague didn’t? What if you don’t know anyone else there? We should acknowledge that the people use it as a space to catch up and meet new people, and that that would be disrupted by lottery tickets.

I totally applaud the idea of opening UKGovCamp up, but at a time where the focus on govcamp in terms of accessibility seems to focus so much on ticket price and ticket allocations when there are simple solutions to those things.

Charge money, get a bigger room, give out more tickets. This is not a fringe conference any more, there are easily over 3000 people in the sector who would have something to contribute and capping at 200 will always limit conversations. We are not pioneers. We are an industry.

Products and Open Source: finding the public good

The other day on the U-Bahn home, I looked up at the screen to see it running on a loop with a shell script that kept getting stuck. My girlfriend was nowhere as interested in it as I was. I stood up. I looked at the My Computer icon, I’d seen it before. I knew it like an old schoolfriend. It almost winked back in recognition. It was Windows NT 4.

The first computer I broke was a Windows 95 box. I installed something and Windows died. I spent the day on the phone to tech support at Dell (who didn’t seem pleased to be talking to an 11 year old), who basically told me to reinstall Windows.

This experience of breaking things got me started on Linux. Finding a 8 CD box of Mandrake in the bargain bin at PC World (2 of the disks were too scratched to use, 2 others could do it with a little coaxing). It was magic to see KDE/Gnome run on my PC back then. To start to understand dependency chains. To play Konquest.

What was great was a product built by so many people, with “good” intentions, packed up in such a way that with a little help from the kid in the Red Dwarf T-shirt in the library, you could install it yourself while only running a 50/50 chance of frying your Windows install.

I wrote something incoherent a few months ago about why GOV.UK should be a ‘product’ like Linux. It was borne of a conversation at UKGovCamp in January 2015 with Simon Dixon from WordPress, talking about products, CMSs and user needs. It was crystalised by going to a German government technology conference where the example of the best user-led web project that anyone could point to was VATMOSS.

It garnered a pretty terrible reaction - people involved with GDS seemed annoyed that I was volunteering more work for them. Many people, entirely correctly, pointed out that without the laser focus on user needs, the code is worthless. Still others pointed out that GDS’s code is a pile of wonky shonk.

Make things open, it makes things..... ah.. who are we kidding?

Putting some stuff on github

isn’t actually properly open sourcing a project. I mean, yeah. The code is there. But I can’t run it non-trivially. I could potentially contribute to the code base, but would I be being useful?

Open Source projects require communities. Rules. Codes of conduct. Possibly even Benevolent Dictators For Life. I’ve worked in agencies that contribute to open source projects (WordPress and Drupal) and it takes time. Staff time that is unbillable on client work.

Maybe government doesn’t have that time. Fair enough. But don’t call it open source then. There is something of a fiction with GDS in calling its work open. Journalists are denied interviews if Cabinet Office’s Press team.

Another day, another interview blocked by press officers, despite the fact both I and the civil servant in question want it to happen. FFS.

— Charlotte Jee (@charlottejee)

October 1, 2015

Code is thrown over the wall. Trying to spin up a GDS service could take you longer than is reasonable. I count reasonable as longer than it takes to make linux sound drivers play nicely with a native install of Spotify (around 2 days is my record there).

https://www.danblundell.com/2015/02/maslow-how-to-spin-up-the-gov-dot-uk-needs-api/

User research is more closed off than it should be, despite it basically creating a national ethnography. It isn’t shared in a way that is useful even if you’re a contractor doing user research but you aren’t a civil servant. The research isn’t available to academics. This data shouldn’t be a secret. [There are bigger reasons for this - it's not for lack of trying to make it better, but it is a near Herculean task that isn't a daily priority]

So maybe GDS isn’t the right bit to ask.

Winning big

Tom Loosemore’s speech at Code For America the other day echoed Richard Pope’s comments

on the future of the digital state - that whoever harnesses its power of the internet will win big by virtue of having better national infrastructure.

Imagine a future where the UK has managed this. But instead of building things custom, half in the open, half in a developer’s head, we could turn it into national infrastructure in a box. I sometimes think about it like railways. The UK ended up defining the standard gauge of tracks. Maybe we can do the same with government on the internet. We don’t say what the trains look like or where the stations are, but we can say how wide the track should be.

I saw at that conference in Munich what the alternative is - governments with systems that actively protect against open data, systems that have more in common with huge ERP systems than user focused services, old blokes in old suits selling inertia as a service.

We need to spin out the approach and services. As the world gets more connected, there are more governments that are going to have to make their systems talk across borders. Let’s set the standards for that. Someone has to. Maybe not GDS, maybe a part of UKTI. Maybe an agency with blessing. Whoever. But soon.

Government and the cargo cult

As part of work with Parliament User Group recently, we’ve been trying to think about the institutional blockers to reform and “progress” (without the Soviet, vanguardist associations).

One of the ones that came up a few times was the issue of “we’ve always done it like this, so we shall continue to do it like this”. Inertia is a powerful force in large organisations, especially ones with an aversion to being seen to fuck up.

There is an obverse to this, especially in the case of Parliament and a lot of Local Authorities: “We’re not GDS, so we won’t do anything that they do”. Wishing to have a sense of individuality is not uncommon. We’ll do this thing - we’ll succeed. We’ll show them.

Both of these behaviours come from a lack of strategic understanding. Simon Wardley has put it around 1000x better than I could: “why does a general bombard a hill? It isn’t because 95% of other generals did the same thing”.

Understanding where your business/organisation is going is absolutely vital. Merely saying (in government especially) that you will do open source because GDS did it (but did they really) or that you won’t be focusing on user needs as you are internal only (no word of a lie, have had this chat) is doomed to failure, and likely an expensive and public failure.

Mimicry in this space is often called Cargo Culting; an allusion to a particular fascination of anthropologists in the post-war period (although it goes back a lot further).

You can read up on it at wikipedia including how it came to be used in this context. But my problem with this is that cargo cults and their formation have very little to do with the way that we use the phrase, and I am hugely uncomfortable with the ethnocentrism involved in using the phrase. It relies on an implicit assumption that people in “tribes” are somehow dumb and that they see an airplane and thus mimic it like children. There is a huge long list of reading on how we treat people in tribes like children, and rob them of a claim to complex thought. I’d like to think we can do better in our metaphors.

So I thought I’d try and reverse this a little bit. Taking the same broad brush strokes as everyone else, I’d like to consider Trobriand Cricket. It’s not quite in the same bit of the Pacific as the cargo cults, but it has a similar use of “western” ideas.

This game is a syncretic mix of cricket as introduced by missionaries to the Trobriand Islands (along with football), and the rituals of war on the islands. It includes teams whose victory dances for a catch include references to WW2 soldiers and military materiel. Many of the islanders are university educated.

Now. Unless we’re going to say that Norwich City’s canary branding is part of an animistic shaman cult led by Delia Smith, this all seems quite normal for sport. Or even life. You have a game, you change it to fit your local context. You understand your strategy. It is also worth pointing out that the Prince Philip Movement’s cult of personality around the Duke of Edinburgh wouldn’t look out of place in the comments section of a number of national newspapers.

Here’s where it breaks down though. The game was imposed by missionaries. The way that these people lived was changed irreparably by British colonialism. The power dynamics in this idea are huge. So changing the game becomes a response to colonialism. A way of maintaining some control against overwhelming power.

And there is where my discomfort is. What people in the Pacific were often doing was a response to the invasion of their land, and its use as a strategic assett in war. Many islanders died. The power relations and political history of the phrase have been shed so that it becomes a glib way of comparing someone’s actions to the old saying about the definition of insanity as being the repetition of something expecting a previous result to the previous time.

As I mentioned yesterday, ‘disruption’ can feel like an existential threat to the status quo. It is all around us, and the internet is going to change how you do your job. Or else someone who will change will put you out of business, or make your organisation seem too inefficient by comparison. Copying your work of the kid at the next desk won’t work. Learning together and open conversation might help. But treat this as a plea: don’t get sucked in by tired cliche. Shorthand is just that - a reminder of the nuance and complexity that has been left out.

Go out and explore that.

#government #anthropology

What’s good?

**** With apologies for co-option of the gif of a different, bigger and more important issue. ****

The thinker Gayatri Spivak used a concept called strategic essentialism (I am told at any rate, I saw her lecture once and was almost entirely lost, but I am not a specialist in post-colonialism studies). As I understand it, it means that while a group may contain a huge range of nuanced histories and lived experiences, sometimes for the sake of getting something done, you put those things to one side and work as a unit for the good of the ‘essential’ quality that has brought the group together.

It leaves out nuance to make bargaining easier, but it excludes smaller voices that need the most thought.

I want to talk about changing the guard. For two reasons.

The first is what Tom Loosemore said in his Code for America speech a few weeks ago that one of the key points around his decision to leave was a senior civil servant asking him to tone down his “internet jibber-jabber”. That was when he thought about a new generation needing to bang their own heads on the wall and take on the entrenched power struggles and fiefdom building of the modern British Civil Service.

The next thing was reading Tim O’Reilly defending Uber from a relatively mild critique. I’m kind of agnostic about Uber. I’ve only recently started taking taxis as they are too expensive for most people to use regularly. So is Uber if you use it instead of public transport. What bothered me wasn’t the usual stuff around whether drivers are better or worse off in a particular system - it was the position of the people writing about it and their needs from transport being transposed to being an a priori standard.

People talk about bad jobs and poverty as an abstract, especially in technology. But it’s real. Disruptive technology seems to be built around the user needs of a white guy in technology who is money rich and time poor and probably over 35.

This isn’t good enough. Disruption isn’t good enough. It’s just pushing over everyone else’s blocks unless you’ve got a way to work collaboratively to make things better.

And so we’re back to the question of “what’s good?”. Not what’s fun, or what’s cool. Not “good” in the Doctor Who sense of heroism each week beset by hand wringing.

Public works. Thoughtful, well architected, researched, ideological (in a nice way). No “will this do?”, no “consultation exercises” that don’t leave the department.

It’s interesting to see writing by people like Richard Pope and practical action by people like Sym Roe and the rest of democracy club amongst many others. Organisations like Democracy Club, mySociety and others are working in a space that’s trying to do the best it can do. Meetups and Unconferences like UKGovCamp, HealthCamp and Redecentralise are, again, trying to fix problems. For ‘users’.

Users (in the sense that we talk about them in Government) are a bit like a fractal, infinitely zoomable into nuance and complexity. We strategically essentialise them into personas for services, which is fine, except that when we’re talking about things bigger than a service, we need nuance, accountability and not the 80/20 approach.

We’re starting to see proposals for the future that look an awful lot like political manifestos. Changing the machinery of government so substantially is going to require a huge amount of thinking and democratic oversight.

So we’re back to the start of asking “What’s good?”. And for whom.

Are we content with the tone of discussions around what public life looks like in 20 years being set by people who would wish no regulation on Uber, describe government setting price upper limits as “price fixing” and make not one single mention of accessibility of transport in a rambling blog post?

I don’t. I want the people who take the reigns of these discussions to be thoughtful, inclusive and to rely on a thread of creating a public space on the web that can hold a whole country and its complexity, not just those with a token of superiority, whether that’s titles, money or a Nando’s card.

I want us to talk about what public good is. It risks us finding significant divisions between groups working on the same goal. I’m tired of appeals to authority that whoever is in charge knows best, whether they’ve organised a conference, written a book or run GDS.

I’m going to dig into this topic over the next few weeks. This is part of a concerted effort on my part to actually answer “what’s good” rather than just banging on about what’s bad.

@blangry

Government as a Platform that Pays

Dave and Phil have both done pieces around how Government as a Platform (the practice of creating re-usable chunks like payments systems to cut costs and increase usability). Phil’s blogpost (What's in GaaP for the Private Sector?) got me thinking - how would things need to be structured for separate agencies collaborating on projects with different parts of central/local government but the same building blocks?

A few words first - you know what is the most boring project in a web agency? Websites. Two hundred page information-and-a-contact-form websites. The kind where the project is about shiny, “pop” and there isn’t a penny in the budget for user research or content design. What is actually interesting to do is digital transformation - working on how to make a business work better using the internet.

So - the way I see GaaP is like Twilio or SurveyMonkey: a tool you use because Dear God there isn’t budget to recreate it custom for a project with 2 devs and six months.

But then, what if it doesn’t meet the needs we find our users have? What if the new (currently fictional) GaaP payments system doesn’t do the trick?

As an agency, would we be paid to do pull requests to a bigger system?

I’d be happy with that approach, but there’s a question here about what openness is and how we do it.

OPEN SESAME

Not a single part of government publishes their user research findings. Sure they blog, but if this was academia, one would expect their field notes to be held by an archive so researchers could peer review their evidence (evidence/review/epistemology debates on ethnography to be dealt with later). It is really weird that all this money is going into a distributed, multi-author study of how the UK uses the internet and thinks about government, but no one outside of a project can see the results or analyse trends.

Nor are any of the GDS projects open source in any meaningful way - the code is open but it is pretty hard to run it yourself. Coding in the open isn’t the same as an open source project with a community.

I think the big questions we have to ask before we can assess whether GaaP is good or bad for business are:

Are the components going to be managed by a sponsor and a community or just dumped on github

How will we assess user needs and changes if all user research is secret

How will contributions be managed and paid for?

And maybe it is too soon to answer all of them, but the first two would help us right now.

One Year On

It’s been a year since I upped sticks for Berlin. In that time I’ve read over 30 think-pieces in the UK press about why everyone should move here and so here are my pros and cons.

Good

Cheap

Government is nowhere near as racist as the UK. Loads of other problems but when you see a right wing politician talking about refugees and human dignity as being inalienable it is quite refreshing after the UK’s toxic political climate.

Food (high quality, easily available, cheap)

Beer (high quality, easily available, cheap)

Bars have enough tables and chairs. You don't have to have your group of 6 stand by the bog doors until someone spies a mantlepiece you can lean against

I've almost won the Das Gift pub quiz with a team of two (Quiz Trek: Deep Quiz Nine)

Even in rush hour there are more or less enough seats on the UBahn.

Summer actually happens (highs of 37 then thunderstorms, rather than perpetual grey/drizzle)

Bad

If you don't know German and your business isn't coding then it can be hard work/no work.

Government is paralysed about civil service/digital reform. Expect no Deutsche Digitaldienst any time soon.

Berliner Schnauze can be trying

There is a bubble of Anglophones and people who are ‘finding themselves’.

You pay more tax. You get more services. You can live in walking distance of your job, nice bars, your friends’ houses, the cinema, the theatre (when they’re showing a play in English or with subtitles) and you can get out of the pressure cooker of conspicuous instagrammed happiness and start to actually enjoy yourself. Here’s to a bit longer in Berlin.

Worst case scenario

Yesterday Mike Bracken announced he was leaving Government. It’s a funny time to announce such a thing, having earlier in the year both in print and to rooms full of people pledged to stay around and implement the Goverment as a Platform plans. It is also after the election, after the budget but before the spending review.

If you think that it doesn’t matter what happens with digital then you’re demonstrably wrong. The government wants to cut costs, and in some places that is quite a large amount of money. Entrenched interests in the civil service do not want to restructure the administration (read: put it online and have fewer civil servants), so the only place left to cut then is the actual delivery of the service.

The mitigation against cuts to service delivery is cuts to service administration. Whatever your views on cuts, one has to agree that it makes sense to try and maintain service quality if possible and to reduce the overheads. If there is just less money on the table because of politics then you need to be pragmatic in spending it as wisely as possible.

Without political cover (Frances Maude out of Cabinet Office), without a sympathetic head of the civil service (whatever Sir Bob’s failings, he was publically on board with the plan of CS reform and restructuring) and without money (rumours suggest a budget and role cut for GDS), there isn’t much hope.

I worry that this is a win for the politics of inertia, the politics of outsourcing and the paucity of ambition in clinging on to the old world.

I hope I’m wrong, I hope that the people in GDS have enough clout to make the argument, I hope that the example of things being done well online isn’t one that we’ll be looking back on in 2 years time as we trudge back towards headlines like this:

Let’s make sure it doesn’t happen.

What does public money look like?

Money.

It can be exchanged for peanuts and websites.

If you have one website, don’t spend money building another to host content that your other one can already do. It is wasteful. Parliament have just done that today with the design, build and launch of a 21 page microsite to host information about the potential closing of the palace for extensive repair work.

If you do insist on a new site, make the IA work properly.

Rather than letting it change from top to right hand side.

Don’t use PDFs as a default. They’re big, ugly, the text doesn’t wrap properly, they’re not as good for assistive technology and more importantly, they’re built for print and you’ve just built a whole new website rather than just sending some paper. Best case - put the content up in HTML and use a style sheet to make it print nicely. Acceptable case - HTML and PDF on an equal footing. Worst case - this ^^^^.

Or maybe, just maybe, rather than building new SHINY things, prioritise fixing the most pressing problems first. The site that hasn’t been redesigned in five years. The lack of a public product roadmap. The slow progress of any change at all.

The priorities of communications of the House of Commons are less important than the priorities of the public. Stop wasting time and money on a whole website for 21 pages hosting PDFs and duplicated information. Users First.

Missives

I miss people at home.

At the moment, chats on facebook, twitter and whatsapp are a weird way of simultaneously feeling connected and distant.

Trending Blogs

Recently Viewed Blogs

Alex Blandford