If you want to make a successful product, size matters – perhaps even more than how well it works

Is it the end of the Airbus A380? That was the question my colleague Simon Calder asked in his column yesterday. The answer essentially is that this depends on Emirates, the principal customer for the super-jumbo. If it does not order more, Airbus may have to shut the line down.

While the aircraft is extremely popular with passengers – if you have been on one you will recall how quiet it is – it is expensive to operate. It seems that people are not prepared to pay a premium for the experience.

This is not only a travel story. It is also a technology one, and as such has wider implications for us all.

The first question to ask is about scale. As a general rule, making something bigger – a car plant, a cruise liner, a power station – increases its efficiency. But this works only up to a point. There is a minimum size below which the costs per unit of output are too high, for example, if you only make 1,000 cars a year they are going to be very expensive.

But beyond a certain point the economies of scale are offset by other costs. Huge manufacturing plants are difficult to manage. People will pay more for smaller, more intimate cruise ships than they will for the mass-market floating hotels. A huge power station is a massive worry when it has to be shut down for maintenance.

What happens is that there are sweet spots where you can get most of the economies of scale but with more manageable risks. In aircraft, the Boeing 777 is the long-haul workhorse for that reason, while the 737 and Airbus A320 family are the preferred short-haul jets. The problem with the A380 is that it has to fly full to make money and there are not many routes that are busy enough for it to do so.

The second question is about technology: why some technologies flourish while others that appear to be equally promising become dead ends. There are lots of technologies that work perfectly well but are not economic. The hovercraft works perfectly well – before the tunnel was built we used to go across the Channel in them – but is too expensive to operate.

Others work but there are few applications: tidal power works but there are not many locations with both a large tidal range and an estuary that you can dam to trap the water. Even with a tried and trusted technology such as hydro-electric power, the world is beginning to run out of suitable sites to dam – and, by the way, these projects carry environmental costs that only become evident years after they have been completed.

I suppose the best example of a technology that works but is too expensive is nuclear power. If it were not for government intervention, no one would build a nuclear plant anywhere in the world ever again. It has proved much harder to reduce the costs of nuclear power than anyone imagined it would be in the 1950s, when Lewis L Strauss, the chairman of the US Atomic Energy Commission, said that nuclear power would make electricity “too cheap to meter”.

On the other hand there are technologies where costs have come down faster than most people expected. In electricity generation the obverse of nuclear is wind power, now so cheap that subsidies are no longer needed – though it may take a while before they are eliminated.

The most obvious example of that is mobile telephony. You could get a phone in your car in the UK in the 1960s but it was fabulously expensive. Now – well, a phone call is indeed too cheap to meter.

The other obvious example of a technology that will revolutionise our lives is the falling cost of batteries. The genius of Elon Musk was to see that the batteries needed for a laptop could be scaled up to run a car, and the switch away from the internal combustion engine was born.

So much has been written about this that I think the only further point worth making here is that if electricity storage costs do fall as fast as seems likely, all sorts of things will become possible that are not practicable at the moment – including the electric aircraft.

Meanwhile the A380 is a wonderful example of a product that works and works very well, but actually will be a commercial failure. Pity, but that is the story of technology.