How to Build a Disruptive Product

Sameer Kalwani

Director Product Management at Amazon Lab126



When incubating my previous company, a surprising product opportunity in the Industrial IoT (Internet of Things) space emerged. At the time, a lot of people were building hardware devices to collect data while others were creating analytical dashboards to better analyze that data. Most of the companies in this space were stuck in a ‘pilot purgatory’ -- they never got out of the pilot or proof of concept phase with their customers -- and never made it to production. While customers needed to collect and analyze IoT data, it seemed that something else was troubling them more.

Actions taken

There were enough market studies highlighting the opportunity, and enough evidence with companies pursuing several pilots in parallel. Rather than interviewing a great number of people we devised a quick mockup that would allow us to understand what the burning issues in the industry were and what we could do to solve them. We recruited a couple of companies to get analytics for “free” and brought in a few Ph.D. candidates who knew the space to perform data science work on the “free data”. This allowed us to quickly develop an analytical result behind the facade of a product. In pulling the insights together for these companies, identified where our customers’ exact pain points were.

We learned that the data was widely available and we could easily get it in a day. We also learned that the analytics were actually valuable and could help companies increase their ROI dramatically. But the biggest issue was putting context around the data so an analyst could interpret an insight quickly at scale. Building new sensing hardware or dashboard didn’t solve the problems, understanding it and delivering it in a digestible form to the customers was what we should focus on.

As we began building a data contextualization SaaS solution for Industrial IoT data, we had to find a way to test what we were building throughout the development process - ensuring each feature was meeting a market need. We brought on our own data science resources to do the proof of concepts as we kept building more products to enable them to help them understand the context faster and more easily. As a result, the people using our product were our own people, enabling us to learn faster. Soon we were able to share the tools with our customers, so they could begin to do the work themselves. We began to get the key insights from helping our customers, sitting side by side with them, and walking through the process with them. They subscribed to our service and eventually began using the software on their own without any of our people there.

Lessons learned

  • To come up with a disruptive product you have to work backward and put yourself in the customer’s shoes. Don’t just test your hypothesis, but test your assumptions in terms of the problems your customers are encountering. Understand a persona who you are trying to solve the problem for -- and not just a problem but a recurring problem. That will help you to find the right market, but moreover, someone who will buy your product.
  • Before you start building something, validate your hypothesis, and build something lightweight.

Be notified about next articles from Sameer Kalwani

Sameer Kalwani

Director Product Management at Amazon Lab126

Organizational StrategyDecision MakingCulture DevelopmentLeadership TrainingPerformance ReviewsFeedback TechniquesCareer GrowthCareer ProgressionSkill Development

Connect and Learn with the Best Eng Leaders

We will send you a weekly newsletter with new mentors, circles, peer groups, content, webinars,bounties and free events.


HomeCircles1-on-1 MentorshipBounties

© 2024 Plato. All rights reserved

LoginSign up