I guess hackers think Twitter users are a gullible bunch. They’ve just started a viral scam which promises to get Justin Bieber to follow you. But this time, the Twitter team cracked down quickly.
Read more

There might be no legal repercussions heading his way, but one Twitter spammer is going to have one heck of a headache today, if an angry mob of Twitter elites has anything to do with it. His “app” – which promised to show you how many hours you spent on Twitter but was really used to send nefarious tweets from unwitting accounts – targeted a big-name superstar in social media, and he unleashed the hounds.
Read more

We’ve all seen it. Hijacked Twitter trends that display tweet upon tweet that claim to let you “find out who your top Twitter stalkers are!” and want you to “check out this CRAZY YouTube video.” These tweets are often jam-packed with four to eight different trending topics at once, and include the requisite link to spam, porn, or a virus. But there is a way around all of this spam – Clean Tweets.
Read more

Search Engine Land editor-in-chief Danny Sullivan (@dannysullivan) has been attending his own Search Marketing Expo (SMX) in San Jose, California, and Twitter’s Trust & Safety team lead Del Harvey (@delbius) was the key speaker in a section entitled, Don’t Be A Twitter Spammer.

Sullivan is moderating the piece and has been live-tweeting Harvey’s comments, which are interesting, provocative and, in a few cases, more than a little controversial.

Some of Delbius’s tips are well-known and even obvious, but others are definitely new statements being made by Twitter, and there are items here that require further debate and investigation.

Don’t Be A Twitter Spammer

1. Twitter doesn’t try to judge content but if it observes a given user being blocked repeatedly, the spam team may step in to investigate further
2. Avoid using multiple accounts to send the exact same message
3. Don’t spam hashtags
4. Don’t use misleading links
5. Avoid reply spam
6. Don’t churn – manually or especially through the use of apps
7. While automated DMs aren’t expressly forbidden, Twitter recommends avoiding them because they’re unpopular and lead to complaints, which results in Twitter stepping in
8. Sponsored tweets must be disclosed by US law
9. If you’ve been unfairly tagged as a spammer, ask for help. (“I’m super into forgiveness,” says Delbius, as the medium is so new)
10. Avoid doing lots of old-style, manual retweets – the main reason Twitter’s internal retweet system was created was because of so many complaints about fake retweets (which, admittedly, are an ongoing problem), and it can appear to be reply spam to Twitter

This final tip is by far and away the most controversial (UPDATED: see below). Like many users, I’m a huge fan of old-style retweets because they allow you to add your own flavour and commentary. I do retweet using the new method from time to time, but only if the content is perfect as-is, which to be honest is pretty rare.

The cynic in me wonders if this statement has been made less out of truth and actual, bonafide advice, and more because Twitter is frustrated that so many users loathe and (in an awful lot of cases) completely avoid the new-style retweet. And so by introducing an element of risk they might hope to move people over in a shorter period of time. To be safe it might pay to be a little wary, although you’d think if they were really concerned they’d compromise and introduce optional annotation. But that would mean admitting they made a mistake – humility hasn’t been Twitter’s most well-played card.

On the upside, Delbius advises asking questions and retweeting items of interest to your followers, and generally being authentic. These are all plusses in the eyes of Twitter, particularly their spam team.

Also, curiously, Delbius mentioned the Miracle Whip (@miraclewhip) Twitter account as an example of an entity that is doing it right. Which, on an initial inspection, seems a little strange. But at least she didn’t go with the obvious. Or Charlie Sheen.

Lots to learn from here, and in our few interactions I’ve always found Delbius very reliable, but it might be worth taking some of these statements with a pinch of salt until they’re backed up with similar remarks from other members of the Twitter elite.

UPDATE: @Delbius has reached out to me on Twitter and clarified my concerns about the old-style retweets, as follows:

Something about it didn’t sound right, and I’m glad we’ve got that straightened out. The rest of her presentation is definitely worth taking on board. Unless, of course, you are a spammer, in which case you won’t care, and probably won’t be reading this.

(Hat tips: Danny Sullivan, Del Harvey.)

There are 5 buzzwords that are almost guaranteed to get you spam Tweets; these keywords are like fish bait for fake corporate profiles. Fake Tweeters will search for Tweets that contain words that relate to their business. And they will follow you in hopes that you follow back – allowing them to spam you with links to their own websites.
Read more

Over on the official Twitter blog, there’s an update of a new way that all of us can help in the mounting battle against spam on the network. Twitter has added a one-click way to report any user as a spammer – which is now visible on everybody’s profile – and once done their ‘Trust and Safety’ team will move into action, investigate the alleged nuisance and act accordingly.

Also, once a profile is reported, it’s also automatically blocked.

It’s important that a team is involved and that this isn’t in any way automated as this naturally limits the impact of false accusations and, as Twitter states, cannot be used to “incite an angry mob” against an account.

Previously, the only way to alert Twitter of a spammer account was to send a message to the official @spam account. I assume this will still be an option. And while this one-click, take care of everything approach is a step forward, it’s likely to only have a marginal impact against the growing spam problem.

What the network really needs is an Akismet or Gmail-style filter system that is configurable at the user level so that it learns. Because Twitter is an open network, and anybody can contact anybody else, whether they’re following each other or not, the biggest concern on Twitter right now is reply spam. (Closely followed by abuse of trending topics.)

I also think we need a stricter sign-up process, possibly via a subscription model. Because it’s so easy to set up a disposable Twitter account using a disposable email address, blocking one user doesn’t mean the end of your problems, as another one is right around the corner. And another one. And another one. All managed by the same despicable teams. This is a proactive move by Twitter, but because it doesn’t really tackle these existing concerns, the benefits are likely to be negligible.

Throughout the Twitter week I get lots of rogue messages from spambots and other ne’er-do-wells who I then immediately block. You know, this kind of thing:

This charming individual illustrates exactly why replies are far more of a threat to your Twitter experience than direct messages. (At least it wasn’t to the level of what happened to me previously.)

What I thought would be a neat implementation from Twitter would be a facility that alerted all the members of my network each time I – or anybody else in that network – blocked somebody, and more importantly why.

This would have to be opt-in, as not everybody cares. But perhaps when you block somebody Twitter should ask you for a reason. It could be a drop down list of choices (spammer, retweet bot, etc) and an ‘other’ option where you could wax a little more lyrically.

When this was completed, everybody who ticked the box within that network would be sent a direct message saying

@username just blocked @troublemaker because reason

For example:

@sheamus just blocked @BeverleyBestg because "it's a spambot."

You could then click on the person I’ve blocked, check them out yourself, and block where necessary.

I could do this manually, but sometimes publically stating why you’ve blocked somebody is not always appropriate, and not everybody will care, as said. And publishing a big list of blocked users is of no interest to anybody but the person who created it.

Moreover, by actually asking us for a reason when we block, Twitter could get a rough indication of problem areas within the network. We know spam is already a big issue, but there’s no real indication how much of a problem trolls are on Twitter. Or stalkers. Or good, old-fashioned weirdos. Many of us have had bad experiences with individuals on other networks – that knowledge could be passed over to everybody else.

Of course, it’s all relative. One man’s guru is another man’s con artist. And likely there will be cases where somebody would use the block system to defame another’s good reputation. And that’s why it’s always important that you check these things out yourself before deciding to make what should be a fairly considered decision to block somebody.

My friend Mark has a neat idea.

Reply spam on Twitter is on the up and because anybody can send a (public) message to anybody else on Twitter (whether you’re following each other or not) there’s not much you can do about it until it happens. As Twitter expands, this is going to become more and more of an issue. A simple security measure that asked new users to authorise their tweets could all but end this problem overnight.

Yes, having to enter CAPTCHA information all the time can be really annoying, particularly if you’re a seasoned user (I’m looking at you, Facebook) but if Twitter implemented this as a requirement for brand-new accounts then you could essentially eliminate reply spam overnight (certainly from bots).

It wouldn’t even need to work on a number of people basis, as Mark suggests. If each new Twitter user had to complete a CAPTCHA for their first ten tweets and/or direct messages, probably 99% of all spambots would be nullified.

Once you’ve satisfied this trial period, no more CAPTCHA for you.

Aside from the marginal annoyance for newbies, can you think of any good reason why this wouldn’t work?

(Another solution would be for Twitter to verify all accounts, and not just a smattering of celebrities and brands. No verification, no unauthorised replies for you.)

Last night at around 9pm, I happened to glance up at my follower number on Twitter and noticed that it had suddenly (and instantly) dropped by around 30 people.

This morning, I received my daily follow/unfollow stats email from SocialToo, and spotted this curious trend:

(this image is too large to fit on here – click to open in a separate tab)

39 spammers, all in a row. All of these accounts have now been suspended.

Which leads me to suspect that Twitter did another purge last night, following the one we saw last Friday.

Is this going to be a weekly event? I hope so. Assuming, that is, that they keep getting it right.

My last nine followers:

They all arrived at once. All come with great names like ‘Giggles Shepard’ and ‘Danger Lehman’. All are following 500+ while being followed by less than a tenth of that.

Good times.