Elephants, stealthy creatures one and all, have been cropping up a lot lately, so I’m running with the theme.
What’s the elephant in your room?
We all have big, nasty problems that are either horribly current or just around the corner. We know they need to be dealt with, tackled, met head-on with a stiff upper lip; but instead we choose to pretend they’re not there, in the hopes they’ll just go away if we wish hard enough.
Now, sometimes this does work! Usually when the problem is time-based, or contingent on a decision, and the “la la la” approach wastes enough time to make it all moot. Oh, I’ve been guilty of this, but I always feel terrible afterwards. Instead of facing up to the situation, I just let it… happen without me. Instead of taking the risk, I wait until the risk isn’t a question, and convince myself I wanted the safe route all along. (So even when the approach works, it doesn’t work satisfactorily).
I started feeling a lot better about the elephants in my life when I started confronting them, though. It turns out half the time they’re not even elephants at all, but out of the corner of my eye, when I’m looking the other way, they seem gargantuan, insurmountable and infinite. Turns out when you start looking closer, they’re really quite manageable.
So, yesterday’s Monitoring Social Media conference is over, and all I have to show for it is a heightened case of RSI (ok, ok, I jest). My live notes from the talk are here – Alice, you were an inspiration, I just had to call up mental images of your GDC typing-at-the-speed-of-light – how could I not publish the notes I was already taking? All that training at videogame events has certainly paid off.
But now it’s time to reflect and put together some marginally more coherent thoughts on social media and the lessons of the day.
Lesson One. Social media is people.
We’re finally starting to get it. Social media isn’t about numbers, or spreadsheets, or models, or calculating ROI to the last tenth of a decimal point. It’s about people, and you can’t (always) chain people down in tidy little tickyboxes and assign numbers to them.
We are not numbers.
This causes conflict in organisations that are used to the ‘old’ ways of doing things and don’t really understand the ‘new’. The case for the new was presented again and again and again yesterday. Look. We get it. Social media matters. People matter. It’s just difficult convincing higher-ups that it’ll impact the bottom line.
There were a few attempts to get some slightly more detailed answers on this subject. What exactly is the investment, when we talk ROI? Is it the cost of a tool? The cost of an agency? The cost of people? What will make the higher-ups listen? In the case of STA Travel, it was pointing out the properties of existing customers (that the STA relationship stopped once customers had booked a trip) and making a clear, coherent case for engagement to extend that relationship. But this brings me on to…
Lesson Two. Everyone is different.
We’re all human, and so naturally we want easy answers. But there are none. It seems that currently the range of social media monitoring tools (in terms of software offerings) is very much an off-the-shelf jobbie – obviously customisable to some extent within that, but still, off-the-shelf. Indeed, some companies with freemium/SaaS products seem to be encouraging this approach.
But if I learned nothing yesterday, it’s that everyone’s totally different, and that works for one client won’t work at all for another. Enter agencies, and humans (see point 3), and customisation, and tailoring. Hell, the agency behind Skype built a dashboard because nothing out there fit their needs! Weren’t all the SMM providers in the audience cringing at that? Speakers repeatedly said that today’s tools aren’t really that great – but some speakers praised them! What a load of mixed messages.
There is method to this madness, though, and it’s all about the human. People praising the tools probably used them well for their specific needs – people dissin’ them probably found that they were looking for something that the tools didn’t do. One thing seems sure though, the tools should work for the clients, rather than the 37signals-etc approach of ‘fit your thinking into the way the tool does it’.
Lesson 3. Automatic isn’t good enough.
This is obviously something I’m interested in, but it was almost disheartening to hear it repeated so much.
Basically, we need humans. We’ll always need humans. Tools help us cut down the humans’ time involvement, but there seems this fundamental mistrust – sentiment is wrong too much and too often, and even humans disagree 15% of the time (bang in line with the kappas I’ve seen in academia).
So even if there were a brilliant, perfect, 100% reliable sentiment detection system, it would be wrong 15% of the time, and so humans would want to check every message just in case. And if all you want is a ‘temperature’ type analysis, well, free tools already do that, and even allowing for error they’re just about good enough.
Lovely.
Lesson 4. We’re too close to the curve to see what’s around the corner.
The whole social media landscape is changing, and the monitoring stuff is just starting to catch up. Two years ago it was rubbish, nowadays it’s OK, and in two years it’ll be great. But the future’s not about technology, it’s about business intelligence, business process, and getting companies to embrace social media and its feedback loops at every level.
Because this is going to become such a fundamental part of how we do business, major players are already getting in the act. Search engines are integrating realtime search, so ’social’ SEO – building social capital – will become as important as keyword-based SEO. But you can’t just add in ’social keywords’ – that concept simply does not transfer.
As well as that, Google and Twitter could well be (hell, let’s just say it, they are already) developing their own social media monitoring systems. Google Analytics is powerful, but not in a social way – but it could be. Twitter could launch their own monitoring product and charge us for API use, creating an.. interesting, albeit unlikely, situation. Sure, cross-platform will still be a need, but we’ve already seen that that need varies so much even by department within a company!
One of the more interesting concepts to come up yesterday was that of an open source framework for monitoring social media, a plug and play approach that everyone could be using in two years – with a company making money where the hard stuff is, consulting and the human factor. I do wonder if this is perhaps viable, especially adding in outsourced human validation (MTurk) and cross-classification to reduce error.
Anyway, this is certainly all food for thought, and <shameless plug>should give me plenty to talk about at the RealTime ChristmasCrunch, at least!</plug>
The afternoon kicks off with a talk about data and customer understanding from Beyond Analysis. The entire afternoon’s talks are behind the jump – usual disclaimer, these are unedited notes, etc.
Recent Comments