A TV show in the late-1990s called “Early Edition” was based on the premise that a guy received the next day’s newspaper a day early. The hero then had 24 hours to set things right.
That scenario is enough to make any marketer salivate. Even better if the time frame was longer. Marketers have been trying to crack the code on hit products forever with limited success. Blame the economics of standard product development and marketing. Product innovation and mass marketing is often supplier-led vs. based on what consumers actually want.
The flameout rate for new products ranges from 75% to 95%. Smart employment of AI and machine learning could improve those odds greatly, not just by identifying trends but by prioritizing them and isolating the ones with the greatest chance of success.
How machine learning and AI can improve the odds
While no one can glimpse the future, looking at the data closely enough will get you far enough along that you can beat the odds.
With product launches, the variables include the progression of trends across geographies and market categories and from early adopters to mainstream, the impact of large consumer trends (e.g., sustainability, fair trade), the impact that influencers have on trends becoming mainstream, the alignment of consumer perception and actual needs, and more.
There is too much of this type of data for a human being to make sense of, but machines can find patterns in huge amounts of data.
In a recent example, K-beauty skin care, a major trend in cosmetics, began in South Korea and then gained interest in Europe before it became a hot topic in the United States. In that instance, a U.S. marketer could have looked at the data and seen a trend. In the CPG industry, trends travel across geographies and subcategories (food/beverage, health/beauty). In this case, an AI system might have predicted the worldwide spread of the K-beauty trend early enough for a marketer to capitalize on it.
While that example looked at growth across the globe, another indicator of strong growth is the type of people who are discussing a product category. In the food and beverage category, topics like dairy-free, vegan, and organic have all become big trends. A recent example in this category is the discussion among trendsetters about pea yogurt, and the data shows the discussion is unusually strong. That indicates that pea yogurt may be just a few months from going mainstream.
The difference between identifying trends and prioritizing them
Identifying trends isn’t new. The difference that AI and ML bring is that they can make sense of the noise and identify the most promising products (like pea yogurt) by showing how such trends evolve and what influences their growth. That’s because AI can analyze data and recognize patterns that resulted in successful past product launches. It can also identify early on what trends are not so hot (like savory yogurt).
In 2018, blindly guessing what consumers are going to like months from now doesn’t make much sense. Nor does bringing a finalized version of that product to market and trying to make it fit.
What does make sense is using all the available data and computing power to make product launches as successful as possible. It also provides a good roadmap for taking the trends to market (e.g., how to market and advertise pea yogurt to consumers). That’s the closest thing we have to seeing into the future and it’s getting closer all the time.