Let’s Compete with Facebookby@arnoldkling
911 reads
911 reads

Let’s Compete with Facebook

by Arnold KlingApril 10th, 2018
Read on Terminal Reader
Read this story w/o Javascript
tldt arrow

Too Long; Didn't Read

I am sick of reading about people who want to regulate Facebook. You didn’t come up with the idea. You didn’t build the business. Now that it’s here, who the heck do you think you are telling them how to run it?
featured image - Let’s Compete with Facebook
Arnold Kling HackerNoon profile picture

I am sick of reading about people who want to regulate Facebook. You didn’t come up with the idea. You didn’t build the business. Now that it’s here, who the heck do you think you are telling them how to run it?

It’s not that I’m happy with Facebook. Far from it. But to me, the best way to fix it would be to come up with something better. I figure that if we really do come up with a much better way of running a social network, then some entrepreneur will be able to make a success out of our idea.

I’ll start with my pet peeve about Facebook.

The Stupid Algorithm

My problem with the infamous algorithm that governs the news feed is not that it’s mysterious or that it’s malicious. My problem is that it’s just unforgivably stupid.

— It shares jokes from friends who don’t have my sense of humor.

— It shares political opinions from friends who aren’t insightful.

— It shares life events from friends with whom I have no personal relationship.

— It gives me frequent updates from people who I only want to hear about occasionally.

— It misses updates from people who I really care about.

Toward a smarter algorithm

With our Facebook competitor, our algorithm will give users much better control. When John and I link up as friends, I will be able to indicate how often I want to hear various sorts of news from John. As John and I link up, I will consider a list of topics:

— items that pertain to me as an individual, such as comments on my writing

— jokes or amusing videos

— food

— entertainment

— news related to a specific hobby or interest (one of the interests could be politics, or a particular sub-field of politics; I would have specified my interests when I created my basic profile)

— news related to work

— news related to friends we have in common

— news related to an affiliation we have in common, such as a previous school or employer

— major life events

— minor life events

— etc.

For each category, I can say whether I want to see what John posts in that category.

Whenever anyone posts something, they check boxes indicating what it pertains to. The algorithm feeds me news by comparing what gets checked by my friends with what I have said I am interested in from those friends. I think that this would eliminate much of the stupidity that bothers me about the current algorithm.

Suppose that John posts something really interesting about politics, but I have turned politics off as a category for John. It still might get through to me if it gets shared through the network to another one of my friends whose political posts interest me. Otherwise, I might never see the post. I am willing to take that chance.

An alternative implementation

It might not be optimal for us to impose a fixed set of categories on users. An alternative would be to let the categories emerge organically through a hashtag mechanism.

With this approach, we do not start with a pre-determined set of categories. Instead, categories emerge from hashtags as people write posts. So if I write a post about folk dancing activities, I include a hash tag “#folkdancing.” Then, when you connect with me as a friend, you look at the list of all the hash tags that I have used. You can decide whether going forward you want to see my #folkdancing posts or not. Instead, you might only want to see my #economics posts.

I can imagine that it would be easier to use fixed categories to start with. But over time, the hashtag approach would turn out to be more robust.

The Business Model

As plenty of commentators have pointed out, Facebook’s reliance on advertising creates adverse incentives. When Facebook and its advertisers study your behavior, it is not with the goal of improving your experience as a user. The goal is to make ads more effective.

Instead, our competitor will use a “freemium” model. Our money will come from subscriptions, but anyone may join for free. People who do not subscribe will have limits on the features that they may use. For example, they might be limited to, say, 5 posts per month.

With the “freemium” model, our incentive is to create the best experience possible for subscribers, while making our service at least minimally attractive to everyone. Unlike the advertising model, this aligns the interests of our business with those of our users.

What do you think?

These ideas may be flawed. They may very well be even worse than Facebook as it currently exists. But I’ll bet that somebody can come up with better ideas that work. My point is that I would prefer to see such ideas tested in the market, rather than imposed by a regulator.