Tag: Twitter

  • “Marketing Experts” on Twitter

    There was a point on Twitter where it seemed like every day I’d check my email and discover that I’d been followed by yet another “Marketing Expert”. I had to wonder how many there were…

    WeFollow was the obvious way to do this. So I made my way through the list, hacked some URLs and I have a number.

    17,450

    Seventeen thousand, four hundred and fifty. Seriously, check out this link. By the time you read this, there’ll likely be more. Crikey! Not saying all these people try and sell themselves as “marketing experts”, and these are obviously only those with the nous to register on WeFollow. But still. If we estimate the number of people on Twitter to be 1 million, 1.745% of the people on Twitter are in “Marketing”. They range from 0 followers to 997,354. The person in the middle (Number 8725) has 244. So there’s a power law graph going on here with the long tail graph. The first 3765 have over 1000 followers.

    Actually, that’s kind of less than I expected. So, good to know that Twitter isn’t made up mostly of people in Marketing. Adds to it’s appeal, for me at least. WeFollow is an interesting resource, I wish they had an API…

  • Spammers Getting Creative

    Spammers Getting Creative

    This morning I found a spammer who I actually have a grudging admiration for. They’ve been a little more creative than posting a headless picture of a young (hot) girl or mass following anyone who mentions whatever they’re pushing in the hope they follow them back.

    Check them out: http://twitter.com/SkiClubs

    At the moment of writing they have 491 followers (to the 640 they are following). Twanalyst, a service I find very useful tells me:

    Tweets per day: 8.5
    Readability index: (?) 18
    % conversations: 3
    % links: 96
    % hashtags: 1

    Personality: popular obsessive cautious
    Style: garrulous academic
    ROBOT

    So there’s actually a website about skiing linked to from their profile, which doesn’t try to sell anything and has some content. Not a huge amount, but enough to appear legitimate from a casual glance – maybe it is, even. A nice background and a profile picture.

    But let’s take a look at their ten most recent tweets…

    Tweet 1
    Tweet 1
    Tweet 2
    Tweet 2
    Tweet 3
    Tweet 3
    Tweet 4
    Tweet 4
    Tweet 5
    Tweet 5
    Tweet 6
    Tweet 6
    Tweet 7
    Tweet 7
    Tweet 8
    Tweet 8
    Tweet 9
    Tweet 9
    Tweet 10
    Tweet 10

    So here’s what I think is impressive (for a spammer):

    1. There’s content that looks legitimate.
    2. 357 Tweets so this is someone who has some history.
    3. Some of the content is actually interesting, I would love to know what “Skwittering” is!
    4. It’s targeted. (They likely added me because in my profile it lists “ski instructor”)

    Here’s what makes me say it’s spam:

    1. Anything which contains anything about “making money”
    2. Same “money making tweet repeated twice (tweets 2 and 8 )
    3. Almost all content links link to the homepage of the website. It suggests there’s content, an article or a blogpost – but there isn’t. So I still don’t know what “Skwittering” is…
    4. The content links as well are recycled. If you go back further through their tweets, there are multiple of the same of those as well.

    Here’s what I would love to know:

    1. Did they make the ski website themselves or is it one they found? How much effort are they actually going to here?
    2. What’s the click-through rate?
    3. Are there multiple areas which are targeted? I.e. are they running a “rock climbing” twitter and website as well.
    4. Do they actually ski?
  • Spam

    I spend a lot of time thinking about Spam. It may be an inherent danger of being a Programmer. Even though I take care of my email address not being posted around the internet I get junk mail. Some of it is quite funny, although I hear some people fall for the bank with unclaimed money in Nigeria… so I guess not for them.

    I bet back in the day people said that spam would ruin email, and I don’t mean to negate it as I know spam costs companies a lot in lost time and productivity. But I think Spam on Twitter and Facebook is even more interesting.

    Take Facebook. The primary source of spam for me on Facebook is Facebook Applications, that send me notifications inviting me to add things I have no interest in and frankly don’t want accessing my personal information to my profile. I definitely consider those to be spam, and block any that I receive. On the rare occasion that I do consider adding an application, asking me to notify some / all my friends will immediately cause me to change my mind. But – many of my friends who I’ve actively acknowledged a relationship with clearly don’t consider these aps to be Spam, as they add them to their profile.

    Maybe this is similar to email. After all, whilst the vast majority of people consider invitations to buy Viagra online, or contact from a bank manager in Nigeria to be spam sufficiently many people respond that it’s worth the cost to the spammer of harvesting email addresses and buying the bandwidth to send the spam out. I think the difference is, I don’t personally know anyone who buys Viagra in response to an email or really believes that they’ll get thousands of pounds out of Nigeria. So the response rate for spam on Facebook is higher? Or one person’s spam is another person’s… ham?

    However spam on Facebook doesn’t (thus far) have the power to ruin the whole experience. And since the Facebook applications are now old news, I definitely get fewer notifications. Twitter though – is different.

    Twitter has the issues of Facebook and of email. You can get spammed by people you follow (like you can get spammed by people you’re friends with on Facebook – the Spymaster game springs to mind) and you can get spammed by people you don’t. There’s an intereting article on Twitter spam here. But the real way that spam annoys me on Twitter is because it takes over every trending topic. So, in an unscientific study I checked the trending topics a little after 19:00 BST today, and counted how much was people just including the topic to get their post showing up in searches for it.

    Trending Topic # of Spam Posts % Spam
    Follow Friday 0 0
    #whentwitterwasdown 1 5
    #welovekevinjonas 2 10
    #flywithme 4 20
    GI Joe 2 10
    RIP John Hughes 3 15
    #tls09 8 40
    UberTwitter 6 30
    TGIF 4 20
    #IranElection 0 0
    Average 3 15

    So there’s an average of 15% spam, which is lower than I expected (clearly things have improved since this article) but still – some trends are as high as 40%. I want to say – really, does anyone fall for this… but clearly like all spam enough people do for it to be worth the very low overhead for spammers to do it. One of the coolest things about Twitter is it’s ability to capture the zeitgeist, and trending topics being full of spam damages this.

    Another thing I consider to be spam is pyramid follower schemes. You know the ones, you follow everyone and give them your details and it spams your followers with the link and adds you to the list, so the next idiot follows you. I’ve un-followed two people (doesn’t sound like much, but I only follow about 50-60 people) for using these. I loathe them – and there are so many! I’ve been keeping track of the ones that come up and (these are just those that I’ve made a note of) have a list of 13 since June. They have such inventive names as: “Twitter Train”, “TweeterPro”, “GatherFollowers”, “BestFollowers”, “FollowersFast”, “QuickFollowers”, “ThousandFollowers”, “EasyFollowers”, “FollowersFree”, “AddFollowers”, “MaxFollows” and they all look the same, and all have the “VIP” option – which is how they make money. Really? People pay to spam the followers they have in the hope of getting more followers who are only following them in order to get more followers themselves? Really?!

    Great article which highlights the stupidity of such schemes and that the best way to “grow your follower count” is organically, by interacting and treating Twitter as a conversation medium not a broadcast medium is this one. For people wanting to run a live Twitter stream on their website, spam in the search results is a real issue. TidyTweet aims to minimize this (and take out inappropriate language as well). Whilst there have been a plethora of “make money on Google” links lately, even legitimate companies have been guilty of spamming, so I can’t see there being an easy fix. And, when this kind of promotion makes it into the trending topics I wonder where we draw the line between what’s spam and what’s just an aggressive and questionably honest marketing strategy.

    Unwanted @ messages are another form of spam that’s pretty common, although Twitter seems to be clamping down. This morning I had a RT that I never tweeted in the first place, but by this afternoon it (and presumably the person who sent it) was gone.

    Finally, Bot’s that follow you because of keywords or just because in the hope that you follow them back and click on their links are the things that annoy me most. Apparently it’s grounds for account suspension, however I haven’t noticed much effect on the number of times I get added by accounts following this strategy. I’ll have to write another post just about this, because this one is already too long. Briefly, for the past 2 months I’ve been tracking who follows me, if I suspect they used a keyword, and whether or not I followed them back. Since the start of June, 120 people have followed me. If you check out the number of people I follow, it’s apparent that I haven’t followed a fraction of them back. I’m also not being followed by much over 100 people, so clearly they haven’t kept following me! They often give you 24 hours to follow them back, and then unfollow you in order to maintain their ratio. In fact, 120 is inaccurate because this contains the same account following me 3 times!  Of this 120, 3 I was already following (when they followed me), and I followed 21, 3 of whom I later unfollowed. So 80% of the follower notifications I received I did not consider to be worth following back.  This article demonstrates an extreme example of this. That’s a pretty high rate of spam. If you’ve been following everyone who follows you back, there is something you can use to get rid of them – mashable has a good article on this.

    Of course these are just ongoing issues I have with spam on Twitter. There have also been isolated incidents, like when loads of legitimate accounts were hijacked, or fake celebrity re-tweets. Twitter is also definitely trying to minimize the issues of spam, although with mixed results, such as blocking bad URLS, and suspending accounts. But with the rise of sponsored tweets, such as this service it may get more complicated.

  • #unfollowdiddy

    #unfollowdiddy

    Earlier, I was trying to work out why “Amedisys” was trending (job postings! Crikey), and then people including it in their posts just to get their posts showing up on that trend. Does that work? Do people really follow people who just tweet all the currently trending topics periodically? One guy was trying to pimp out his facebook group follower scheme just including all the popular trending topics with a link to it. The group had 2 members.

    And it reminded me of this article that I read this morning, analysising the #unfollowdiddy trend and concluding:

    “…the most common reason for unfollowing Diddy was because others were doing so…”

    Anyway, he thinks that period had zero effect on Diddy’s follower count. But I graphed the trending topic earlier and we can also graph follower counts, so I put them together below:

    Diddy Followers and #unfollowdiddy Trend

    I’ve tried to line the graphs up so you can see if there’s any effect. Looks like the trend didn’t reduce his number of followers but the graph did plateau a little. So even if the trend didn’t have the effect it was advertising, and even if it’s not bad publicity, he wasn’t gaining any followers as a result of it.

    Looking for this, I found myself at TweetStats. Diddy’s top 5 words are apparently: lets, people, rt, day, lol and the person he most commonly “rt’s” is himself. Hmm.

    Twitter trending can be a great way to see what’s going on, what’s hot right now, what are people engaged in, what’s interesting… but there’s so much noise. It can hard to draw conclusions without looking at each trend individually. We know why iPhone is trending today, because the service pack is out etc – but do we need Twitter to tell us that? The blogosphere and the news are buzzing with it too. Weird, unexpected trends are much more interesting – and what I’m taking from this is that they don’t always mean what we think they do.

  • Wow

    Wow

    The US State Department confirmed that it intervened to keep Twitter going and downtime was recheduled to take place in the middle of the night in Iran, not in the US. The #IranElection hastag has been trending for days. See:

    Trends on Twitter

    I saw this, and thought – wow. This is big big news on Twitter. Right now the number 2 trend is #haveyouever and it’s way below #IranElection. Gordon Brown, who was trending the other week can barely be seen. #followfriday, which once a week generates a load of spam is still below #IranElection. The #unfollowdiddy trend was really shortlived. People really care about this, if you search for Twitterers in Tehran, well you’ll find a whole lot more than you did last week. In order to confuse censors, people are changing their timezone and greening their picture in a show of support.

    Then, I graphed this…

    #IranElection and iPhone Trends

    Well I guess we all have our priorities. Democracy – and people – may be dying in Iran… but at least we have the latest iPhone. The iPhone’s trend has this really sharp peak, but also is quite broad. It might not be number 1 right now, but the US is asleep. I really wouldn’t be surprised if that changed later today.

  • The Iranian Election is Trending

    The Iranian Election is Trending

    There’s a large Persian population in Ottawa, especially in Computer Science. So, a lot of my friends are from Iran and I follow the international news so I’ve been watching the election results with interest. One of my friends wasn’t going to vote, so the other day I was embroiled in an argument about that (it’s his responsibility to, don’t complain if you don’t vote) during which another (Iranian) friend declared that I should have his passport, which I found rather funny.

    Anyway, living in the West where our media coverage is no doubt biased against Ahmadinejad (although comments like this speak for themselves) and all my Persian friends hate the guy with varying degrees of venom, I’m likely biased. None of them seem to buy the election results. The margin is unconvincing. I’m inclined to agree.

    So, lets check the trending data. Note though – this is likely to be skewed in favor of Ahmadinejad because he’s the current President.

    I’m going to start with Twitter this time. You can see the graph from Twist below:

    twistThis is interesting, because it’s actually Ahmadinejad that was more popular – I expected the opposite as Twitter likely has a more educated, liberal population. Although it looks like Mousavi has been more popular since the results came out – they’ve even been using Twitter to organize protests.

    Blog Trends are below:

    Blog Trends for Iranian Election

    Not much to say here, but we can see the curves follow each other from around June 2nd, likely the election boost – but – Ahmadinejad is ahead of Mousavi.

    Finally, Google Trends. This is more interesting, I think. Below is the graph:

    Google Trends for the Iranian Election

    Underneath the chart we can see where most of the traffic comes from, unsurprisingly it’s most from Iran (particuarly Tehran), in English and Persian.

    Google Trends Breakdown for Iranian Election

    So if we look just for Iran:

    Google Trends for the Iranian Election (Iran)

    This is interesting – because it’s so close. It’s a shame that the graph just stops because it doesn’t show what’s the case right before the election but on the face of this it’s hard to beleive the announced result.

    However, if we look at the breakdown:

    Google Trends Breakdown (Iran)

    Mousavi is only winning out in English, Ahmadinejad is doing better overall, in Tehran, and in Persian.

    Given that Facebook was temporarily blocked in the run up to the election, I would love to see the trending data there.

    So – hard to draw conclusions. At the end of this, I still don’t really beleive that Ahmadinejad won by that margin – although ahead in trending stats I don’t think more than would anyway be biased towards a sitting president. I’m not saying he couldn’t have won fairly (in which case, any perceived tampering can only be bad for him), but the margin is just huge. But the real tragedy is how may of the Iranian people don’t beleive that democracy has played it’s course. Can that ever have a good outcome?

    Update: Mashable has a good article on how social media can help you track what’s going on with the Iranian election here.

    Update: Protesting by crashing websites. http://www.pagereboot.com/ will let you put in a URL and auto-refresh it. Not advocating this, but if you wanted to participate some URLS are…

    • http://www.ahmadinejad.ir/
    • http://www.farsnews.com/
    • http://www.leader.ir/
    • http://www.moi.ir/
    • http://www.shora-gc.ir/
    • http://www.irna.ir/

    Update: Website crashing is slowing all traffic in Iran, causing problems. People listing useful ways to help are Green Revolution and apparently also this one which has been taken down, hopefully will be up again later.

  • Tweeting Normally, Mr. Average

    This week I’m at another conference, this one put on by FOSSLC. It’s an Open Source conference, which isn’t strictly related to what I’m doing but it’s good to get out there and meet people, and hear about what people in industry are saying.

    I’ve had a few conversations about Twitter… of course. I’m really fascinated by how people use Twitter. I follow a little over 30 people myself, don’t actively look for people to follow, but if people follow me and I know them or they look interesting I follow them back. You can see my “twitter personality type” (and check out your own). Apparently my tweeting is now “tip-top”, how flattering! Finally an electronic device that doesn’t disparage me… my wii fit puts me down constantly. However until quite recently it was telling me that I needed more followers and I should try following people in order that they follow me back. I hate that tactic, and, by extension – things like TopFollowed and, this one even more horrifying TopLinked (do they know what LinkedIn is for? Really?!?). If that’s the future of social networking, I want to cancel my internet connection.

    So yesterday, I got talking to a guy about Twitter and he said he had 1000 followers and was following about that many people and of course I wanted to know how he managed it. Following 30 people, my feed is pretty busy (especially first thing in the morning, a lot of my friends and the people I follow are on UK time). He can’t possibly be reading the time-line! His strategy: he has a search for his username running and only saw those tweets directed at or referring to him. Twitter is useful when he wants to ask a question, he gets real-time answers. But clearly given how he uses Twitter he won’t be answering other people’s questions unless they are directed at him. Perhaps his followers don’t mind, and those that answer just enjoy being helpful. Perhaps they haven’t yet reached critical mass in their own following that they need a similar strategy, but eventually, if we all try to use Twitter the way he does no-one will be answering anyone’s questions, except when asked directly. But all this is speculation. Even if his strategy sounds a little free-loader-esque, it’s certainly not going to bring down Twitter any time soon.

    I also met two other people who said they “didn’t get” twitter. One had abandoned it, the other was just passively following and wondering about business models.

    There was also another guy (back to him later) who, like me, had been convinced to try Twitter by the NYTimes post, “Twitter? It’s What You Make It”. Not a very insightful statement on the basis of it – isn’t the same is true of any social network? I might use Facebook to keep in touch with my friends, just to passively stalk people I barely know, to vet prospective employees, or to try and pick up far younger and more attractive women. OK the last one would be hard – but a creepy guy did add me who was clearly doing just that, looking through his friend list it was just beautiful woman after beautiful woman… I blocked him, but I was flattered to be chosen! But I think Twitter is more versitile still, especially with the open API.

    I’m a typical NetGenner I guess, I use Facebook to keep in touch with my friends. I use Twitter to keep in touch with my friends too, but also to get news stories from a specific area of the paper (technology) that interests me (via @guardiantech) – so I don’t get distracted by the front page. And I follow various others who link to interesting / useful articles. But… I want the majority of things that show up on my feed to be interesting. And I’d like to be able to read pretty much everything that shows up on it. Perhaps that’s weird. But I think the truth is – no-one’s weird. I was talking to my supervisor the other day about Grown Up Digital: How the Net Generation is Changing Your World because one of the things Tapscott says is that his daughter has over 700 friends on Facebook, and that’s normal. Not normal for my circle, and in fact overall on Facebook the average user has 120 friends. Of course you’d want to adjust that for age, which you don’t have the information to do. Anyway, so I said, I don’t think this is normal, and my supervisor said, but what is normal any more? Do we know?

    Probably we don’t.

    So now I’ll come back to the other guy, Mekki. Before, I wrote about Average Users, which was driven by a paper I wrote about Vista. So when I saw there was a presentation on the user’s perspective of switching between Microsoft Office and Open Office I was keen to see it. And so he started his presentation and he too was talking about Average Users, so a little way in, I asked if he’d ever Googled “average users”. And he said, no, what do you get. And I said – sod all. The only definition I could find was in Wikipedia. We talk about our average user a lot but we don’t know who they are, we don’t know what they want, and we don’t know what they do. And he’d come to a similar conclusion. More – Sun had realized this back in 2004 trying to promote OpenOffice. That’s a company trying to promote use of free and open source software, not the company charging hundreds of dollars a pop. Wow. You can find his presentation and thesis here.

    So we don’t know what normal is. We don’t understand average anymore. But we’re trying and, even if it makes our lives as programmers a little more difficult… I think it makes our lives as people a little more exciting.