Finding customers' stories and voices in the data


Read a TDWI Checklist report: Marketing Analytics Meets Artificial Intelligence

I’m going to start with a question. What’s your favourite Internet of Things (IoT) device, product or service? A Fitbit, perhaps? An app on your smartphone?

My next question: why do you like it? Because it’s exciting, techie and interesting? Because you like to have things before anyone else? Or because it is genuinely useful to you? I’m going to go out on a limb and suggest that for most of you, it will be one of the first two, with far fewer in the final category.

A tale of product adoption

IoT is still at a relatively early stage of product adoption, as innovations go. Most of those using IoT devices are ‘early adopters’ if not ‘innovators’. And to my mind, that’s not surprising, because most IoT applications currently available, with a few notable exceptions, are still largely driven by technical possibilities, rather than user need.

IoT applications that have caught on in a bigger way—smartphone check-in to airlines, for example, or control of home devices such as heating and lighting—are successful because they meet a real need among consumers. But if anyone can explain—I mean really explain—the overwhelming need for Alexa, I will be surprised. Fitbits? Yes, they’re fun. They are becoming more widely used as people see that they can compare what they are doing with friends, and share data about exercise. But many other devices, like the Apple Watch or Google Glass, have failed to take off widely, rather than among tech enthusiasts. They have not answered the basic question ‘But what is it for?’

For the majority of people, this is crucial. Purchases have to meet a real, defined need—and for most of us, that has to be more than ‘I want to be the first person to have this’. For most people to buy something, they must see a genuine need for it.

This is where data science meets product development.

Finding the story in the data

One of the most important uses of IoT data is to explore user experience. Direct customer feedback gives information about customer views, but use data shows how they actually use the device or app in question. Both derived and direct data, in turn, can be used to provide solid information about what customers want and need, to inform product development.

But what questions should you ask to get the right answers? I think that there are three to ask initially:

  • Is there valuable information within my data? In other words, does it have a commercial value? And is it relevant to what I am trying to understand?
  • How can I extract it?
  • How can I feed it back into the product development process?

When you are deciding whether something is relevant to what you are trying to understand, it is important to consider how seamless the experience is for the customer, because this is the fundamental aim of the process. Looking at patterns is both interesting and useful, but customer experience should be the focus.

This is a fantastic opportunity for product manufacturers that never before had direct and immediate access to the way their products are used in the field. There is no comparison between lab experiments observing users of a product, or analyzing data from a survey on product usage versus when you can get access to realtime data collected from a device in the hands of a user. And even better, data from ALL the users of ALL your devices where you can simply run some tests to compare feature adoption by users of different software versions.

Faster learning cycles

I think, however, that there is another, and more important, question than any of these: how do I learn from failures, to improve or even totally change the product for next time?  IoT is fundamentally changing business models. With usage data, it is possible to see how the product is being used, and start to provide services around it. Predictive maintenance may be the most obvious and best-known example, but there are likely to be many others emerging over the next few years. This service development model means that customers can be sold services around existing products to improve their experience, as well as to improve the product itself.

This is important because there are dangers to trying to improve products after initial launch: unlike software, you can’t just roll out an upgrade to a physical product. And you can’t do a product recall every five minutes. Sooner or later, customers will get tired of being expected to buy a new version. Apple is pushing hard at this particular boundary at the moment; changes to the hardware around the latest iPhone were not greeted with enthusiasm by the tech press.

Hearing the voice of the customer

Using data to ‘hear’ the voice of your buyers is, in my view, likely to be one of the most important uses of data science in product development. I am confident that it will be the key to improving products and services alike. And with IoT, the availability of realtime insights into specific behaviours will change the discipline of product design.

Suggested reading: White Paper - The Internet of Things: Marketing’s Opportunities and Challenges



About Author

Mathias Coopmans

Business Development Manager

In my work I combine business curiosity with technical skills. I am passionate about big data and business analytics, and help companies across South-West Europe create value from their IoT initiatives. Often pioneering and innovating but always seeking for ways to integrate IoT into their business, both on the technical side and at the strategic level. The role and impact of data and analytics on our daily lives fascinates me. Outside work you can find me tinkering with 3D printers, programming small devices and in winter on the ski slopes. Follow me @macoopma

Leave A Reply

Back to Top