Twitter Verification: A Case Study in Product Management Failure and How to Fix it

The Problem with Verified Users on Twitter

Recently Twitter got a lot of heat for verifying the account of Jason Kessler, organizer of the Unite the Right rally in Charlottesville, Virginia, that resulted in Heather Heyer’s death.

In a rare example of self awareness, Twitter’s official support account acknowledged that part of the outrage was due to the fact that Twitter had caused confusion by treating verification as both an indicator used to authenticate an identity and an indication of celebrity status on the platform.

This is an understandable logical error for the product team to have made. Basically

1. Verified users are usually celebrities who you want to authenticate as the real thing.
2. Celebrities are special users who deserve perks to encourage them to use and endorse your platform
3. Verified users are special users who deserve perks

Twitter started making this mistake as far back as in 2014, when they first shipped special features for verified users. This was clearly an unintentional mistake and it was a good thing for Twitter to acknowledge this confusion they’d caused although a few years too late.

Twitter’s About Face on Verification Being a Celebrity Perk

However this moment of insight and clarity from the Twitter team was short lived. Yesterday the official Twitter support account posted the following

This is an odd about face since it implies that Twitter is effectively acknowledging that being verified is a reward and they are going to take away this perk if you disobey their rules around civil behavior. But that means Twitter is acknowledging that it gave a perk to known Nazis and white supremacists like Jason Kessler and Richard Spencer which is an odd message to send. That is extremely odd for them to admit out loud.

More importantly, Twitter is also conceding that they don’t have a solution for authenticating users. Shouldn’t I be able to tell that @SixGodDrake is fake and @Drake is the real account even if Drake writes mean tweets that cause him to violate Twitter’s rules around civility?

Is Twitter going to take away @realDonaldTrump’s verification because he posted mean tweets about Hillary Clinton? Where do they draw the line? Why are some offenses bad enough that you lose verification but not bad enough that you’re kicked off the service?

Twitter went from acknowledging that they’d generated confusion around what verification means on the site to sowing even more confusion in a handful of days.

Product managers are supposed to create clarity. In Twitter’s case, the product team has actually generated confusion.

Fixing Twitter Verification

With that said, Twitter does have real problems that require a workable verification feature.

The first step in fixing Twitter’s verification feature would be to concede that there are actually two different scenarios that have been conflated into a single feature

1. Showing that a tweet from a newsworthy person is actually coming from that person.
2. Providing an extra tier of recognition and functionality for favored users (i.e. celebrities).

For the first, Twitter should democratize and standardize the process for getting verified with objective criteria. Some sample criteria could include anyone with a Wikipedia entry that has existed for at least 6 months (to avoid people creating Wikipedia entries to get verified) or anyone who’s been mentioned in a news story from a list of trusted news sources (to avoid people being referenced in their friend’s blog claiming they are newsworthy). These are just examples and one can imagine other objective criteria that could be used. The important thing being that they are objective and open to all who are newsworthy.

In addition, when a user is verified, it should be clear from their profile why they were verified including a reference to their newsworthiness(e.g. link to Wikipedia entry). Since users can change their display name it’s not unusual to find a verified account with a cryptic user name and a fun display name with no idea who it is or why they’re verified.

For the second item, Twitter needs to introduce a new concept for “special” users. Facebook already has shown you can have users, brands (i.e. Pages) and verified users or brands coexisting without users being confused.

With these simple changes, verification on Twitter would go from a cause of confusion and angst to a source of clarity and understanding.

Now Playing: 21 Savage & Offset — Ghostface Killers