0.0 Station Ident
I'm writing this poolside, somewhere in (very) northern California. We're on a family vacation, and I have to admit that part of me was seriously considering not writing an episode today, or at the very least just sending one that contained the words "Dan Hon is on vacation," or sending one that just says "Dan Hon returns on Wednesday 27 August". But no, I'm sat here, Macbook Air in my lap on a lounger, watching a pool cleaner robot do its thing. Robot looks like it's enjoying its job, though. 

The family vacation was off to a wobbly start - partway on our drive down from Portland we were amusingly crashed into by another driver which wasn't fantastic. You don't especially *expect* drivers in front of you on a main road to slow down, stop and sit stationary for a few seconds before reversing into you, but sometimes apparently these things just happen. And then just to top things off I'm pretty sure I stubbed my toe.

Anyway: I seem to have picked up a few hundred new readers thanks to this TechCrunch article on newsletters[1], so some housekeeping is probably in order. 

Things That Have Caught My Attention is a fairly eclectic mix of anything that's caught my attention. I write it every weekday, and Episode One Hundred contained a look back, clip-show-style at some of my favorite bits. They've included things like writing about the Californian Ideology, a streak on empathy (and its lack in corporate and organizational environments), a little bit about dealing with depression and a bit about the quantified self. In any event, the archive at TinyLetter[2] will always be most up-to-date, followed by an archive at newsletter.danhon.com[3], running off the back of a Wordpress instance that relies on me updating it by hand and adding the right metadata tags.

[1] Why Everyone Is Obsessed With E-mail Newsletters Right Now - TechCrunch
[2] Tinyletter newsletter archive (always up-to-date, less metadata)
[3] Blog newsletter archive (more metadata, slightly less up-to-date)
1.0 Here In My Car
The Paul Ford piece on The Future had a few points that are still pinging around in my head. Ford takes concepts that we have right now - subscription services and access instead of ownership and then brings to life a tale of how they might be experienced during a regular day about forty years from now. 

For starters, there's often a debate as to how quickly the driverless car future will come about. Ford's piece doesn't really talk about holdouts still manually driving cars, so we'll ignore that part, and instead look at another parallel: adoption of consumer electronics. I got my first mobile phone in 1998, but I was an early adopter - it was a tie-in with a student bank account that I got from Barclays, and was a super-early GSM cellphone running on the Cellnet, pre-O2 network - the late 90s were super early in the development of mobile phones and you didn't even have cross-network minutes, or even cross-network SMS compatibility back then. Anyway, I digress - less than twenty years later, you've got pretty much ubiquitous coverage and people who don't have mobile phones are the exception rather than the rule. 

This type of change always feels like it happens both faster, and slower, than people anticipate. I imagine that a similar situation might happen with self-driving cars: that we'll always imagine them to be about forty years away, and then wake up twenty years in the future and find out that they're mostly - but not entirely - there. 

Now, I'm just coming up with this off the top of my head and I don't really admit to having done any particular research. These aren't facts or projections or anything - just barely researched intuition, so make of it what you will. 

So take a look at the car market in the US - in the 7 months to 2014, about 9 million cars were sold[2], at an average selling price of around $31,000 - so you've got a market size of around $279bn before discounts are taken into account. For comparison, Comcast's annual revenue is around $65bn, and Verizon's revenue is around $120bn. 

The idea of subscription services is interesting, because one of the first things that a Disruptor will tell you is to sell the service, not to sell the product. Car manufacturers aren't necessarily selling cars, they're selling freedom of movement or transportation. Changing the automotive/car business into a personal transport business that has to grapple with concepts like Average Revenue Per User sounds like it's going to take some time for Detroit, and the rest of the industry, to deal with (modulo companies like Daimler, who have their Car2go pay-per-minute experiment - in fact, I'd argue that Car2go feels as much an experiment as Apple's Apple TV experiment - knowledge that the market may well be heading in a direction but that there are problems and issues in terms of go-to-market. But anyway.)

We learn about two "transportation services" in Ford's future - LessTravelled, which markets to single people under 40, and FamilyVan, which markets to, well, families. Both of the services offer value-added bolt-ons - LessTravelled subscribers get restaurant discounts, FamilyVan subscribers get things like free delivery of groceries whenever. 

So here's the back-of-an-envelope reckon on how you bring about a self-driving car future:

First step, obviously: figure out a way to make self-driving cars, and makes lots of them. 

Then, pick a metro area as your beta-test launch. Follow the model that Google have done - find a city with lots of young people with disposable income and do build-out city-by-city. Get people to campaign to bring coverage to your city - and there's another turn of phrase - service coverage. Because what you're selling isn't a car - it's a transportation service. 

OK, so you've done a deal with somewhere like Austin where you've come in and made sure that the regulations are friendly. You're making self-driving cars, and because you're a company like Google (or whoever) you've got lots of capital lying around so you buy up lots of land in strategic areas around the city so you can model distribution. This isn't a hard problem, because you've already got lots of (anonymized) data from Android phones about where people live and where people want to be, so you can model, roughly, what people want to do. Oh, it helps that you also have lots of (anonymized) calendar and search data, so you can model demand and intent. So: deploy all your cars, and do a deal like Car2go where you pay a flat-fee to the city so your cars can park anywhere at any time. 

Now: either do some introductory pricing, or, and this is the easy one: just work out what you need to price your cars at to make a profit *and* to be more attractive than the lease-option. Make your cars super covet-worthy, make sure they have lots of USB ports in them for charging Android phones and iPhones, and even have some silly ones in them that have fully stocked bars or whatever, or even hot-tubs. Because remember: you're selling transportation, not cars. 

The last one, I think, is the kicker. Work out what sort of usage plans you need to offer: do you go for the "unlimited" plan (that, thanks to the telecoms industry, is actually a "reasonable use" plan) that offers unlimited miles? Or do you offer Fast Response plans that guarantee access to a car within 3 minutes? Or even cheapo-plans that offer access to a car within 15 to 30 minutes? So many pricing options! So many upgrades! Do you let people pay a premium one day to upgrade to Instant Availability? Or, do you start bundling with other services? 

This is why it feels, to me, that the question of ubiquitous self-driving cars is a bit premature. They, like all futures, are just going to be unevenly distributed for a bit. In the UK, rollout of broadband was, well, rolled-out on a geographical basis because of the lack of penetration of cable services. Until local-loop unbundling was introduced across Europe, you had to wait for the incumbent telco to install DSLAMs in your local exchange if you wanted broadband internet access. It feels like there'll be metro pockets of opportunistic cities that will be up for introducing self-driving car areas in the same way that Google markets "fiber hoods". 

[1] Wednesday Aug. 20, 2064 - Paul Ford, Medium
[2] Monthly Sales Data - Automotive News

--

Signing off, because I'm on holiday. And as ever, send me notes.

Best,

Dan