All about Technology: Forget the Internet of Things – the Internet of Data is where it's really at

Introduction and business intelligence

Everyone's getting in a fuss about the Internet of Things, whether it's excitement about smart appliances like Nest, Belkin WeMo, smart bulbs and smart TVs, self-quantification fitness trackers, or about new genres of connected gadgets like smartwatches, Google Glass and the spread of Bluetooth beacons in stores. Oh, and the smartphone…

What all of these Internet of Things (IoT) devices – as many as 50 billion of them by 2020 – have in common is that they create detailed data about their users, and lots of it. From transactions and customer profiles to usage patterns and even location, the Internet of Things is really an Internet of Data that is begging to be cross-referenced and analysed. It exists not to help people per se, but as a new tool for collecting valuable business intelligence.

We asked a few business intelligence experts for their views on the Internet of Data – what it is, how it's organised and how organisations should best make intelligent, but ethical, use of it.

Data from smart devices

"By storing and analysing this data, companies have the opportunity to gain valuable business insight from IoT," says Matt Pfeil, Chief Customer Officer and Co-founder at big data applications company DataStax, who gives an example of a device that is being internet-enabled – the thermostat.

"Now you get a clear picture of how much energy is being used, when it is being used, and can compare this to other energy users," he says. "Link this up to other account data, and the company can observe patterns to save energy – for example, setting the default temperature a degree cooler at installation." It may seem a trivial amount, but across millions of devices it adds up to a huge amount of savings for customers.

"As more devices connect to the internet, the data they create can be linked together, and if you don't have that record over time, you can't get value from it," says Pfeil. "Time-series data provides the potential for insight around how we live and interact with our homes, our jobs and our cities."

Matt Pfeil, Chief Customer Officer and Co-founder at DataStax

The Internet of Data equals business intelligence

Companies are collecting data from all quarters to gain insight, but it's not just about understanding customers' 'buying journeys'.

"Operations and supply chain data would not normally be looked at alongside sales data, as they are viewed as separate departments within the business," says Southard Jones, Vice President of Product Strategy at business intelligence and analytics company Birst. "However, using analytics, you can see if there are common patterns in the customer base – are customers that spend more on service levels or speed of delivery more profitable than those that focus on price alone?"

This analysis then allows marketeers to target potential customers based on their favourable patterns of behaviour, whether that be the time and location they use an app, how and when they heat their homes, or – in the near future – when and where they drive their car.

With data storage so cheap, businesses find it hard to let go

Is money being wasted on collecting useless data?

With MarketsandMarkets predicting that the big data market will hit $46 billion (around £30 billion, AU$65 billion) in value by 2018, there has been much investment going on.

Existing systems and software are creating a lot of the data that companies collect. "A lot of money is being invested for potentially minimal short-term return," says Fletcher, who thinks that the potential of big data is in the interconnections and interrelationships it will reveal, not the volume of it collected. "Realising the benefits of big data in the sense of interconnectedness will inevitably take time and investment," he adds.

Since the cost of storage has plummeted, this is at little cost to businesses. "When storage costs peanuts, it becomes more economically viable to store data that you can create value from over time," says Pfeil. "Each transaction that a customer takes part in can create more value for the future, improving the service or providing further opportunities – the opportunity cost of not storing that data is less than the investment required to keep it."

The problem with big data analytics

Are marketeers simply collecting user data 'just in case'?

If we can, we should. That's been the attitude towards all kinds of data in the business world in the past few years, and after being constantly told that data is the future, it's no wonder. "The majority of activity is more 'just in case' rather than about gaining genuine or additional insight," says Gordon Fletcher from Salford Business School's Centre for Digital Business. "In being confronted with multiple terabytes of data, marketeers – just like everybody else – are more inclined to stumble than fly."

However, marketeers are the furthest ahead in using data to analyse and make decisions. "Take personalisation on e-commerce websites – this functionality is used to sell to us, but it also makes us aware of things that we might want or need," says Pfeil. "Collecting data over time helps make those predictions more accurate, and therefore more useful."

Beacons are about to begin collecting reams and reams of location-based data in retail stores

Is this 'spare' data being sat upon to sell later?

If every business has been told that big data has financial value now and in the future, you can be sure that the accountants have cottoned on to the value of so-called 'exhaust data'. "It's unlikely too many businesses would actively admit to this, but this is a real prospect as more and more businesses recognise data as one of their tangible assets," says Fletcher.

Such residual data is fuelling a growth in data aggregation companies, which at the moment is on the edge of the Internet of Data, but much of this data belongs only to those who first generated it – customers – few of whom will look kindly on companies known to sell-on personal data.

"For many companies, the value of big data is the knowledge of the customer that they get, rather than the strict monetary opportunity that could come from the data," thinks Pfeil. "Look at the energy sector – as we bring more smart devices into our homes, the utility provider can build up a better picture of how we live and where services can be updated to save energy over time, which adds up to a far closer relationship than would be possible in the traditional energy provider market," he says, stressing that 'monetising data' should only be about prolonging the customer relationship.

"Selling data would potentially backfire here," says Pfeil, who thinks that clarity is everything. "If people understand how their data may be used beforehand, they will be willing to share data in return for a service that provides value to them."

What's the problem with big data analytics and tools?

Many business intelligence platforms were designed only for data specialists to work with. "People across the business would ask them for reports, and they would generate them and send them back, often times after a period of weeks," says Jones of this now outmoded way of working with data. "The challenge with this is that you have to know what question you want to ask, and if the data you have will support what you want to find out [but] so much was invested in these applications, they could not be touched."

Jones explains that business users responded by bringing in their own tools so they could upload their spreadsheets and create visualisations of their own data, saving a lot of time. But not relying on a centralised team brings a big problem. "When more people are analysing data sets in different ways, you run the risk of getting many different answers to the same question," says Jones.

The answer? The cloud, of course. "It's important to marry both traditional business intelligence skills and ease of use together," says Jones. "Networking the business intelligence infrastructure so that everyone can work from one source of data is one part of this." With the Internet of Data growing every day, developing best practice within organisations – and employing data scientists to manipulate it – will become hugely important.