How to Avoid Drowning in a Flood of IoT Data




At LiveWorx 2016 Harvard Business School Professor Michael Porter, PTC CEO James Heppelmann, and iRobot CEO Colin Angle provided a guide to how companies should think about their customers, their value chain, and their organization in response to the vast amount of data that the IoT will generate.

Facing the data challenge

All three participants spent time on one key issue: the choices organizations need to make to respond to the data streaming back from their smart connected products and relate effectively with their customers.

“The scope and scale of the impact on the organization is immense,” Colin Angle said. “It touches virtually every area of the business.”

Where the data comes from—and where it should go

Data streams are generated by factories, marketing, sales, and the products themselves, and everyone is still learning how to gain value from this vast amount of data. Angle observed that there is no one logical owner of each data stream, no matter how much easier it would be if there were.

Jim Heppelmann pointed out that engineering skills will need to be retooled to meet the requirements of these data streams. They currently know mechanical, electronics, and perhaps embedded software, but they don’t yet know cloud architectures, security, and all of the other considerations that are now key to smart connected product success. And that’s where IT needs to step in.

“The relationship between IT and engineering is a strange one,” Heppelmann said, “perhaps not even a healthy one.” Now both have to work together to develop and create these products, which is a cultural, management, and procedural challenge.

Product as relationship

“What smart connected products do is they generate a treasure trove of the most relevant data that would ever want to have,” Michael Porter said. “’What’s happening to this product? Is it working, is it on, is it off, is it failing, how is it failing?’”

This data can be combined with what companies already have, such as CRM data, as well as relevant external data, such as local weather conditions. 

Early smart connected products just piled up data that no one looked at. It takes a lot of work to turn data into information. But who runs what Heppelmann called the “data refinery”, and ensures that marketing, product development, manufacturing, customer service, and all the other functional areas get the useful, actionable data they need, and only what they need?

Introducing IT to the product

Angle said IT is the logical collector and manager of the data, but has traditionally been fairly divorced from products. So iRobot has brought mechanical, electrical, and software engineering together into unusually integrated product teams, creating a new competency within IT: “product IT”. This has diffused knowledge of IT through an existing cultural mechanism where these product teams get together to solve problems.

It is working, but Angle is not sure if this is optimal. He thinks it may be succeeding on the backs of good behavior, rather than as a result of a well-designed structure. Companies that have a shallower pool of high-quality talent should consider Angle’s observations carefully, and recognize how much thought will need to go into coming up with workable structures, institutions, and corporate cultures to solve the problem of data integration while continuing to focus on their core competencies.

Products now report back on how well they are working, and, perhaps more importantly, how they are getting used. Is the customer getting the full value of the product? If it’s being underutilized, whose problem is that? Heppelmann pointed out that the responsibility didn’t seem to lie in any existing areas: the product worked to spec, so engineering didn’t necessarily get called in, and it didn’t have a problem reported by the customer, so customer service wouldn’t handle it.

And how do you develop these smart connected products when the natural cadence of mechanical development is considerably slower than that of software development? Angle says you need to develop a proper library of hardware surrogates, so that the software people aren’t slowed down. Managing these development cycles will be an increasing challenge.

The slow emergence of the new organization

Porter focused on the changes IoT will require of the organization. It is a fundamental principle of organizational design that different functions should be separated into groups that can focus on them, he said. But you always need to integrate across these various functional groups. With the changes wrought by the IoT, the balance between integration and differentiation has changed dramatically. You can’t just hand things off from one group to another. They need to work together consistently. Porter mentioned that when he and Heppelmann started working on analyzing the effects of IoT on organizations, they did not have enough appreciation for how big a deal the organizational issues were, as compared to the technical or conceptual issues.

Technology tends to be more visible than organization, easier to describe—and easier to modify. But success and failure in the new IoT-based market will depend on how effectively organizations manage to transform themselves while retaining the value they already have with their customers.

LiveWorx 17 will be held on May 22-25 in Boston. Registration is now available online at http://www.showreg.net/LWORX1706. To receive a discounted All-Access pass for only $695, register using the promo code ATTEND17 by August 15.


Related Articles: