Interesting.

I like that she calls out Twitter for having no product innovation at all.

Indeed, like she says, Twitter has not demonstrated much in the way of internally sourced innovation, no. Users have dictated how the service is used, and Twitter has codified those behaviors once they are entrenched.

She also gets that it's Twitter's inconsistent treatment of devs which has been problematic. Facebook has consistently kept devs at arms length. We loved Twitter because they seemed to be so dev-friendly in the beginning. Shit, even at the Chirp conference they seemed dev-friendly compared to now. Remember how Annotations were going to be so awesome?

Ed, I do remember that!

And then there was a Twitter regime change.

And Twitter buried Annotations and redefined what they wanted from developers, like something out of 1984.

We developers are a crabby bunch.

I think you're right that it's the inconsistency that gets our goat more than keeping us at arms length.

This is so right on:

If Twitter can actually turn this legacy into a successful, lasting public company, if it doesn’t flame out and become the tale of what could have been, we will have witnessed a singular event in the history of Silicon Valley...

And so is this:

Facebook was always an outlier — a once in a decade success story by a young kid who came out of nowhere and built something huge. But while rare, we’ve seen those sorts of outliers before. Twitter is something even more uncommon: A company that has violated nearly every spoken or unspoken rule for how you build a successful company in the Valley. You hear it muttered everywhere in startup circles these days: This company should never have made it this far. And yet, here it is...

Expectations matter:

Had the private markets valued them more rationally, we’d be marveling at the size of Facebook’s market cap — even 50 percent lower than at its debut. We’d be stunned that something as seemingly frivolous as the many iterations of Farmville was worth anything close to the market cap of EA. And the fact that anyone in the hard to win local space had created any company worth $2 billion in just a few years would be impressive. These three would be successes, not failures...

I use Twitter so much less now than even a year ago, for three reasons:

1. The content is mostly unintelligent noise. The gems are few nod far between.

2. The API rate limiting discourages counting on Twitter for anything.

3. The Web client is slow and often locks up or crashes my browser.

I love that Twitter, unlike Facebook, gives me the pure, firehose feed of everyone I follow... And I hate that Twitter, unlike Facebook, gives me the pure, firehose feed of everyone I follow.

I'd really like an algorithm in the middle to cull the crap -- just not one with Facebook's agenda.

What percentage of Twitter tweets do you think are worth reading? 10%?

I believe I remember seeing an analysis from Gnip or some other social analytics company that only 2-3% of tweets are "real content of consequence."

Twitter is an information slot machine. Presentational gimmicks make even mundane, repetitive things seem extra-interesting -- for example, by interleaving unrelated items, or obscuring necessary context, or using attention-grabbing interruptions, or imposing a urgency-faking 'decay' (towards the bottom of your attentional window).

This is not healthy, even though other aspects of the Twitter medium are transformational.

Twitter's choice of an advertising and control-the-client model makes the necessary innovation -- innovation which actually *makes Twitter less abusively compelling* -- less likely. In fact, the latest rules about Twitter display seem to preclude exactly the sorts of auto-filtering/grouping/folding/reordering that could make Twitter a healthier medium.

Just as DVRs redeem the boob tube, Twitter needs a Tivo. But Twitter's attention-hungry revenue model makes a 'Tweevo' illegal.

A study said 36% of tweets were worth reading. Personally, I feel it's far lower than that.

Eric, I conur: no way Twitter is 36% useful. mko's estimate of 2-3% seems right.

I love Gordon's idea of "TiVo for Twitter". In theory that's what Twitter's summary emails are about.