4 ms·
The article posits that instead of trying to focus on display ads and impressions, they should have focused on sentiment analysis, data mining, trend forecastin
by niftich 10y ago
The article posits that instead of trying to focus on display ads and impressions, they should have focused on sentiment analysis, data mining, trend forecasting as avenues of monetization. Shazam, the app that can recognize a song from a recording, did this [1], and they can predict what songs are going to become really popular based on their users' interest.
Of course, actually doing this in 2010 would have been extremely ambitious, given that sentiment analysis was an up-and-coming field having recently overtaken by extreme hype (just look at the edit wars on the Wikipedia article [2] in their effort to revert marketing drivel). It was a far more sure bet at the time to pivot towards display ads and a more locked-down ecosystem, and about new user onboarding [7] and content curation [8], than the even more vague promise of being able to mine useful data out of aggregates of Tweets (preferably in near-real-time).
This particular bet may have been the wrong one, as the next year academic papers were coming out by the dozens about sentiment analysis on Twitter; by this time the company had doubled down on display ads and keeping more users inside the application with embedded videos, a captive shortener, a captive image host, the acquisition and relaunch of native mobile clients, and the like.
After yet another year, in 2012 Facebook smoked them with their ad onslaught [3] and also bought Instagram for good measure. Twitter acquired Crashlytics in 2013 and then the ad network MoPub and reworked it into an embeddable Twitter SDK that app makers would want to use [4], then bought a machine learning startup and I finally was sure the omnipresent sentiment mining Twitter would materialize [5] but this ship sailed at last when they sold most of that to Google after all [6] to focus on their 'core product', which isn't growing... a problem the article, despite its title, doesn't offer a solution for.
[1] https://news.ycombinator.com/item?id=12083782 https://news.ycombinator.com/item?id=12083782 [2] https://en.wikipedia.org/w/index.php?title=Sentiment_analysis&dir=prev&limit=500&action=history https://en.wikipedia.org/w/index.php?title=Sentiment_analysi... [3] https://www.theatlantic.com/technology/archive/2012/12/2012-year-facebook-finally-tried-make-some-money/320493/ https://www.theatlantic.com/technology/archive/2012/12/2012-... [4] https://news.ycombinator.com/item?id=11914620 https://news.ycombinator.com/item?id=11914620 [5] https://news.ycombinator.com/item?id=11942293 https://news.ycombinator.com/item?id=11942293 [6] https://news.ycombinator.com/item?id=13429179 https://news.ycombinator.com/item?id=13429179 [7] https://thenextweb.com/twitter/2011/12/08/a-new-version-of-twitter-is-coming-and-you-can-get-it-now/ https://thenextweb.com/twitter/2011/12/08/a-new-version-of-t... [8] https://thenextweb.com/twitter/2011/12/08/why-twitters-discover-feature-is-a-game-changer/ https://thenextweb.com/twitter/2011/12/08/why-twitters-disco...