Monthly Archives: October 2012

How You Use Twitter Can Prevent You From Ever Working There

In years past, I noticed some Twitter employees didn’t even have a Twitter account before they were hired.  No longer.

Now potential hires have their Twitter accounts carefully studied by CEO Dick Costolo’s staff before they are hired, to make sure it meets their criteria.

Only started your Twitter account in the last few weeks? Don’t expect a job offer, because it looks suspiciously like you might have set up your account just to try to get the job. Costolo and staff want to see Twitter accounts that are “truly embracing the organisation and its ethos.”

“I think it’s great [to] use that as a measurement of how suited they are to Twitter as a company,” Costolo enthused at the Internet Advertising Bureau’s Engage conference in London on Thursday.

HT @ej_butler 

How Twitter’s Worldwide Outage Changed Klout Scores Worldwide Twice

On July 27, 2012 many folks had Klout scores that dropped dramatically.

The day before and the day after were not affected. This was the day of Twitter’s dramatic worldwide outage. On that day, Klout lost around 20% of the data it needed to calculate Klout scores accurately.

How Does Twitter Affect Klout Scores?

Klout measures a person’s “influence” each day, and then publishes an overall score that is the average of the last 90 days. The overall average is the number they publish. It’s your “Klout score.”

That means that a score from 90 days ago won’t count towards your score tomorrow. So if you had a really high score 90 days ago, it will go away tomorrow, lowering your overall score tomorrow (unless tomorrow is equally high.)

Klout’s “dashboard” shows how much each signal affects your Klout score. In the case of our @TweetSmarter account, Twitter accounts for just over 90% of our score.

How Much Will Klout Scores Increase?

On Thursday, October 25, the day of the Twitter outage was 91 days old, and no longer counted towards anyone’s Klout scores. When that day was no longer counted, scores went up around .2% for users whose Klout score was influenced heavily by Twitter.

Since Klout averages to only report whole numbers, scores that were just below a .5 (such as 33.4, 41.3, etc.) saw their score increase a whole number, as they pass over the .5 mark and are averaged up to the next whole number.

A very rough guess is that about 1 in 700 Klout scores went up…affecting hundreds of thousands of users.

HT to @Karen_C_Wilson for some proofreading

Twitter Bought 18 Companies—Do You Know Why?

Twitter never stops changing, it seems.

And when they acquire a company and/or its assets or employees, you can often predict that whatever that company does will be involved in upcoming changes to Twitter.

Early changes were to improve things for users, but more recent changes have been more about business, branding and advertising enhancements. (For example, recently Twitter announced that it would be controlling ads that run through Twitter more than ever before by changes to its API use guidelines.)

Here are all the companies Twitter has ever acquired, in one graphic, from a report by JEGI (click to enlarge):

Twitter’s first purchase, Summize, was a Tweet search engine, leading to Twitter’s own search engine.

Twitter’s two purchases this week were Cabana for mobile apps, and Vine for video sharing.

Have Millions of Fake Accounts Fooled Twitter?

Twitter CEO Dick Costolo has mentioned several times that millions of people use Twitter only for reading information, and rarely if ever tweet themselves.

But how much does he know?

Fake Twitter Accounts

Companies that sell followers create fake Twitter accounts that rarely tweet, and just have them follow people for money.

So the fact that there are millions of Twitter accounts that rarely tweet doesn’t mean those are people that use Twitter passively for reading without tweeting (much) themselves. Many—most?—of them are fake accounts being sold to people who want to buy followers.

Of course, some of them are real people, but Twitter can’t tell the difference! Twitter hasn’t figured out how to stop fake accounts from being created, and can’t tell the difference between them and real people in order to suspend them

So how many people are just using Twitter “passively?” No one knows…except perhaps the spammers creating fake accounts. But since automated hacking tools are widely available for creating fake Twitter acc0unts, and it’s estimated that selling Twitter followers brings in nearly $50 million each year, it’s fair to say there could be millions of fake accounts.


Of course, some fake accounts are created in order to send spam. Here’s a list of the latest spam, and spam accounts, for example. But accounts created only to be sold as fake “followers” never spam, because that would risk them getting suspended, and if they were suspended, they couldn’t be sold as followers.

Fortunately, even if hackers find a lucrative way to break into your account or computer just by getting you to click a link, those fake accounts are unlikely to switch over to sending spam, because they make a LOT of money being sold as followers.

And that’s good, because otherwise, millions of fake accounts could suddenly start spamming Twitter.