As we all now know, we live in an age of massive centralized digital platforms: Facebook, Twitter, Microsoft LinkedIn, and the like are the conduits for a vast amount of our personal and professional interactions, and they profit accordingly.
They are much in the news, especially Twitter and Facebook, because they had such impact on the U.S. presidential election.
I feel that when I talk about platforms and products most people’s eyes glaze over, and yet these things are big and important, and they make culture. So I wanted to just take a minute and explain. This way we can refer back to these concepts from time to time in future newsletters.
What the hell is a platform?
It’s pretty simple, actually: A platform handles transactions. For example:
- Buy something
- Sell something
- Send an email and see if it was opened.
- Save a blog post.
- Save an image
- Show someone an ad and track that they saw it.
Because they require hardware and software at the least, transactions always cost something to pull off, even if it’s a billionth of a penny. So an effective software platform lowers transaction costs. And a truly successful software platform lowers transaction costs as the number of transactions increases. This is called scaling.
Or in one-syllable words: If a thing costs less to do as folks use it more, that’s good.
If you discount the cost of marketing, it was far more expensive for Twitter to get its first million users than it would be for Twitter to add a million users today.
So what’s a product then?
A software product, or program, is just a thing that satisfies a want or need in its user. Many software products are made to let you access platforms. Your email client is one example. Apple News, or Slack on mobile and desktop. A successful software product encourages the growth of the host platform. They look good, feel good to use, and are fun.
Your bank has a platform. Their ATM is a product. When you use the ATM instead of a teller, it’s cheap for the bank. The more you use the ATM the cheaper it is.
Remember how an effective platform lowers costs as transactions increase? This means that a successful software product will lower transaction costs. If you have a good platform, a good app saves you money.
Or in one-syllable words: If it is fun to use a thing, folks use that thing more, and if a thing costs less to do as folks use it more, that’s good.
What about decentralization?
Since centralized platforms are so incredibly powerful and profitable, people are suspicious that these platforms will abuse the power they have, and point to examples of such abuse. Some people say: Let’s create decentralized alternatives. There have been many attempts to create more open, secure, and private versions of Twitter and Facebook, but none have taken off.
Rather than have one big Twitter, you could have everyone with their own little personal Twitter software and the various Twitters would federate together on the Internet. This actually could work! It’s technically feasible. The Internet works that way. But it involves lots of people installing software and working together and being good citizens. So it’s easy to get bogged down in the details.
People are totally willing to use decentralized software when it comes to pirating movies—BitTorrent is a decentralized protocol, for example—but less so when it comes to communication. Unless they could get arrested or sued for their behavior, people tend to prefer using well-maintained, well-designed, highly centralized services. Even when a service is decentralized, like email, we tend to use things like Gmail, which recentralize it, and block spam and abuse. Google profits accordingly.
One more thought
Some of the largest and most successful platforms are not-for-profit centralized platforms. For example, Wikipedia and Archive.org are highly centralized with reach that must now be in the billions of people. (Neither one has truly world-class mobile apps, though.) They are centralized and fairly virtuous, and their centralization is itself a virtue because it makes it easier for people to find and use their services.
The conclusion I draw from the above is: If you want the fastest possible global impact, and to compete with other centralized platforms, build a centralized platform with great apps, and understand your transactions, what they cost, and what can happen if you lower that cost.
This is not an argument against decentralization, because within ten years we may all be communicating via tiny network peer computers that cost $5 instead of going to centralized services for our Internet connectivity. It’s just an observation of what works today.
Paul Ford is a co-founder of Postlight.