Started thinking about two things yesterday:

1. The conversation Ian/Flockhart/Josh were having on Monday about how to decide what the "right" types of user behavior are and how to measure them.

2. On Twitter, Flockhart and Ian coexist with me and Josh — even though we both produce and consume totally differently.

This leads me to think that Twitter didn't decide that Josh's and my behavior was the "right" way to use Twitter and architect their product entirely around it. Instead, they built an interesting tweet/follow mechanism and let users hack it.

>>> TL;DR - Don't arbitrarily decide on a type of user behavior and build a framework around it. Build a compelling framework and let users behave how they want. <<<

Do you agree?

Yes, and that's what we've been doing thus far, I think. There are all types of branches: live branches, long-form, lots of people, two people, etc., etc. That's what makes Branch awesome and we shouldn't be prescriptive about a single use case.

That being said, I think whether a branch is "good" or "bad" is binary, in the sense that if it has less than two people and less than, say, 3 posts it likely didn't make the creator feel great.

It's a spectrum. It's not a "yes we are going to build mechanics and ignore potential behaviors" VS "no we are going to impose behaviors directly".

Josh — I actually think our existing product does impose a certain use case: that of long form, thoughtful, elite conversation. Again, not binary, but a spectrum, since there were people who "hacked" it and used it for other things, but the product is very much suited for long form. We don't force people to do so, but if they don't, it feels kind of off.

Back to my spectrum point: we're always making assumptions. I don't think it's possible to only build mechanics and NOT expect certain types of behaviors.

I agree that our UI, copy, etc. encourages certain types of conversations — the ones you allude to — but I think we should change that, to be more neutral, not double-down on any specific use case.

I don't think our definitions of 'specific use case' are the same. Is a use case "create branches" vs "discovery of branches"? or is a use case "having a casual GIF conversation with friends" vs "an intellectual, live, debate"?

The latter, for me. The former will develop over time. You can't build a publishing platform without focusing on the creators first, I don't think (per anecdotes from Ev and Karp).

But this is part of my point.

Ian has more than 5,000 Twitter followers and his last tweet wtas a retweet of an account called "IM A FUCKING CAT" that said:
"#SINGLEBECAUSE CONSTANT FEDORA USAGE, I HAVE A FEDORA THAT I WEAR TO SLEEP, IT'S MY SLEEP FEDORA."

Creators are gonna create, whether or not you think they should be having intellectual, live debates or trading GIFs with friends — and both things probably have an audience if we build mechanics they can both take advantage of.

Ian speak up plz

I don't think I understand your point then, Libby. What are you advocating for?

Twitter is actually very single use case. People who use the product and gather a large following base kind of have an issue where they gain popularity for a certain "style," and branching out of this style is impossible. @Fat and I always want each others follower base (I want his tech followers, he wants my weird twitter followers). But if I tweet something about tech it falls on deaf ears (other then the select group that crossover) and when he tweets weird shit about smoking weed it falls on deaf ears. Branch solves this use case by letting you follow groups rather than people.

The debate we have been discussing is about features which enforce these use cases, specifically "chatty" vs. "intelectual." Twitter is a poor example of this because it was always based around text messages. This forced it to be both short and realtime.

++ Ian.

imho even Twitter isn't that "neutral" when it comes to things that get attention (if that's a measure of what the platform is optimizing for). Popular tweets are either insightful or funny, and it's hard to be insightful or funny all the time.

The bar for posting stuff on Instagram that'd get likes seems way lower than Twitter. So maybe in a way it's more "neutral"?

im confused now. libby what are you trying to say?

re: ian's
"The debate we have been discussing is about features which enforce these use cases, specifically "chatty" vs. "intelectual.""

i think when there is a significant enough gain to one use case, we should do it. for example: live branching features that hursh built during the hackathon is awesome.

and yea i don't think twitter is a good parallel. it is neither at the complexity of a product that we are, nor is it at the early-stage that we are. we should look at complex systems like FB in its infancy.

I AM SORRY. Guess that post was only a quarter baked.

I misread what you said. @Josh - I thought in response to @Julius you were advocating for "the latter" — but between choices of "a casual GIF conversation with friends" and "intellectual, live, debate."

Separately, re: Twitter not being a good parallel, I'm not so sure. @Ian - I agree 100% with your point about silos. Con: You're locked into your personas. Pro: You both still feel like you are "doing Twitter right."

On Branch, @Ian, you and Fat should be able to branch both about your weird stuff and your tech stuff and feel like you are "doing Branch right." (So +1 to Julius and Josh's moving toward neutrality.)

@Julius - Good point about Facebook. (And I'm guilty of oversimplifying about Twitter for the sake of being controversial here...!) What do you guys think Facebook did right/wrong?

Fb had very little complexity at the start. It was profiles. Then at some point walls. Pokes. Groups. Feeds. Pages. Apps.... But it was fairly straight forward I think. Now it's really bloated.