Twitter’s video deals mean it’s giving up on business model innovation

So says Ben Thompson in his newsletter today:

This is why Twitter’s increased focus on securing these video deals feels like such an admission of failure: the company is basically admitting that, despite the fact it contains some of the best content — given to it for free — in the world, it simply can’t figure out how to make that into a business, so instead it is (presumably) paying to create content that it can monetize more easily. The Bloomberg deal, which was first reported on Sunday, is particularly poignant on this point: Twitter is (again, presumably) paying for content about business and financial markets even as the most valuable business and financial market information is being posted for free on Twitter. That the company cannot build a business on that fact is certainly a disappointment.

You’ll have to subscribe to read the rest. For $100 a year, it’s worth it.

Meanwhile, some stats from Sara Fischer at Axios:

In total, Twitter has closed over 40 live stream partnerships around the world with sports leagues, media companies, etc. The company increased live programming by 60% last quarter and aired roughly 800 hours of live content reaching 45 million viewers. Of those hours, 51% were sports, 35% were news and politics, and 14% were entertainment. Above all, Twitter says 55% of its unique viewers are under the age of 25, a stat that directly competes with Snapchat’s coveted millennial demographic.

There’s also an extensive list of the video partnerships and shows to be broadcast in Twitter.

Tips on using social media from analyzing how celebrities manage their brands

Some highlights from the article that seem to apply to any marketing use of social media:

  • “If staying on message is the first rule of corporate communications, it is also the cardinal sin of social media.”
  • Each medium has it’s own format and expectations: “corporations can and should differentiate their approach to each platform, digital-marketing experts say.”
  • “Instagram is stylish, behind the scenes” – well, for most of us, “stylish” won’t apply. But the “behind the scenes” part is interesting.
  • “Validate your followers with likes, comments and retweets. It builds goodwill.”
  • Frequent factotumia – “It’s about showing up every single day and showing pieces of their lives rather than when they have a premiere or something to promote.”
  • “Instead of trying to get followers to buy their product, companies can gently boost their brand by commenting on current events.”

Source: What Celebrities Can Teach Companies About Social Media

I’ve developed a little obsession with Twitter Analytics. It’s facinating to see all the stuff people do with my nonsense, and much more helpful than things like and SumAll.

The addition of the “impressions” metrics is the new thing – I’m not sure there was way of actually counting how many views each tweet got previously. It’s also interesting to see things like detail expands, emailing, etc.

The click-through rate seems pretty low for most of my tweets. The CoreOS one, above, is predictably high because it’s offering a free report.

Some Observations

I haven’t done a deeper analysis of what all the data means. For one thing, I’m not really sure what my goals are. However:

  1. Images work – Just for the “get more attention” metrics I have learned one thing: put images in your Tweet. People love images.

  2. The Tweet [is|can be] the post – Now that I know the “impressions” being tracked, I’m not so worried about people clicking through to my blog. I’m trying to think of how Twitter can be used as a “primary channel.” That is, the “end of the line” or final thing in a long trail of clicks. If I look at “firehose” tweeters like James Governor, I think they treat Twitter like this.

  3. Engagement? – Building on this, the “engagement” rate is a curious metric. It somehow summarizes “conversions” of tweets to clicks, replies, favorites, and follows. That is, how many people “did something” with this tweet other than viewing it? The screenshots above don’t list that, but the CoreOS one has an engagement rate of 7.7%, while the KACE one has a rate of 0.8% It’s probably a pretty good heuristic for sorting tweet popularity.

  4. Twitter is most likely my “front door” on the web – Long ago, my blog(s) were my primary source of (pardon the word) “engagement” with people. My RedMonk blog had over 1,7000 RSS subscribers at it’s apex, for example. The blogs I now have are pretty piss-poor. Interaction is tumblr is low as well (though the occasional like from Robert Brook is always the highlight of the day). In comparison, Twitter is much more of a front-door. The consequence is to focus more on #2 above.

Anyhow, I check Twitter Analytics all the time. It’s much easier to understand than the mess that is Google Analytics (where I also spend a lot of time but am never sure what’s happening). I’m interested to hear y’all’s feedback on how to use it and what it “means.” For starters, I have no relative idea of how my numbers compare to others.