As ones and zeros eat the world, data science is the new process of innovation. Good product managers have always been data-driven, bringing the voice of the market to bear on the new product development process. This was traditionally done by analyzing internal data on existing customers and data generated by a variety of market research tools. The explosion of data available on the Web has changed that by adding a vast new source of intelligence on potential and existing customers, competitors, and new technologies, materials, and approaches to product development.
For many product managers, more data has meant much more “noise,” confusing rather than clarifying analysis and decisions, as the external and (mostly) unstructured data is much more difficult to manage, manipulate, and analyze than the internal, structured data of old. Only ten years ago, social networks barely existed. Today, they are a rich source of data on what works and what doesn’t work in the marketplace, what customers like and what they say they want. Only fifteen years ago, it was difficult to find data on your competitor’s product plans, new hires, and patent filings. Today, all of this and more is accessible via your preferred search engine. But how do you mine and make sense of the exabytes of digital breadcrumbs that individuals and corporations are leaving online?
“Companies rely a lot on their ability to launch new products as a growth engine,” says Gil Sadeh, CEO of Signals, “but there are not enough innovative ways today to bring products to market. We thought that big data is a great opportunity for the space that’s called innovation.”
Big data analytics is the killer app for new product development and Signals has developed the app store for innovation , for successfully bringing new products or permutations of exiting products to market. Sadeh, co-founder Kobi Gershoni, and their team have spent three years developing a technology infrastructure that would allow them to pull into one unified environment large volumes of data from numerous sources, in real-time.
Once the infrastructure was ready, they decided to focus on a specific business problem that could benefit from the intelligence gathering, modeling, and analysis capabilities they have developed.
To identify the right business problem to focus on, they followed their own prescription, running both top-down and bottom-up analysis. The top-down analysis found a “white space,” a business domain requiring their expertise, one that could use more rigorous and comprehensive data collection and analysis: new product development. “Bringing new products to market is a well-established process with budgets and measurements,” says Sadeh.
Similarly, conducting a bottom-up analysis, looking at the successful engagements they had with early customers over the course of the first three years of their company’s existence, “we found out that companies like P&G and J&J,” says Sadeh, “came back again and again to us with questions around new products, new technologies, decisions that are part of the new product development process. We realized that this is where we show clear value.”
Once the specific business opportunity was defined, Signals embarked on developing and offering its customers an “app store,” a library of generic applications geared towards different stages of the product development process. These applications are customized by Signals’ customers to their specific situation and needs, by filtering the data by selecting different parameters and weights, and by making changes to the ontologies and taxonomies suggested by Signals.
This process brings together the customer’s domain knowledge and specific approach to product development with Signals’ data science capabilities. But even before that, in the development of its tools, Signals makes sure to incorporate domain knowledge. Three types of Signals employees develop the application: data scientists, subject matter experts, and business analysts. They start with the business analyst defining the business objectives, actions, and decisions. Then the subject matter experts help develop the relevant ontologies and taxonomies and define the relevant data sources. Finally, the data scientist takes this input and develops the application and its associated models and data sources.
“Our strongest asset is the ability to model in a dynamic way different business problems and create big data applications that are very personalized and actionable for our clients,” says Sadeh. “We take a business question, build a model to answer this question, and then we build a data model that allow us to bring the data into one unified space that allow us to connect the dots from very disparate sources of data.”
An example Sadeh likes to use to illustrate the power of data science in new product development comes from the medical device business. Looking at data from social networks, blogs, and online forums for new insights regarding stroke patients and their rehabilitation, a medical device company found an unmet need in the market: most of the patients were complaining about lack of products dealing with their legs, as opposed to products helping with the upper body, arms and back. Further analysis driven by Signals’ applications confirmed that competitors were not planning to provide a solution to this need anytime soon. Finally, Signals helped identify the set of technologies that could help bring a solution to market—in this case, robotics.
Signals’ business model is based on three revenue streams: Customers pay up-front for the license to use an application; they pay for each data cycle, when they update the application with fresh data; and they pay for advanced features, used mostly by their own data scientists to conduct their own analysis.
Given the projected continued growth of more data, from more sources, that could be used for product-related decisions—the latest entry being the billions of connected Internet of things (IoT) devices and sensors—the level of deafening noise will soon be turned up to eleven. Identifying and connecting the most relevant dots, Signals turns the volume down to where product managers can hear the voice of the market and make the right decisions.
This article was written by Gil Press from Forbes and was legally licensed through the NewsCred publisher network.