I listen to the podcast Data Driven. One of the questions that they ask all their guests is: What do you think is the coolest thing in technology? This article is my answer to that. The short version is: standards, and the benefits they bring to users. The longer version is below.
Standards might seem the opposite of cool, particularly in the world of technology where the parade of mayfly-like Latest Shiny Things seems to never come to a rest. It feels like saying the coolest thing about haute couture is its health and safety policies. (For another thing I’ve written about things that seem opposite to each other, please look at User experience and data quality.) I hope you will let me explain.
I’ll start with an example. I can go into many different shops on the high street (or online, if I prefer) and buy a mobile phone. It’s not a big deal – many (but not all) people in my country can afford at least a simple one. If I wanted to, I could relatively easily get a landline contract instead, which would end up with a phone socket in my home, and I could essentially repeat the mobile-buying experience when I bought a phone to plug into that socket.
With either kind of phone I could phone someone in Australia if I wanted to, as long as I knew their phone number. The things I wouldn’t have to worry about or even know is quite large, including things like:
- Whether my phone would work the network I’m with;
- Whether the different components that make up my network’s equipment would work with each, despite possibly coming from different suppliers;
- Whether my network’s equipment would work with any of the other networks involved in getting from the UK to Australia, e.g. via satellite or undersea cables;
- And then the equivalent questions for the person I’m talking to and their network.
This means that if one of my children decides to go on holiday to Australia and has an accident, they can phone me for help or to just hear a reassuring voice (assuming they have mobile coverage, credit etc.). They won’t need to worry about any of the points above, and instead all the many moving parts of this system fit together well enough behind the scenes to do something very valuable to both of us.
One sign of the success of the phone system is that we don’t think about it most of the time, because it just works. A large part of what makes it work is standards, such as those governed by the ITU.
I mentioned this example to my son, and he pointed out how this example rests on other standards from other fields. I can charge my mobile phone using a charging cable I bought on the high street, that wasn’t made my the phone vendor. This is possible because of the USB standard. The electricity flowing through the cable does the job because the plug at the other end of the cable complies with UK power socket standards, and the electricity itself complies with UK power standards (in terms of current, voltage, frequency etc.).
The unglamorous underbelly
This is the result of lots of work, and didn’t happen by accident. There were lots of probably boring meetings, with agendas, drafts and minutes. There was lots of work around the meetings, going back and forth between the meeting and the organisations its members represented. There was a lot of project management to make sure all the paper was shuffled in the right way, nothing was forgotten, and so on.
Then, once all this had been accomplished, people got out of vans wearing high-vis clothing and did more hard work outdoors. This was things like installing cables and phone masts, and the equipment that connects them all together, and the power that keeps it all working. This equipment had to be designed, built, tested and so on. Then other people had to develop and install the software that monitors and controls these networks.
Not only is there hard work involved in standards, there’s other less pleasant stuff too. A standard is an instrument of power, and so people fight over it. I’ve described above how users can benefit from standards, but suppliers can benefit more or less depending on how closely the standard aligns with their interests, such as existing stuff they make or do.
I have a little insight into this. I played a very small part in specifying the design of domestic energy smart meters for the UK, by being part of a working group run by Ofgem. I was just some random programmer from a random IT company, but everyone else was from the industry – there were people from energy suppliers, people from organisations who ran the network (DNOs), and meter people.
The DNOs were slightly different from the others because they’re in a regional monopoly, so aren’t competing with each other. Everyone else was competing with other people from their own sector. So supplier A wanted the standard to match the meters they had already installed, which were different from the meters that supplier B had already installed.
There was also people in sector A jockeying over the boundary with neighbouring sector B, and this was influenced by how important sectors A and B were or seemed to be. Each sector wanted as many of the benefits of the project as possible, with as many of the costs dumped on the other sectors as possible.
In other situations this understandable jockeying between rivals can lead to an explosion of standards, with each rival putting forward their own standards. If there’s not enough external power to control the standards, there can be a kind of Tower of Babel where the world gets divided up into N incompatible standards. One approach to solving this is to then create some kind of umbrella standard that doesn’t replace any existing standard, but describes how to translate between them. This means that there are now N+1 standards, which is not always ideal.
There is an argument against standards, that they slow down innovation. Everyone has to move at the speed of the slowest, slowed down even further by extra bureaucracy. There is some point to this, but I’d like to make some counter arguments.
About two thirds of the time, the features added by software development teams have zero or negative value. The thing that’s slowing you down (in terms of rate at which you grow your customer base, increase your income etc.) is more likely to be your choosing to build the wrong thing, rather than a standards process preventing you from building the right thing.
Also, standards can lead to new things, i.e. to innovation. A standard can act as a watering hole, around which an ecosystem of companies can gather. If you think about HTML – there are web browsers, IDEs and code generators, reports that write their output to HTML etc, that can all be created because it’s safe to make that investment due to HTML being a standard.
There’s another issue to do with standards, which is ownership. Is a standard a de facto standard, such as logging into an account using Facebook, or a de jure standard such as HTML? I can’t come down too hard on one side or the other. A de facto standard can often start having an impact more quickly than a de jure one, so at least some of the user benefit will start more quickly too. However, I’m not sure I feel comfortable with Facebook, an already powerful company that already knows a lot about quite a lot of the world’s population via their use of the Facebook sites and apps, knowing even more about users as they log into completely separate 3rd party services.
But I have a similar uncomfortable feeling about companies being forced to give up control of de facto standards they have created, so that they can become de jure standards. (I’m aware that this isn’t the only way that de facto standards become de jure ones.) I know it’s about much more than just standards, but China shows how governments can have a chilling effect on business when they throw their weight around clumsily.
As I said earlier, standards can be instruments of power, and it’s important to think about who wields that power and for whose benefit.
It’s probably a weird thing to say, but I think that standards can be one of the coolest things in technology. Despite their limitations, they can deliver benefit to users and support ecosystems of companies.