GSME #13 :: Boulevard of Broken Dreams

social-media-asides

I’m not going to lie today, I’m pretty hacked off. When you start something as an experiment, one does so with the understanding that the data you’re using is going to be constant and untainted. In the last few weeks, it has been impossible to accurately gauge anything from Twitter’s analytics. If I was being paid to do this for a living I’d feel angry and upset, and I doubt that paying for the service would give me that much extra recourse either. I understand enough about how the Internet works to grasp that this ‘software’ has to be upgraded on the fly: you can’t realistically shut down the Twitter servers for twelve hours a week for maintenance. That means if summat’s broken, it stays that way until you can find the means to realistically fix it ‘live.’

The irony is, of course, that on the other page I’ve quite obviously tweeted, quite a lot, and that’s gathered the interest of a few people:

realtimefuckups

Here’s the issue: according to the metrics, I didn’t tweet anything on June 1st yet garnered over 20k impressions. In fact, the first time Twitter acknowledges I did tweet anything is June 5th. What this has done is made my engagement rate not simply plummet but become a completely pointless statistic to gauge anything by:

tinfoilhat

If I was wearing my tinfoil hat, I’d say May 19th was when all this ‘trouble’ began. I’d also love to know whether the means by which Twitter reports engagement has subtly altered too, because this whole decrease for me now seems off, especially as I know I’m getting bigger engagements overall using threaded content (that is, if I go off on a rant I makes sure each tweet is posted as a reply to the ‘parent’ message.) Needless to say, I can see organic impressions doing nothing but raise as the Summer goes on, because I will be doing my absolute damnedest not only to sell the Intenet of Words, but to use Social media as effectively and ruthlessly as I possibly can.

dramatickangaroo

However, in the interests of balance, I have to make a few observations with relation to the amount of crap I now see in my feed… and, to be fair, it has dropped dramatically. By ‘crap’ I mean random followers who’ll pick up a high performance tweet and effectively flag it with a bot account. As was mentioned last week, hashtagging content used to result in a flurry of robot follows (presumably by websites who charge you to link up with like minded content to extend your reach.) With the demise of CoPromote at about the same time as all these changes came about? Well, it doesn’t take a genius to work out that Twitter might not just be taking steps to remove automated followers from its software. It could well be clamping down on those who use Twitter to make money for themselves.

With the potential value of the company long term, I really can’t say I blame them.

I look forward to seeing if a) anyone even bothers to respond now I’ve cc-d in Twitter’s Support account and b) on the amazingly slim chances they do, whether there’s any comment other than ‘we are aware of the issue and are working to resolve it for users as soon as possible.’ I understand how this works with gaming UI after many years of banging a head against the same wall.

I don’t really expect any change now.

GSME #12 :: Everybody’s Changing

social-media-asides

It’s been a few weeks away from the Experiment, what with operations and the like, but I’ve not forgotten the task in hand… except some rather important things have changed since the last time we looked at stats. They are nothing to do with me, and everything around how Twitter itself does business, and they have had a significant affect on the client’s reporting tools at a basic level.

blogs

I knew this blog post on the 17th was coming well before it arrived. The standard infographic data stopped working for a few days beforehand. Tweets would not register any hits, the real-time reporting went tits up and some key figures were royally screwed.

changestwitter

As you can see, I didn’t gain 3389 followers this month, but for some reason this is is how the details on my account have been ‘reset.’ You can work out when that was: that dip in followers corresponds in the 55% and 47% dip in profile visits and mentions. Something major happened mid month that (I suspect) may also have removed a lot of bot accounts and automated followers, as the number of random new people I used to get as a result of using hashtags in posts has effectively vanished overnight. The meat of the Twitter blog is however far more concerning: focussing on advertising data, syncing accounts across devices and effectively tailoring your ‘advertising’ experience if you don’t have the good sense to opt out.

engagement_truth

The reality of my engagements over the last couple of weeks is that, like it or not, I’ve just not been trying that hard. Mostly that’s been because of recovery from my operation: there was also a Tweet last week that acted as a salutatory reminder that reach is not everything. I made a comment on a message by Duncan Jones (@ManMadeMoon) which he then subsequently retweeted. He justifiably commands a massive online audience, but getting them to interact?

brilliant_not

broken_gubbins

So I manage 38k views, and only 17k of those are self-generated. The rest are that one tweet, that garnered a less than 1% engagement rate. I can’t even be confident that these numbers are correct, considering that I didn’t do anything last week, according to my metrics. Until there’s confidence that what I’m seeing is in fact correct, it is probably an idea to not place too much store in what I’m presented.

It’s fine, there’s plenty to do in the meantime instead, including addressing a significant issue with a tool I was beginning to enjoy using:

CoPromote has been broken since April 29th, and the last active tweet from the company was on May 13th. If I were a betting woman I’d say that that interruption I saw happen pretty much put pay for good to this form of promotion, which made money that should really by rights should have been Twitter’s to begin with. It reminds me to go and see if my credit card was charged for May, and if it was to make sure that before June that amount is denied. I’m betting that we’ve seen the last of this Company, and that other such ventures will be quietly yet firmly shut down by Twitter going forward.

It means, starting in June, that there’ll need to be some serious consideration on how I use Social media for effective promotion.

GSME #11 :: Personal Jesus

social-media-asides

I should apologise for being a couple of days late with the Guide, but as history can show this as being Monday when it is in fact Wednesday, I doubt there’s too much of a fuss to be caused anyway. Health issues have inevitably derailed my progress, but they’re also a powerful means of driving social media. If you didn’t know already, drama generates interest.

100517.png

Let’s be honest here: I’d rather not have all this fuss and bother right before one of the most important periods of my life, but the Universe does like to mess with your head at the most inopportune of moments. Therefore, my views last week were exclusively driven by the fact I’ve become my own news story, like it or not. I could choose to read these figures in an number of ways, but at this point I’d just like to think that more people than normal care about my welfare, and that’s where we’ll stop. The better news, if we look at the graph on the right, is that I’m keeping close to exceeding a 2% engagement rate.

How am doing that is part of a structured change to my Tweeting habits.

nervous

This Tweet has a picture, from my phone, of me waiting in the Hospital to see my surgeon on Monday night. If I’d have posted it *without* the picture it would have received less interest, because what the image does is make this less about having a conversation. The image, like it or not, draws you into my story. That’s why SnapChat and Instagram are as immensely useful and compulsive as they undoubtedly are as tools for generating traffic: the brain, in most cases, reacts more favourably to visual stimulus. This means that if I post something with a GIF attached, more people will read it, as is the case with a screenshot or a photograph. Once upon a time mobile download limits might restrict this as useful, but as WiFi becomes de rigeur and mobile devices become portable newspapers and magazines as well as gaming rigs… use more pictures.

surgeon

On the flip side, one well worded, concise and honest tweet is worth an entire blog post.

Picture + personal affirmation, in this case, results in a 9.3% engagement rate, if you didn’t believe me when I said that pictures sell. On the flip side however, being overly personal can turn people off. There’s the individual on my feed for instance who’s almost permanently ragging on their ex-partner for being horrible, and although there is a continued measure of sympathy for the situation, that person does go on Mute when it all gets too much. As is the case with life, balance is everything.

thread

Finally, you don’t need pictures, a personal dilemma or the end of the World to engage an audience. With the right words, a bit of thought and some planning, you can use Twitter to write blog posts, by ‘threading’ a series of tweets together with an appropriate picture at the top. This method of communication is proving increasingly useful for me, and I’ll be making the most of the format going forward. This week, there’s not much else to report really, and with an operation scheduled for the 16th, it will be a little quiet around these parts on Monday anyway.

Time to go practice what we’ve learnt and be ready to start again once all this drama is over.

GSME #10 :: Back and Forth

social-media-asides

Ten weeks into our journey, and there’s a moment to stop and take stock of the last two and a bit months:

April was, like it or not, my quietest month for a while, which probably has a lot to do with me taking time off at weekends and being less involved with the gaming side of of my feed overall, as I prepare for the new project. I’ve also started quite aggressively blocking and ignoring people who simply join in the hope I’ll follow them back, simply for audience numbers. Despite this, I’ve seen an up-tick in new followers. All told, I’m pretty satisfied with where things stand, even though (as I discussed in the week) the CoPromote attempt to get the new venture some publicity pretty much was a waste of time and effort. However, things aren’t as woeful as they perhaps first appear.

0105engagement

I effectively took the weekend off, and that’s demonstrated by the end of graph tail off. However the rest of the week was really rather robust: only Tuesday and Wednesday had anything to do with the CoPromote push, Friday was all my own work. In fact, these numbers give me hope that my forward planning could really start to show some results. Plus, having wanted to keep my engagement at 2% or above, today’s rate is already showing I don’t need an excessive number of tweets to do that, just to use the right content when I do.

The next couple of weeks will show a real split for focus: I’ll be setting up analytics for the IoW site once we hit early July, but until then I see no point in worrying over the details. There’s no intention to stop running or creating any of my other content either, but I will be amending posting days and content in the next two weeks. Basically, it will be all change yet with a focus on getting a better quality of response overall. I’ve decided that numbers really don’t matter nearly as much as getting a decent return on interaction and education.

I also still have some CoPromote reach to spend, and there are some new plans for that going forward.

GSME Special :: Baby I Don’t Care

social-media-asides

Normally I’d wait until Monday for an update on the GSME, but I feel that this needs to be written now, whilst the revelation is still fresh in my mind. I woke up to a mass of emails this morning, happily informing me that I’d almost exhausted the 500k’s worth of CoPromote shares I’d stored in three days. The results of that share, when you look at them, are pretty spectacular:

notcaring

50 plus retweets and 482 thousand people. WOW. That is indeed impressive, right up until the moment you translate that to my web stats:

scraped

That post was read 100 times. 31 times on Monday, 37 on Tuesday and 33 on Wednesday. Not a single person signed up for web updates, and (possibly) one person joined my Twitter account. I stopped the boost this morning because, honestly, I think there’s no point.

CoPromote’s entire business revolves around the understanding that if you’re willing to share someone else’s work, there’s a benefit to yourself. The problem, of course, is that many individuals simply won’t care about anything except what they stand to gain from the equation. As long as you can be seen to be doing something that appears to benefit others, it doesn’t matter about whether you take an interest or not. Looking at the people who shared me, and I have a long list, many undoubtedly did so so just in order to add my reach to their totals. How do I know this? Because that’s exactly what I’ve needed to do myself, sharing content that is not appropriate to my feed just to keep the numbers ticking over.

The truth of ‘sharing is caring’ is, in this case, pretty much a lie.

sharingiscaring

There’s a deeper truth here to be considered: I can see on my list people who shared my work with (it appears) nothing in common with me, except the fact I shared their posts previously. One assumes therefore their reasoning goes along the lines of ‘well if they did this for me and I do that back, maybe they’ll do so again.’ This is the moment that my husband would accuse me of being overly cynical and I’d look at that glass of water over there and know it is both half full and empty simultaneously. It would be lovely to give everybody the benefit of the doubt, right up to the point where you’re proved correct in your outlook. Assuming most people will be using this service for free, that means an awful lot of sharing to allow you to do the same. That means a feed full of stuff that could effectively be curated without you ever having to write a word. I can totally see the appeal of that for certain people, and then it stops being about one person hand curating their output and becomes something completely different.

greatjobguys.gif

What you have then is a Twitter account full of other people’s work making your output look both relevant and vibrant, that nobody effectively reads. The problem with CoPromote for me, like it or not, is that it’s not a network that meshes with my interests long term. I could use it to float the occasional general interest piece, maybe with lots of pictures or a lighthearted subject matter but honestly, it is not what I need to sell my extremely niche interest project. Fortunately to discover this only cost $40, so there’s really no harm done. If I’d have gone full out with Twitter I would ironically have a far better chance of reaching the exact audience I want, because I can specify the particular type of user I’m trying to reach, but if I attempted to spend the same amount of money doing so? I’d undoubtedly have gotten about the same return, probably less. Effectively, if I want to sell the IoW, I’m going to have to do so for myself.

When I sit here and think about that conclusion, it isn’t a surprise at all.

GSME #9 :: Talk to Me

social-media-asides

This week, it must be said, has the potential to be a game-changer.

240417_1

My numbers are down overall for the week, which is due to one thing alone: I didn’t do the work. Instead there’s been a phenomenal amount of back end work, which will continue I fancy for the next four weeks, as I organise and put in place all the parts of the disparate puzzle that will become the Internet of Words. I’ve made the decision to have a separate Twitter account for that ‘brand’, away from what is being done here, because the subject matters are so very different: it is only where the two worlds meet here that the overlap will be noticeable. I’m looking at separate tools for that account’s potential growth, but for now we’ll focus on what happens here in the ext seven days. As you can see, I’ve quietly tipped into another multiple of 100 this morning. What I’m more interested in is the 6% increase in mentions, and the fact I’ve got some idea now of what is effectively doable in terms of long-term engagement.

240417_2

The plan therefore this week is to stay above 2% engagement and to try and work towards 3%. Next week I plan to make my own graph of how engagement has panned out in the last ten weeks, as that’s a reasonable time frame to show how effort relates to result. The way to keep that number high is to lay off the unscheduled ranting (which I’m getting progressively better at, it must be said) and focus on content people like: pictures, useful articles, insights on my life. I successfully scheduled from WordPress this morning, and have reactivated Hootsuite after an absence.

I’m paying for it, so I may as well use it, and it is proving already quite useful for showing up the holes in my social media ‘strategy.’ That means that this week’s maintenance task, like it or not, is to go redesign a bunch of effectively dilapidated Google+ pages to make sure that niche is covered come Patreon launch. I also have a Post-It note with ‘Pinterest’ on it for my art projects, plus the visual side of the project which is also going to be featured via Instagram and Flickr.

Needless to say, this morning my brain hurts.

verypointed.gif

If I can get all this right, I’m hitting large areas of effectively undiscovered audience with a concerted effort. My CoPromote 500k reach is sitting freshly delivered for April too, and that means that tomorrow there’ll be a post here with some mission goals and a Call to Action, which I hope to repeat on the Facebook page. I’ll consider it a success that I get one person to be interested at present, especially when there’s effectively nothing to promote until June 1st. The key, of course, is to build interest in an engaging and attractive manner.

If I can get everything to mesh together effectively, who knows what I might be capable of achieving.

GSME #8 :: Your Cheating Heart

Last week, if you were paying attention, was fairly significant for this site. That means that after today things won’t look the same, but this Guide will remain plugging away at improving the reach and interest in what is about to become a fully fledged ‘brand.’ In fact, if all goes to plan, I will be updating WordPress on this account to a Business Plan. That means that SEO is going to become rather an important part of what goes on around this parish in the months that follow, but for now I’ll simply mention it in passing. For now, we have Twitter analytics to look at and some more foundation work for the months that follow.

170417stats

The peak on the Engagements graph last week was my resignation as @MMOGames’ Warcraft columnist (in anticipation of the Patreon in June) and because of illness, the effort to engage has, I’ll admit, not been as full on as in previous weeks. However, I’m definitely feeling more awake, aware and willing to go this morning: I managed to CoPromote enough posts to get me to 44k Reach, but I’ll wait for my purchased 500k’s worth to hit the account before I use it on summat worthwhile. However, my 28 day stats are looking distinctly encouraging:

28summary170417

What is happening is the constant drip of what I know are genuine followers and not either a) robots or b) people with their own promotion to highlight. It helped at the weekend that a post I wrote on the Warcraft site appears to have been ‘shared’ somewhere that’s not on my radar. This normally means Reddit but as my WordPress analytics simply list the hits as from ‘search engines’ it is just as likely someone used me to mask traffic for summat nefarious. However, the number of relevant follows that could be as a result means I’ll take whatever, without complaint.

views2

As you can see, my Warcraft site generates only a handful of hits on any given day. Ironically, the post in question is a less than glowing review of current content, and isn’t something I intend to repeat on a regular basis. However, it gives me an interesting blip on the stats to write about, so you take your anomalies wherever you can find them.

I also considered this week promoting a post using Twitter’s own boost feature. However, when I tried to work out roughly how much this would cost, I could find nowhere where pricing was listed, except from 2012. A quick look at Twitter’s knowledge base revealed why:

nomoney

If I have to hand in my credit card details before I am even charged for the service, I’m really not interested. I appreciate this may be standard practice for other services, but as I’m not a massive company but a single person with the most limited of budgets? Nope. Yet again, Twitter are a complete waste of time and CoPromote continues to look like the best move I could possibly have made in order to extend my reach into a completely new and unknown audience. It is odd how these things work out: I’d thought the service would be a complete waste of time, but for what I’m paying (set amount, easily budgetable for) it knocks spots off the ‘professional’ alternative.

Sometimes, the biggest provider is not necessarily the best choice.