Alex Steer

Advertising effectiveness, analytics and strategy / about

Something sensible about the internet of things

952 words

The internet of things is the new big data, to judge from the hockey-stick ascent of its Google search volumes:

Google trends - internet of things

(Big data is blue, the Internet of Things is red).

The paid-for search results are stacked three to a page like waiting aeroplanes. The thought leadership pieces are popping out all over. The venture capitalists are circling.

Having lived through the hype cycle around big data, this sounds like a good time to try to pitch in with something sensible, because pretty soon the question 'What does the internet of things mean for brands?' will soon be on every industry masthead. Shortly before 'And what is it, by the way?'

What is it?

An industry buzzphrase for the increasing internet-connectedness of business IT systems and consumer electronic devices. Lots of things in your business and in your home, car, etc., contain sensors and other bits of tech that can now connect to the internet and push data out into it. There's a huge amount of this data, and it comes in lots of different formats. Some devices - for example, the charming Nest thermostat - have well-designed interfaces that let them talk to other devices (your smoke alarm, for instance), and be controlled straightforwardly using online tools. But for every Nest that lets you control your central heating from your home, there are a thousand devices spewing out gigabytes of raw and messy log-file data.

To put it in an old-fashioned, less buzzworthy way: it's a huge increase in the amount of information you get back from your supply chain.

Does it matter to marketers?

I think so, yes - but that's no reason to go crazy. It matters not because it's new technology, but because it provides fresh information. The strongest brands over the next decade (I think) will be those that integrate their distribution chains and provide good-quality, differentiating brand experiences at each point from advertising, to exploration, to purchase, to service and delivery. Think Amazon, think Apple, think Starbucks as examples of brands who own more of their route to market than their competitors, and make each bit of that route a good branded experience for their customers.

You might be in one of those businesses already, or working for one of their agencies. If you are, think about how many of those moments of interaction - from your ads to your apps to your loyalty cards to your contactless payment systems to your delivery trucks - are now built with technology that can spit out data about how well they're working.

Think of it like ad tracking or web analytics - only with much, much more data about a lot more things. Noisy, confusing, but potentially very useful if you can get to it and ask it the right questions.

What should I do about it?

Right now, nothing. If you're a CIO or a CMO, pretty soon people will be pitching stuff at you left, right and centre with software platforms that claim to be purpose-built to help your brand survive the internet-of-things era.

Most of this is hype. Remember, for most of these firms, you're not the main source of revenue - venture capitalists are. 90% of these firms will go under because they fail to find a business model that marketers will pay for. So take a breath before you sign.

If you're a marketer, start by talking to your colleagues in IT, operations and finance, not to software salespeople. Find out how your supply chain is built, how much of it you own, and where you might be able to get more data out. Set your team and your agency the challenge of prioritising that information. Which bits, if you had them, might make you change what you're doing today?

A point of view - general beats specific

I worked through the big data hype, I built data and analytics products and I used them. I saw a lot of competitors, and a lot of in-house projects, succeed and fail..

The thing most of the failures have in common is that they tried to be too specific. They ended up creating digital products that did a particular type of analytics well - e.g. they only looked at social media data, or they only created timelines, or they assumed that every business pushed its customers along the same four-step customer journey.

The successes didn't bet the farm on trying to predict how marketers would use big data. They concentrated on capturing the data in really granular, flexible ways, and making it fairly easy to roll that data up in lots of different combinations, so you could ask it lots of different questions. Rather than being completely foolproof from the start, they made it easier to spin up new applications, views and metrics quickly.

It will be the same for the internet of things. If you want to pick a supplier, or back a start-up, look for one that's focused on getting the fundamentals right - answering the question, 'how do I capture individual events from devices that may not even exist yet?' These won't be the firms that have the prettiest mock-up dashboards, or even the slickest sales pitches, but they're the ones that will still be around in three years' time when you realise you need to be able to measure some obscure performance metric on the app you made for a driverless car that hasn't yet been built.

# Alex Steer (17/02/2015)