Winning IoT Isn’t about the Analytics of Things – It’s about the Analytics of Your Customers

A convergence of improvements in some key technologies – wireless communications protocols, embedded sensors, mobile processors, batteries, memory, and internet data transfer protocols, among others – with increased smart phone penetration and engineering innovation to cram more and more computing power into ever-smaller spaces has enabled the genesis of an industry that produces internet-connected devices.  This “Internet of Things” (IoT) has been and will continue to be a “buzzy” topic for tech pundits for some time and I’m completely buying the hype – I have very little doubt that connected devices will fundamentally change transportation and infrastructure, our homes, and our workspaces over the next couple of decades.

The Analytics of Things (AoT)
Thomas Davenport, one of the most brilliant minds in contemporary business analytics (and author of one of my favorite books, “Competing on Analytics: The New Science of Winning“) has popularized the term the Analytics of Things (AoT) to refer to “new” practices of analytics that are enabled by the streams of data created by IoT devices.  These practices are focused on devices’ (and their services’) ability to use analytics to serve their owners in two contexts: using external “analytics” to drive device output when the devices are connected to the internet and having the ability to take local sensor inputs to control the devices’ behaviors when they aren’t connected to the internet.  As an example (which I lifted from this article), the Nest thermostat provides insight to its owner on usage when connected to the internet, but also has “internal analytics” which triggers thermostat responses based on a home’s occupancy (or lack thereof).

While these are interesting and important concepts when it comes to IoT product design, the companies that are going to “win with analytics” in the consumer IoT space aren’t going to do so purely on sensor data, but on an overall customer data strategy that informs both their product development and marketing activities.  The major missing piece of the discussion of the AoT as it relates to the consumer device sector is how these devices will enable feedback loops to the business, allowing the companies that build them to leverage device data to drive business outcomes.  The ability to execute on customer data will make or break a lot of device manufacturers.

Why are the stakes so high?  Consumer electronics is already tough business with risks on platform bets, low-cost overseas manufacturers poised to attack margins, and a complex and competitive retail environment.  Layer in on top of that the increased cost of maintaining connected infrastructure for your existing customers, increased spending on product innovation, and shifting platform requirements and the business does not work if you cannot sell devices to new customers, up-sell existing customers into premium services, and/or drive repeat device purchases.  The winners in the sector will be the companies that sell the most devices and services and “AoT” as it is being currently defined is a very small piece of the data picture that will enable marketing efficiency and product excellence.

The Customer Analytics with Things (CAwT)
In stark contrast with the Internet of Things revolution, there has also been a convergence of technology and practices in customer analytics, particularly within the discipline of digital analytics, that represent an evolutionary step forward in the ways that we collect, analyze, and act upon our customer and marketing data.  Integrating app usage, device activation, and even sensor data into a well-engineered data ecosystem is, at this point, table stakes for an IoT company to be able to compete on analytics.  You could label this as the “Customer Analytics with Things,” but that is silly (and poor English) – device usage data is simply another data source in what would be, for a typical consumer electronics company, an already rich customer data environment.

Let’s look at the data sources and enabling data technologies that an IoT company should leverage:

Clickstream Analytics – Since most IoT devices are controlled via company-developed web and native apps, clickstream analytics tools create one of the key data sources for an IoT company to develop an understanding of the who, what, when, and why of their customers’ interactions with their devices.  Clickstream tools, like Google Universal Analytics or Adobe Analytics, record these interactions and aggregate them across useful segments for the purpose of developing insights into how these interfaces are being consumed.  Clickstream tools are also critical to develop an understanding of the effectiveness of digital paid/earned/owned media channels as they relate to eCommerce, assuming the IoT company is selling direct-to-consumer or layering in additional paid services.

Voice of Customer/Customer Support/CRM Data – Clickstream can only get you so far.  To truly understand how the devices are being used, user task completion success/failure, the general use cases for a device, and many other questions, there is no better source of data than the end-user.  A smart IoT data strategy encompasses both Voice of Customer (user survey) data with customer support data to develop deeper understanding and model user segments that are based on emergent product use cases which can be extended across all users.

Retail and eCommerce Data – This not only encompasses direct-to-consumer sales, but also includes data on device sales through other retailers.  Depending on the retailer, this data provides individual-level to channel-aggregated device sales statistics.  By tracking which devices go to which retailers, individual device activations can and should be tied back to the originating retail channel, which would be useful for analyzing channel effectiveness as well as the secondary market for the devices.

On- and offline advertising data – Collecting and integrating media data is useful to contextualize clickstream and retail data.  Knowing when and where media is in the field will help an IoT company attribute first- and third-party (retailer) media spending effectiveness.

Device and Sensor Data – The “icing on the data cake” for an IoT company is device and sensor data.  Although some find it creepy, collecting and using this data for marketing and product development could be an important data segmentation tool for an IoT company.  Since activations and connected features are usually driven through an external (to the device) server, the device-level data can be used to directly or predictively place device users into user segments based on how they are using the devices.

Data Aggregation Technology – These data sources can deliver a lot of value on their own to an IoT company’s marketing and product teams, but their value grows exponentially with integration.  Aggregating customer data sources into a warehouse permits organizations to have greater control over the definitions, access, and usage of the data (data governance) as well as providing easier integration with other tools for reporting and analysis.  Due to the cost of infrastructure and the difficulty of integration, data warehouses with business intelligence tools used to be the realm of companies with large IT budgets.  Now, they can be developed and deployed at a fraction of the cost thanks to technology like:
Amazon Redshift, Google BigQuery, and MS Azure SQL Data Warehouse for cloud-based data infrastructure,
Domo for cloud-based data infrastructure with a reporting interface,
Looker, Periscope, Mode, Chartio, Sweetspot Intelligence for querying and reporting,
Tableau, Excel, Power BI, and R for data discovery and statistics,
and Segment or Databox for integrating data sources across all of the aforementioned sources (along with data API’s).

In closing, device data should be the icing on an already sweet, sweet data cake for companies in the IoT space.  Success in the sector is going to rely on effective marketing and innovative product development that will be enabled by a holistic customer analytics strategy.