The combination of beacons, sensors and big data capabilities has opened up a whole new horizon for marketers. When it comes to having real-time relationships with consumers, we are literally doing things today that we only dreamed about a few short decades ago.
The way I define beacons and sensors is as any device that is either actively or passively sending content out to the internet that can be organised and tracked. It could be something such as your local Starbucks’ WiFi, pushing signals out into the surrounding area. You’re coming down the street with your mobile phone on, you’re in the Starbucks database because you have the Starbucks app, and as a Starbucks consumer the brand can initiate a communication with you.
If you’ve enabled the geo-location feature on your phone, we know you’re outside the store or in the area, and we know which Starbucks you tend to frequent. Obviously this is extremely valuable for one-to-one marketing in terms of crafting the offer and understanding a particular consumer’s behaviour – not only when they visit the Starbucks store but also when they use the app to redeem discounts or send an SMS text to respond to a promotion. All that multichannel behaviour is what marketers are after, but the problem is that although they might have had all this data for some time, each channel has historically been working in a silo and each silo does not know what is happening in real time with the consumer.
Wherever you have a human in contact with a customer, whether it’s in a retail store or a contact centre, you want them working from the same data. However, about 90% of the companies on the planet today are still locked in silos, where the retail experience is completely disengaged from the contact centre experience, the contact centre experience is completely disengaged from the mobile experience and the mobile experience is completely disengaged from the point-of-sale experience.
A couple of years ago, HP bought a UK-based company called Autonomy, the reason being for its speech and text analytics. By using its technology, we can effectively look at what we call neuro-linguistic business sensors. For example, we can pick out adverbs, adjectives and contexts from customers’ speech, whether that’s on the phone or in an online chat. The text and language patterns give us clues that can be translated into a standard set of service and marketing protocols, such as ‘this customer is annoyed, so don’t try to cross-sell’, or ‘this customer is using contextual phrases and syntax that suggests they are ready to buy’.
Natural language processing is not 100% reliable, but the way I look at it is in the same classification as consumer ‘intent to buy’ research, which is notoriously unpredictive of actual purchasing behaviour but it gives you a directional set of data that helps you know where your brand stands. Natural language processing is not predictive either, but what is predictive of real orders is behaviour in real time 500 feet from the Starbucks store, so the way brands are going to win today is by integrating behavioural, demographic and attitudinal segmentation in a way that enables you to turn the Rubik’s cube and start to unlock what is in the consumer’s mind.
At HP we have learned that about 70% of that data is noise and only 30% warrants further investigation. Only about 2% is an actionable service or selling opportunity, but once you get down to that 2% it is invaluable. It is an order waiting to happen.
If the ‘internet of things’ is the physical infrastructure that connects items to the internet, to me the applications and business processes that lie on top are the ‘internet of everything’. The way that HP and SmartFocus – our key partner in marketing automation – see it, the ‘internet of everything’ is the future of marketing to the 18- to 34-year-old demographic of digital natives. We are making a big bet on it.