Why AI Is The Next Frontier Of Product Innovation

Why AI Is The Next Frontier Of Product Innovation

Cofounder and CEO, overseeing Zappi’s business and growth strategy across regions.

In the early 1980s, Atari shelled out $21 million for the video rights to Steven Spielberg’s blockbuster film E.T. the Extra-Terrestrial. The company spent another $5 million on a flashy marketing campaign. Unfortunately, the game was a flop, selling only 1.5 million units. The development timeline was too tight, and the design was confusing. Atari wound up burying the remaining 2.5 million copies in a landfill.

In the late 1970s, Clairol launched a new hair product called Touch of Yogurt shampoo. The goal was to capitalize on the growing trend of including healthier ingredients in beauty products. But using “yogurt” was confusing for customers. Some thought the product was a food item, and they ate it.

These examples are certainly extreme, and they are also from another era. So maybe the situation is much better today?

Not really. There continue to be plenty of misfires.

Just look at Walmart. Last year, the retailer wanted to honor Juneteenth with a new ice cream. The packaging included Pan-African colors and visuals of hands at the bottom of the container. One had the peace sign, and two others were high-fiving. But this looked more like a shameless—and tone-deaf—move. As should be no surprise, there was an uproar on social media, and Walmart apologized.

Understanding Product Launch Failure

There are plenty of reasons for the abysmal failure rates for product launches. Just some include lack of innovation or differentiation, lack of understanding of the customer needs, lackluster marketing and suboptimal pricing.

Harvard professor Clayton Christensen estimates the failure rate for new product launches at roughly 80%. Research from the University of Toronto claims that roughly 75% of new grocery products fail each year, and Catalina projects the failure rate for consumer packaged goods at 80%.

But all these are really about insight failures. In other words, there is often insufficient consumer feedback to gauge the prospects of the new product before launch. As Steve Jobs once said: “You have to pick carefully. I’m actually as proud of the things we haven’t done as the things I have done. Innovation is saying no to 1,000 things.”

Granted, this is easier said than done. Few companies have product geniuses like Jobs on the payroll.

Despite failure being an intrinsic part of innovation, it’s become paramount to understand what resonates with consumers (and why) before products hit the market. But it is still the norm for marketers to use guesswork and hunches. They are awash with post-launch data such as sales, social media and CRM data. But before launch, they lack quality insights to inspire, validate and optimize their next big product innovation.

Today, enterprises need to work faster and in more agile ways to stay customer-centric and innovate at the speed of consumers. Consumer data needs to be available immediately. Businesses need answers to their business questions without barriers so they can adopt an ongoing iteration cycle to corroborate so they know they’re on the right track without impacting agility. But traditional market research isn’t fit for a technology-driven world: It’s too slow, too costly and too analog, serving as more of a blocker than an enabler of winning innovation.

The business world has evolved to embrace technology, but the insights world has been left behind.

The AI Factor

The answer lies in AI-driven consumer insights that can help change how companies create products.

By adopting AI-driven consumer insights, businesses can gather and analyze data more effectively throughout the product development process, allowing them to quickly nix the bad ideas and have more time to validate and optimize the promising ones—whether claims, packaging, tag lines or flavors. What’s better, they can do it at scale.

Let’s take an example of a global CPG company to see how AI can dramatically improve product launches. A team has 20 product concepts for its new soda flavor, and they need to validate their final decision on which concept to take to market.

Simply put, the CPG company has two options: retain a traditional market research agency or use AI-driven technology.

For the former, it’s likely they will end up paying $20K-$30K for some useful data sitting in a PowerPoint deck that will, at best, take six to eight weeks from start to finish. For the latter, research results will come back in closer to six to eight hours and, in most cases, at three times a lower cost. This means businesses can test more ideas more frequently and increase their confidence in making critical business decisions at critical times.

Not only that, technology can rely on other historical data sources to identify which concepts have the best commercial potential—such as consumer trends, research of competitors and social media sentiment—and use sophisticated algorithms and generative AI to produce insightful reports that provide impactful calls to action.

Now it’s true that online panels can have skewed results and biases. Yet AI can help mitigate the problems and elevate the quality of the data, weeding out bots or fake and low-quality respondents.

Another critical advantage AI offers is longitudinal benchmarking to understand how new product ideas compare to market peers in real time. This goes beyond the generic survey questions of “what” and “why.” Smart tech compares answers to previous products created by competitors, helping improve predictions of success.

And at the core of the AI promise is its flywheel effect: More adoption generates more data, building a learning loop that means every time tech is deployed, it gets better.

The Future Looks Bright

The days for traditional market research are numbered. Once a technology reserved for the most elite and disruptive companies of Silicon Valley, 68% of the most digitally advanced companies have already integrated AI into their product development, according to PwC.

I firmly believe this percentage will only grow. This combination of immediate availability and more predictive consumer data eliminates one of the major challenges enterprises face in reducing the number of product failures because they’re continuously learning.

In the next installment of this series, we’ll take a look at specific tactics to help business leaders get started with AI to help with product innovation.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?