How Artificial Intelligence Is Helping Companies Identify And Nail New Product Opportunities

Artificial intelligence is helping companies identify new product opportunities and iterate quickly to get them closer to perfection in a number of different ways.  What AI does well is:

1)  Search through mountains of data quickly to report patterns that can be analyzed for white space new product and service opportunities, and competitiveness.

2) Iterate new product or service concepts or prototypes through trial and error virtually and simulate consumer response, in a fraction of the time and at a lower cost than real-world testing.

3) Predict demand for product offerings and adaptations on a local market basis by analyzing search and purchase patterns in each geography.

What follows are some of the most exciting examples of how artificial intelligence is being used at different stages of the new product development process, and different approaches to deploying the technology.

IntelligentX Beer Products Creation

Jordi Torrent, Open Innovation Manager at Damm, a multi-national brewery based in Barcelona with activities in other sectors including logistics and distribution, shared how AI is being utilized in the beer industry.  British start-up IntelligentX developed a line of beers using AI.  They worked with the machine learning firm Intelligent Layer and the creative agency 10X.  Here’s how it worked.  A bot asked consumers questions regarding flavor preferences about different beer prototypes to get feedback.  Then an algorithm learned from the feedback how to optimize the products and how to ask even more helpful questions.  The learning was done on a continuous basis to keep refining the beers.  The AI gathered lots of data and interpreted it in a fraction of the time it would have taken in traditional product development.  The 4 beers, golden, amber, pale and black, were sold at UBREW, an “open brewery” in London where members could brew their own beer. The AI beers were continuously improved 11 times over a one-year period, based on recurring feedback.

NetBase Quid Market and Product Need Identification

The AI company Quid, that just merged with social listening firm NetBase, uses AI to spot opportunities in virtually any market you can think of.  Quid founder Bob Goodson, now President of NetBase Quid, built a unique database of 2 million venture and angel backed startups around the world with descriptions of each firm’s product offerings. For each sector, Quid visually maps the new product and service ideas that have emerged, to show where there might be as yet untapped new business opportunities.

Now, with the social listening capabilities of NetBase, NetBase Quid can “hear” what consumers are saying about these unmet needs and dissatisfactions with incumbent offerings, to infer what a better product or service would look like.  On a call with me last week, Goodson generated the following three analyses in a live demo. In just minutes I watched the AI create reports and insights that would have taken a traditional consulting team several weeks.

The chart at the top of this article shows clusters of start-ups in the sensors and wearable devices market, enabling firms to spot where there are gaps. The chart below shows startups in each sub-sector of the same market, how recently, on average, those companies were launched, and how much investment each sub-sector has attracted.  The data indicates how competitive and developed each sub-sector has become, and whether it’s being heavily invested in.

The last chart shows the recency and amount of investments in each sub-sector on a quarterly basis. It shows, for example, that patient and workplace monitoring sensors and wearables recently attracted considerable investment.

CPG Product R&D and Consumer Testing Acceleration

Start-up Turing Labs (, backed by Y Combinator, Moment Ventures, and Eric Reis, Author of The Lean Startup, is using its AI software to supplement, expedite and reduce the cost of R&D for CPGs and retailers who make household products and foods.

Their AI collapses the time to thoroughly test new products with consumers from as much as a year and a half, to approximately 6 weeks. The software analyzes historical product and survey testing data and chemistry principles and uses machine learning to simulate experiments on potential new products in categories such as soap, laundry detergent, shampoo, salad dressing and beverages.

Turing also quantifies “human data” (the knowledge of the R&D and marketing research teams), and adds this to the mix, to predict how changes in ingredients will impact formulas, costs, and consumer interest in the products.  Iterations are quick and virtual, enabling fast optimization and subsequent learning, rather than the much slower and more costly testing and iteration with physical, human subjects. During situations like the Coronavirus quarantine, when many R&D labs were closed, the software could partially replace large laboratories and consumer tests with lots of researchers working side by side, with machine learning programs on computers from home.  Turing is already working with several large multinational CPG and retail firms.  Accuracy of predictions are validated through control test data and/or live control tests.

Using AI In Design Research and New Product Opportunity Definition

Smart Design, the strategic design agency, uses AI to complement its human-centered approach.  AI adds another lens to a research tool kit, allowing teams to uncover different insights and spot behavior or opportunity patterns that might not be visible using traditional quant or qual methods. AI is biased differently than humans and can find less expected or contrarian patterns. It should be combined with human insights.

It helps if there are large data sets available, that are rich or complex (like voice or image based), or very dynamic, changing quickly and often.  Available tools include:

  • Open Data provided by institutions, governments or cities
  • Subscription services, such as Statista
  • Client owned consumer data including purchase, use, and social behavior
  • Rich, complex data from primary research (activity logs, transcripts, videos, and images)

As a great example, Smart Design conducted an open source data analysis of accidents in bike lanes from NYC, Boston, and SF, to examine how bike lane design effects the rate of accidents, which turned out to be surprisingly substantial and instructive for future designs.

AI can also perform text-based analysis of transcribed interviews of a large number of respondents using analytics software like Voxpopme or Luminoso. The AI finds patterns between phrases and concepts in the responses, revealing patterns that might not be seen by humans reviewing the individual interviews.

Take-aways for CMO’s

For many companies, AI is a whole new, unknown world. Because its undoubtedly where new product development is heading, it’s worth starting to experiment with. For firms that figure out how to use it well, AI will be a competitive advantage in finding opportunities, iterating faster and more accurately, and saving money in the process. It’s here to stay and will continue to get better and easier to work with.


Original post:–nail-new-product-opportunities/

Leave a Reply

Your email address will not be published. Required fields are marked *