Episode One Hundred and Fifty Seven: The Mezzanine; Doing The Job; Minimum Viable Whimsy

by danhon

0.0 Station Ident

4:37pm, whilst the spare car, the one we would’ve sold by now but for the idiot accident that happened to us, is being cleaned. The strange situation where the car wash offers free wi-fi, but a hotel that costs a couple hundred dollars a night doesn’t. Because hey, customer service. Anyway. On with the show.

1.0 The Mezzanine

If you lived in the United Kingdom in the early 2000s, one of the things that you might have noticed was the rapid supermarket growth. It was about that time that Tesco, one of the leading retailers, cemented its position, owing its success as much to a property portfolio, a data warehouse connected to a loyalty card scheme and supply chain management. At its height, about 14 and a half pence of every pound spent in retail in the UK was going to Tesco.

One of the many factors that went into this was the mezzanine planning permission loophole. Imagine the situation: to increase revenue and profit, you can either increase the revenue per square foot of your stores, or you can build more stores. So when someone comes along and says that you can increase the square footage of your *existing stores* – ones that you already know are at capacity in terms of retail space and have high traffic, you’d be interested, right?

This was the mezzanine planning permission loophole: some clever person figured out that there were no planning permission restrictions affecting the building of mezzanine floors – the equivalent of the platform 9 and 3/4 at King’s Cross, the realisation that these big box stores had enough ceiling height for you to insert a brand new floor, giving you around up to 50-75% as much retail space, without having to worry about any regulatory requirements or paying out for the same amount in buying or leasing actual land.

Of course, the regulatory loophole was closed. It had made a mockery of the government of the day’s promise to restrict out-of-city development at a time that high streets were feeling increasing competition from new sub-urban developments.

This feels like the kind of disruption that we in the software industry like. Finding problems to change the world, right? It makes me think of Amazon. One of their major costs right now for the part of their business that is significant (selling and fulfilling physical product) is distribution and fulfilment centres[1].

So here’s a silly thought: forget about all those giant warehouses out their in algorithmically-optimised distribution areas for transport logistics. The land costs money, and sometimes they could be closer to the customers than they might be. It’s an n-body problem, right? Plus, there’s all those pesky regulations about how big a warehouse you can build and where you can put it.

So put them in the sky. Giant redundant arrays of inexpensive zeppelins, serviced and supplied by smaller drones, where picking and packing are performed by Kiva robots, deliveries made by shoving the packages overboard, GPS and inertial guidance systems tweaking fold-out fins deployed from the packaging cardboard itself, yet another use for weather-resistant Tyvek. Put them high enough and it doesn’t matter, right? And anyway, how long is it going to take the government to legislate for a right-to-light? You want fast, cheap consumer goods before you’re going to care about the Amazon Distribution Centre high above your city.

Or, take the distribution centre and explode it. In the same way that Zipcar tries to cut a deal with you if you own your home and have off-street parking, let Amazon park one of their distribution shipping containers – oh, okay, a *half* shipping container – on your property. It offers local-pickup for people in your neighbourhood, and in exchange for becoming part of the distribution network, you get a free Fire phone and Amazon Prime!

All because, of course, you want cheaper access to fast moving consumer goods.

[1] Why Amazon Has No Profits And Why It Works – Benedict Evans, Andreessen Horowitz

2.0 Doing The Job

I signed up to do speed mentoring sessions at How Interactive DC last week – one-hour sessions on Thursday and Friday with 10-minute slots to spend time talking with people. I like talking with people. Between those mentoring sessions and the prep that I did for the talk, it felt like I’d been reminded about how tricky it is to do our jobs sometimes.

At least a couple of the people I met with asked me to give feedback on designs or wireframes, and again, it felt like sometimes the most valuable advice (at least, if I’m to believe what they told me) was to look at things from an outside perspective and ask: well, what are you trying to accomplish? A non-profit, for example, was launching a new service, but had prioritised putting a video that explained what it did above helping people use the service in the first place. Frequently, designers would say that they didn’t have clarity from their superiors on what the service was supposed to do. Or again, to use the above example again, “our organisation is launching a new service, and this website is for that service” – so why is the first thing on that website a video that explains the organisation?

I used to read the Economist and let my subscription lapse, but occasionally pick up a copy when I’m at an airport and desperate for something to read. This week it was an easier decision – it’s their Technology Quarterly issue, and I like knowing what other people think is worth knowing (bio-printing, Internet of Things, connected cars, air traffic control and so on stood out this time.

But it was the Schumpeter column on business and management[1] that caught my eye this time because of its promise of pulling out three core issues from the latest McKinsey Quarterly, ones that would “preoccupy managers for the next 50 years”. And yes, some of this is because I’ve been infected by the Cult of the Government Digital Service, but it just felt like the three core issues were a bunch of management theory fiddling while your organisation gets burned from the inside out and Disrupted by a bunch of people who understand Software.

Let me mansplain: the three issues identified by the Economist were smart machines, boosting productivity and the third wave of globalisation: mid-tier developing cities. From my point of view, the first two appeared to be subtly hinting at, but not really pointing out, a bigger issue, and again, I think it’s because of a fundamental misunderstanding of what software can do when allied with good organisation and management.

For example, the gist of the Economist’s focus on the first part – smart machines – is that much of the work of executives and management will be automated: “much of the work of bosses, from analysing complex data to recruiting staff and setting bonuses, will be automated.” The article pulls out a couple of answers here – one is of Google’s “human-performance analytics group”, the data-backed group that tries to put some science in human resource management, and the algorithm that’s been “appointed” to the Deep Knowledge Ventures board of directors and has a vote on what sort of companies the VC invests in. I covered the latter back in issue TKTK. Senior managers, says the Economist “will have to rethink their roles dramatically if they are not to become latter-day Luddites. They will have to hand some of their functions to intelligent machines, which will always be better at data analysis than humans, and some to the heads of business units, who will be in a better position to make use of the crunched data.”

I think this misses the point that defines the continuum between people like Marc Andreessen with his software-eating-the-world position and the GDS organisational change linked to IT done right position.

Nowhere is there the consideration that part of the job of the manager or executive is to identify areas in which smart machines can be deployed in the first place. Nowhere is there this idea of the executive focussing on the way that automation and technology can be applied to make what they and their teams do more efficient.

This ties into the second point, that of boosting productivity. Apparently, we should be optimistic about improving worker productivity because the “IT revolution is turbocharging what once looked like mature management technologies such as lean production and supply-chain management. Cloud computing lets small startups harness computing power that was once reserved for big firms. But the biggest potential gains will come from focusing on areas of the economy that have either been overlooked, because of a lack of imagination, or have stagnated, because they are protected by powerful interests.”

Schumpeter appears to miss the whole point of the IT revolution in the first place by focussing on productivity-boosting uses of industrial materials, whilst ignoring the already-impressive example of the UK government saving at least fifty million pounds a year through better systems.

It may be because I’m being overly simplistic, but part of the big changes that we’ve seen in terms of efficiency haven’t been because of cloud computing, although the cloud has certainly helped in terms of helping people get to product faster and a shift from capex into opex at the stating phase. They’ve been instead from using software as a material to solve problems from first principles. Instead, I feel it’s been people who’re familiar with what software can do – and the nous to know how and where to do it – and they’ve finally snapped and had enough. We should be able to do better.

[1] Schumpeter: Three issues that should preoccupy managers in the next 50 years – The Economist

3.0 Minimum Viable Whimsy

157 – I haven’t even *used* Slack[1], so this is the absolute worst kind of armchair internet punditry, so take everything that I’m writing here with a grain of salt as big as the amount of money in a Facebook acquisition. So.

One of the interesting things about Etsy is the way it’s taken a culture of dev/ops from initial beginnings at Flickr and exploded it into something that’s a part of the company culture of Etsy. I may well be projecting here, but it looks like what Stewart Butterfield and Cal Henderson have done is taken a look at the way developers and, I suppose, in a way Tim O’Reilly’s alpha geeks, have used a service like IRC and said to themselves: what would it be like if we made IRC (and its bots) as easy to use for regular people as it is for us?

It feels like irccloud tried to do something like this, but ended up creating IRC for people who know how to use IRC. But it feels like there was value in having the distance to step back and say: what are the bits about IRC that are, well, *good*, and would they be good for other people to use?

So you get channels and you get bots through webhooks. You get persistence and you get scrollback. You get conversation – which we know humans are good at – as opposed to correspondence. And maybe you can through the unbundling metaphor to look at what Slack is doing too: we experience email-as-conversation, when it happens to be the medium at hand, and perhaps not the medium best suited for the type of communication that’s needed. But Slack is also an organisational change: I first got an invitation a long time ago when I was working at Wieden+Kennedy, and it didn’t feel like something that could work unless a whole team was using it. And when those teams had only just discovered Basecamp, well…

As much of this feels like finding the good part of an old piece of technology and working out how it might be made accessible to a wider audience now that enough of the workforce has net-connected computers and phones. And perhaps Slack would’ve been a harder sell to a workforce that hadn’t grown up with IM, that wasn’t acutely aware of mobile messaging. From the outside, and the way people are talking about it, Slack is a tool for collaboration because it’s a better tool for conversation in a way that email was so far the least-bad tool for conversation.

At the same time, you can look at something like Slack and something like Google’s ill-fated Wave and see that they both are pointed in the same direction, but something went wrong with Google’s – dare I say it – user-empathy. Butterfield and Henderson (and doubtless others who I’m wronging by not knowing and not giving credit to out of lack of research) have proven themselves twice now about being able to find the thing-that-resonates-with-mass-users that they accidentally built. Or, perhaps, it’s just their whimsy, that they’re not afraid of designing and producing solid things that have Minimum Viable Whimsy (oh god, I just wrote Minimum Viable Whimsy, but see what I’m doing, I’m not going back and deleting it, am I).

There’s a few factors going on here: the fact that (then) Tiny Speck identified (yet again) a need and also delivered against that need with a product that worked. Then there’s the fractal attention to detail and the Butterfield-esque friendliness that infuses the types of things that he’s involved in producing. I mean, notice the casual and accessible tone of voice pioneered by Flickr (and now adopted by Facebook, of all places, in their usage of Dinosaur Stickers to educate users about their privacy tools) that at times was compared to the Innocent Smoothie-ification of the English Language. That was a bit of a run-on sentence, but what I’m getting at is a sort of sideways nod to the whole GDS (if you’re playing Newsletter Bingo then you’ve probably won this episode) “digital services so good people prefer to use them” to a wishy-washy sort-of “digital products that are pleasant to use.”

There’s a difference in tone here – Government shouldn’t be matey with you, it’s fulfilling an institutional and societal role. Consumer/”enterprise” products occupy a different space – and it’s perhaps that “enterprise” bollocks that does a sort of context switch in the kind of language used in enterprise software. Enterprise software is serious. Enterprises are Serious Businesses full of Serious People in suits. But it turns out that a lot of those Serious People in Suits also take off those suits every now and then and apparently do things like go and watch The Hangover. This isn’t to say that there’s a casualification of language in business software, but that there’s value in getting to the point. An enterprise is just a business that’s pretending to be more serious than it is. We do this all the time, and in a way, you’d think that all of this context-switching that we have to do when we switch between different language registers takes a toll on our poor hardly-evolved brains.

[1] Slack

Thinking about Reddit. Annoyed about Reddit, like everyone else. Thinking about a recent car accident – well, the incident in which someone accidented themselves into the car I was in, and who, it appears, has subsequently claimed that our stationary vehicle instead was moving into theirs at the same time. At that, nothing to do, other than sigh and install a dashcam for next time. Or, in other words: get that dashcam, because when you have a word-vs-word insurance account and another party to whom you have no idea as to their trustworthiness, well. Get backup. No Destiny on pre-order. And, of course, the elephant in the room: Apple Keynote Day tomorrow, 10am PST.

Oh, and the one sobering thought, when I saw my primary care doctor for a follow-up after my fall. People have died falling from lesser heights. Time to get that life and disability insurance sorted out.