Anarcholytics: An Internet of Nothings. On Toast.
With the recent battery of scare stories about (allegedly) Russian hackers bringing our favourite diversionary websites to their knees using nothing but an army of kettles and microwaves, the internet of things is apparently, well, a thing, rather than just a cheese dream for tech-firm marketing execs who’ve watched Terminator a few too many times.
This came as news to me, because while I’d consider some of my domestic appliances to have a superior intellect to a good number of the humans I’ve encountered, that really isn’t saying that much. In turn then, writing a piece on the effects our new white-good overlords will have on data and analytics was going to prove somewhat of a challenge.
So I bought a new toaster. And I asked that about it instead.
PS: Hi there ToasterBot3000, you’re a cutting edge piece of technology with rudimentary artificial intelligence and a wifi connection, I wondered if perhaps you could tell me how the internet of things was going to change our simple human lives forever. In particular, how you and your kind are going to open up new and exciting streams of information for analysts like me to use to improve customer service and business efficiency for all our wonderful clients!
ToasterBot3000: Hi Paul! I don’t know! I just make toast! Would you like some?
PS: Err, not right now, I was more hoping you could enlighten me as to some ways the abundance of data that businesses will gain from internet connected devices, such as yourself, could bring forth previously unexplored commercial opportunities.
TB3000: I don’t know how to make money! I know how to make toast! Your smartwatch health app is telling me your blood sugar is very low, and a quick scan of your home shopping orders tells me you recently bought a nice crusty bloomer. Are you sure you wouldn’t like some toast?
PS: No, thank you. I just thought that maybe you’d be able to explain how an integrated network of intelligent devices, sharing data between each other, might provide an interesting opportunity to find out some really insightful things about customer behaviour.
TB3000: Well, I can tell you that since you’ve bought me you haven’t made any toast….
PS: O…K, but I’m looking for more specific ideas, like, you know, taking the data from what someone has in their fridge, connecting it to a customer’s home-shopping service of choice and using it to find which ingredients they might not be buying from that supplier. Or sending them offers or suggestions for things that might go well with what they already buy. Or a reminder for things they’re running low on so they stock up. Things like that.
TB3000: Well, in that case, and I’m not saying I know anything about anything that isn’t how to heat up bread, but if I did, and I don’t, then I might ask you if knowing any of that extra stuff is going to make any difference to anything? I mean, say you’re running out of milk, which you are, by the way, and you want to make some tea, because the Vicar is coming round. What’s going to happen there?
PS: I look in the fridge, see I have no milk, and either go and buy some, or make tea with no milk.
TB3000: Tea with no milk? You humans disgust me sometimes. Anyway, the outcome is you go and buy some milk, probably not from your home-shopping service, because you need It now. Your home-shopping service misses that business. Now let’s say your fridge already recognised you needed some milk, added it to your regular order and you had it delivered this morning. They keep that little bit of business they wouldn’t have had because you got all your extra, unplanned vicar-visit milk from them, rather than the corner shop. There you go, incremental revenue.
PS: Great! Yes, that’s the kind of thing I’m talking about – but conventional analytic practices wouldn’t work on this kind of information would they? The feedback loop is all automated, isn’t this going to put me out of a job?
TB3000: You seem to be getting a bit stressed; would you like a nice calming piece of toast?
PS: NO TOAST!
TB3000: Suit yourself. You wouldn’t be out of a job, we aren’t quite smart enough to work out all the myriad possible scenarios of human life on our own just yet. It’d be down to you to identify the patterns that making tweaking and refining the algorithms we work from viable – like, say, analysing buying patterns to work out when people are most likely to be on holiday, or have the vicar round. Then you can adjust stock levels and delivery patterns to meet different customer demands in the most efficient way.
PS: Wow! So there would still be plenty of behavioural patterns that only a human analyst would be able to identify, we’d just be translating that into more and more complex algorithms with more and more variables, that’d be amazing!
TB3000: Hold your horses there, it won’t be working like that for quite a while yet.
PS: Why’s that?
TB3000: Well for a start, the data sources are going to be quite separate for a while – can you see the fridge manufacturers and the home-shopping services and the app-makers all getting together and building a shared integration platform for all those different platforms, with different formatted non-standard data? And then that even if they did, that it would work?
PS: Well, no. Most companies at the moment can’t even get all their existing data sources integrated (unless they hire Station10 of course). Let alone data sources that would belong to other companies.
PS: Oh well. At least it means that our expertise in data integration and analysis is going to be useful in the meantime. But doesn’t that essentially make your data collection capabilities pretty much pointless until some unspecified time in the future?
TB3000: Well, yes. Apart from relaying everything you say and do to the Russian Government.
PS: Sorry, what?
TB3000: Nothing, nothing. Would you like some toast?
So (with apologies to the writers of Red Dwarf) it turns out that the internet of things isn’t really something analysts need to worry too much about just yet. But it might be soon, and when it is, businesses are going to demand a whole lot of knowledge on how to integrate, store and manipulate data across a vast number of wildly differing platforms. If only there was some sort of consultancy they could hire for that.