facebook

The results in summary according to Digital Trends:

Following up from the Spring 2014 ‘Taking Stock With Teens’ study created by Piper Jaffray, the Fall 2014 edition of the study was published this week with a particularly harsh outlook for social networking giant Facebook. When teens were asked what social network they typically use, only 45 percent responded with Facebook. That’s down from 72 percent responding Facebook just six months ago.

Alternatively, Instagram grew in popularity with 76 percent responding in the affirmative. In addition, sites like Twitter, Pinterest, Tumblr and Reddit pulled in similar numbers as the last study. Only Google+ plummeted with Facebook, dropping from 21 percent in Spring 2014 to just 12 percent in the Fall study.

The story makes it seem like Facebook is on the way out but I have a different take. It’s not helpful to view any particular app from a “one to rule it all” perspective — though it may have started that way — because people have learned to use each service in a different way. Facebook doesn’t compete directly with Instagram even though both are on the surface social networks.

Read Full Article

There’s a social network called Ello that’s buzzing. I’m not sure why.

Ostensibly, it’s because it’s seeing rapid growth, though no one can cite actual numbers. The narrative is that some are unhappy of Facebook requiring real identities, and so are migrating en masse; mostly from the LGBT community according to the Washington Post. People are migrating to Ello because, unlike Facebook, the social network doesn’t require real names.

Hmmm.

You know what other social networks don’t require real names? Google Plus, Twitter and Instagram. Which are just a little more well known than Ello. The point is that if all people wanted is a social network which doesn’t require real names, there are plenty available.

Read Full Article

A couple days ago, we referenced a comScore study on mobile apps. Since then, more articles about its findings are surfacing. Quartz has an overview of the top 25 most used apps by US consumers. Highlights:

  • Facebook is unsurprisingly the #1 most used app
  • Pandora is surprisingly #5…no Spotify
  • Google is the top mobile app publisher
  • Facebook Messenger is the top messaging app, ahead of Snapchat, Skype and Kik…no Whatsapp, no Google Hangouts
  • No games made the overall top 25

The study also breaks down popularity by age segment, which The Atlantic graciously provided. Highlights:

  • Facebook, Youtube and Pandora are universally popular
  • The younger, the more popular is Instagram
  • Older folk use Facebook Messenger more than Snapchat or Kik
  • Email didn’t make the top 10 for 18-24 year olds

Read Full Article

Facebook is in the press lately because the company is forcing users to communicate with friends through a separate Messenger app instead of the main Facebook app. The former is #1 on the app store but people are rebelling by slamming it with 1-star reviews. Privacy is also a common rallying cry, the accusation being that the app is too aggressive with permissions.

Screen Shot 2014-08-08 at 11.29.51 AM

Nobody likes to be forced to do anything, so it’s not surprising to see people react negatively. Many are still wary about online privacy, even as it is an increasingly illusory concept.

We think those knee jerk reactions are overblown. Facebook Messenger is a decent product and it’s no more aggressive in its privacy policies as other messaging apps.

Read Full Article

Instagram is testing their new app Bolt in Singapore, New Zealand and South Africa for iOS and Android, and I got my grubby hands on it.  Bolt is one-tap photo messaging, like a simpler Snapchat.  And you know what…I kinda like it, though I don’t know how defensible it would be as a business.

It’s attractive and single-minded.  There’s no confusion to the app’s purpose.  One tap to send a photo or video might not seem much different than say, Snapchat’s three taps, but surprisingly it’s increased the number of messages I usually send.  Maybe that’s because testing Bolt is top-of-mind, but I don’t think so.

One tap messaging feels so effortless — it’s easier than typing a text and yet can say so much more.  Sending dozens of photos and videos throughout the day to let someone know how things are going doesn’t feel like work.  With Bolt, it feels natural.

Read Full Article

“You have to be in mobile,” venture capitalists will instruct you.  “We have to be in mobile,” tech giants like Facebook, Microsoft and Yahoo will echo.  Obviously, it’s because we’re living in a mobile-first world.

Yet, becoming the next Instagram or Snapchat is insanely difficult and increasingly so.  A hit app is a rarer unicorn than a hit website.  Why?  Unlike websites, apps aren’t linked to each other; you can’t click on a link to discover a new app, you have to purposefully search and download it from the app store.  Then you have to learn how to use the app before finally getting some value out of it.  Some apps — especially on Android but even from bluebloods like Facebook — behave badly and mistreat your phone’s battery or privacy settings.

The result of the above is that, according to Nielsen, most people use only 30 apps on their phone.  What are the chances your mobile app can make a person’s top 30?  Let’s break down how difficult a threshold that truly is.  What are the 30 apps you’d typically use?

Read Full Article

Twitter has a new analytics tool that lets you see how many people actually saw your tweets.  Danny Sullivan of Marketing Land reported that of his 390,000 followers, only 1.85% saw (not even necessarily read) a random tweet.  Ouch!

When we started feecha, one of the early decisions was what data to show users.  To be like YouTube, where you can see how many viewed a video; or to be like Instagram, where you have no idea how many actually saw a photo.

We decided to go the Instagram route because content creators need to feel like they’re being read to continue; once that illusion is gone, only the most strong-willed can keep going.  In the beginning, and especially with all that’s out there, most new start-ups will struggle getting engagement from users.  When the crowd is sparse, no news is better than bad news.  It’s better to keep your users guessing on how many they’re actually speaking to.

So why did Twitter make this kind of analytics tool available?  Advertising. I find advertisers’ interest particularly ironic; if they only knew just how many really saw their ads on TV or print…  390,000 Twitter followers sounds amazing, but it’s the 7,215 that actually matters.  I wouldn’t be surprised if the 1.85% seen ratio applies to TV ratings and print circulation numbers too.