Bad reviews. On September 30th, the soon-to-be-launched app Peeple was described by the app’s co-founder, Julia Cordray, as a “Yelp for people,” that is, a people-rating platform that would allow users to assign number ratings to anyone—anyone at all, fellow Peeple user or not—as long as the critic was 21, had an established Facebook account, used his real name, and could provide his subject’s cell phone number. The Washington Post reported that “you can’t opt out—once someone puts your name in the Peeple system, it’s there unless you violate the site’s terms of service. And you can’t delete bad or biased reviews—that would defeat the whole purpose.” But by October 4th Cordray had changed her tune entirely, describing Peeple as a “positive only app” that will not allow a review to appear until it has been explicitly approved by the subject of the review. An outraged public can take credit for Cordray’s about-face. In the comments section of the Washington Post piece in which Cordray described Peeple’s original iteration, many readers expressed fear and disgust at the prospect of a site that would doubtless facilitate cyberbullying and, for some, encourage crippling self-consciousness. In other words, the very online lynch mob whose power Cordray hoped to profit from wound up turning on her. The irony of that did not go unnoticed. Cordray’s new version of Peeple, which she promises will be part of a “positive revolution” designed to fight back against the kind of online negativity she faced after the Washington Post piece was published, is expected to launch before the end of the year.

Picture perfect. Any social media marketing plan necessarily includes these two goals: Identify your target audience, and attract and engage that audience with compelling content. We’ve written about how neural networks are facilitating marketers’ use of photographs to help find target audiences by automatically identifying the products being used and the activities being undertaken by the subjects of photos posted to social media. Now, companies like EyeEm Mobile GmbH and Neon Labs Inc. are rolling out software to help marketers choose photographs that will engage those audiences with attractive content. The technology that these companies have developed automatically identifies images with qualities that potential customers are likely to find compelling. The software, according to the Wall Street Journal, identifies “patterns common to images selected by professional photographers” or “characteristics shown to trigger brain activity in neuroscience experiments.” The human eye has proven to be more attracted to photographs in which the subject appears to be speaking in mid-sentence, for example, or in which the subject appears to be attracted to something happening offscreen. Since EyeEm debuted two months ago, the startup’s customers have enjoyed a 30% increase in interaction with their online content. The software works for online video posts, too, helping companies to select the most compelling frames for posting. Neon Labs reports that its partners and customers “made 16% to 40% more revenue on their videos due to increased clicks.”

The powers that tweet. Twitter has a handbook explaining its platform to people running for elected office and it runs just under 140 characters pages. That’s right, the social media company notorious for its draconian limit on post lengths took 136 pages to “make sure that people feel empowered with the full story of what Twitter is,” according to NPR’s interview with Bridget Coyne, one of the people responsible for the manual’s compilation. Coyne told NPR that the creators of the “wildly popular” guide went into painstaking detail explaining how the platform works because they “didn’t want to make assumptions.” Still, it’s difficult to comprehend how the manual’s target audience—whom Coyne says includes congresspersons, chiefs of staff and Congressional interns—aren’t likely to be insulted by a manual that dedicates whole pages to topics such as, for example, “The Anatomy of a Tweet,” complete with numbered illustrations pointing out which part of a Twitter post constitutes the “profile photo” and which part is defined as the “tweet text.” Nevertheless, Coyne insists that Twitter continues to get positive feedback about the manual, and says she hopes to eventually publish a second version of it explaining new Twitter features like Periscope. Does all this mean an end to political scandals due to Twitter misuse?