April 17, 2013 at 6:48 PM ET
Ads on Twitter will soon appear based on keywords in your tweets, such as where you're shopping, what you're looking for. It makes for highly targeted advertising — but some may find it a little creepy.
The new feature, announced Wednesday in a blog post, is just the latest by Twitter in a push to made ads more appealing to users and attractive to advertisers. Last year saw the categorization of users based on their tweets, so someone who tweeted about coffee would be put in a "coffee lover" group. But this is far more specific.
Businesses can pay to have Twitter watch for certain words, then display a promoted tweet when one is detected. So someone tweeting about craving an apple won't just be put in the "produce lover" category for later reference, but may actually then see a tweet promoting a local farmer's market.
Natural language processing will prevent ads from appearing when someone is tweeting negatively — so complaining about a TV show, for instance, won't get you inundated by ads for that same show.
Some may think it's a bit much to have such specific and timely targeting, but Twitter believes it will make for a better ad experience. The information was already being collected; it's just being put to use faster.
"Users won’t see any difference in their use of Twitter — we’re not showing ads more frequently in timelines, and users can still dismiss Promoted Tweets they don’t find relevant," wrote Nipoon Malhotra, Twitter's product manager for revenue, in the blog post.
Devin Coldewey is a contributing writer for NBC News Digital. His personal website is coldewey.cc.