Interesting Links for 16-06-2017

[andrewducker]

Microsoft AI plays a perfect game of Ms Pac-Man

(tags: ai games )

Definitive proof that lesbians cheat

(tags: lgbt cheating games funny )

The founder of Pinboard on why understanding fandom is good for business

(tags: viaSwampers bookmarks delicious pinboard )

City of Edinburgh Council to be run by coalition (a Labour/SNP one)

(tags: politics Scotland edinburgh )

List: Writing Advice to My Students That Would Also Have Been Good Sex Advice for My High School Boyfriends

(tags: advice sex writing funny ViaDrCross )

How the .NET Runtime Loads a Type

(tags: dotnet programming computers )

The idea that humans have a poor sense of smell is an outdated myth

(tags: smell humans brain )

Spotify hits more than 140 million users (Still losing money. Not sure it can actually make money.)

Spotify passes on 70% of its income to the labels. I wonder what the fixed costs are which push them into losses.
(tags: spotify music streaming business )

Streaming Drives U.S. Music Sales Up 11% in 2016 - Streaming was 51% of the market

(tags: streaming spotify music business )

Comments

[drplokta]

I've worked for a bunch of Internet startups. The highest fixed cost is always salaries -- all those software developers and marketing experts don't come cheap. Next is marketing, which I think Spotify does quite a bit of. Then premises costs -- rent, rates, electricity (not counting the data centre) and so on. And then data centre costs. Nothing else is likely to matter very much.

[danieldwilliam]

So is the question - at what point do Spotify stop needing to spend lots of money on uncapitalisable software development?

[drplokta]

Never. The most successful startup I worked for has been going for 18 years now, and its development team is as big as ever. There's always more to be done, the competition doesn't stand still, new devices and technologies become available, and so on.

[danieldwilliam]

There must come a point where software development falls as a proportion of revenue though?

[drplokta]

If your revenue increases, then sure. It's a fixed cost. It doesn't really depend on revenue.

[danieldwilliam]

Sorry, that was more a normative statement about loss making companies. They ought to arrange their affairs so that either software development costs fall against fixed income or they ought to increase income and hold development costs fixed. If fixed costs continue to be larger than gross profit then they are bust.

[theandrewhickey]

Yeah, but the business model is to burn through money as fast as you can in the hope that either you get bought out and the losses become someone else's problem, or your competitors spend even more and go bust first, leaving you with a monopoly so you can then increase prices a trillionfold.

[danieldwilliam]

I don't see Warren Buffet going for that.

[momentsmusicaux]

I really hope the SNP don't bring their rabid anti-tram policy and prevent the extension to Leith!

[danieldwilliam]

The Edinburgh tram seems both popular and profitable. It would be difficult for the SNP to be anti-tram. I mean they could argue that the capital required to build the extension is better used elsewhere at the moment but the tram seems to have proven itself.

[nojay]

The tram is popular but I don't know about profitable. The council/Edinburgh Trams is responsible for paying about £250 million of the construction costs of what's been built already (total pricetag about 750 million, Government contribution as an offset for Crossrail is £500 million).

Passenger numbers are about 5-6 million journeys a year IIRC. The tickets are £1.60 except to the airport which is £6. My guesstimate for the annual gross income for Edinburgh Trams is around £12 million or so. They have to pay for track and rolling-stock maintenance, wages for staff, electricity etc. out of those revenues. Servicing the debt (at say 2% p.a.) is going to take the rest and then some.

The lowest cost estimate I've seen for the Leith extension all the way down to Ocean Terminal is about £60 million. Things could go wrong like the main line's construction did and end up costing more. Leith has a lot of buses that cover a larger geographical area than a single tram route could match, including night buses (the tram stops running between 23:30 and 5:00 a.m.) so the tram is not totally necessary, it would only be an extra public transport option.

[danieldwilliam]

So the most recent stat accounts for Edinburgh trams are for the year to 31st December 2015. 2016 stat accounts are not due yet.

Those stat accounts showed a loss of £25k on a turnover of £9, 227k or 0.27%. In order to break even the trams would need to increase tickets sold by about 15,600 single tickets or 3,200 return journeys to the airport. Given that they've been increasing passenger numbers over recent year by between 5-10% I am confident that that the trams will have been profitable in 2016. 2016 for example saw the trams single best day for ridership due to the Six Nations matches at Murrayfield.

The extension down through Leith Walk takes the tram along the most densely populated parts of the city. Leith has a pretty good bus service. In fact the whole of Edinburgh is remarkably well served by buses but I think there are serious capacity issues coming in and out of Leith. In particular, getting from Leith to the west of Edinburgh is tough. The stretch of Princes Street from the top of Leith Walk to the Mound seems to be full every morning and there is another bottleneck at the West End junction too. Andy will know more about that, my bus journey is north - south through the West End junction so I don't experience Princes Street.

The geography is unhelpful. You could run more buses along the 36 route and jink around the city centre to the north but I'm not sure how much capacity that actually gets you. The bus service in Edinburgh is excellent but it might well be full along important corridors.

This is before one considers that much of the new housing that Edinburgh needs is going to built along the shore.

Part of what went wrong with the original build is that preparatory work was done along the whole route which ended up being stranded when the route was reduced. So something could go wrong with the extensions but in many way it already has gone wrong and been paid for.

I'm not saying the tram extensions are a slam dunk investment decision or even that they are the best option for a capital rationed civic investor but they turn out not to be the white elephant many Edinburghers feared they were going to be.

[nojay]

How much did the earnings of Edinburgh trams contribute to the £250 million debt still owed on the construction bill? That needs to be paid down by someone -- the fear for a lot of Edinburgh folks was that the profitable publicly-owned bus service would be sacrificed to cover this debt. Luckily this didn't happen.

Part of the problem with the congestion in Princes Street is the trams. They get priority at traffic lights which bottles up buses behind them, especially during the rush hour periods. There is also the bottleneck at the tramline turn up to St. Andrews Square where the bus traffic heading east is limited to a single lane to make space for the tracks to turn. I think too many buses go along Princes Street but I don't know there's any real solution to that because of the geography, although running more buses along George Street might help.

There is already a lot of new housing at Ocean Terminal where the proposed Leith tram line terminates which is dealt with by the 11, 22, 34, 35 and 36 bus service (including night services, not something the trams offer). The original business plan for the Leith tramline of moving thousands of office workers living there out to Gogarburn and the business park area of the Gyle each day fell through after the 2008 recession and the cutbacks in staffing levels.

[danieldwilliam]

Difficult to tell from the stat accounts how much either the tram operation, or the bus operation or the whole of the transport operation is contributing in cash towards the financing costs. I've spent more time than I really ought to have done trying to follow the cash trail through the consolidate accounts.

It's also difficult to tell from the stat accounts how much of a network effect there is from combining the new tram operation with the existing bus operation. Not that stat accounts are a useful tool for that sort of thing but I do wonder everytime I swipe my monthly travel card how income is apportioned between the two services and how the data is used to influence decisions on changes to services. There's also a separate discussion to be had on locating postive externalities when one is a government.

I'm less fussed about the night buses but no night trams issue. For me the tram is mostly there to move larger number of people than can go on a bus though the middle of town during rush hour and to run people out to the airport. At night the night buses aren't caught up in congestion.

I'm not sure that the logic of shifting people from the shore to Gogarburn is entirely destroyed by the financial crash and the demise of RBS. Edinburgh is still booming and continues to grow in population.

Ahead of the trams being built I'm not sure I'd have been in favour of them. I think putting the money in to new buses and new traffic layouts for buses would probably have been better. The build process was certainly not well managed. I'm not convinced that any public transport is going to survive unchanged after contact with autonomous vehicles. But so far they seem to be performing better than expected and better than budgeted so I'm happy to treat any proposed expansions on merit. I think running the existing line down to Leith is pretty obviously the right decision from where we are. It is likely that extending the line further west would make sense.,. I'm not convinced that expanding the network beyond that, adding extra lines or what not is that clear cut.