Founded in early 2018, the US-based firm launched its artificial intelligence (AI) insights program earlier this year. The AI program tracks real-time online engagement and opinion around beauty products, scanning publicly available images and video content from major brands, retailers and influencers. Advanced image recognition technology identifies exact product SKUs and links this back to consumer engagement and reaction, enabling granular insight on how specific products are perceived by specific consumer groups.
Predictive intelligence for product development
“At its core, what Cherry Pick is building is what we call predictive intelligence for product development,” Justin Stewart, co-founder of Cherry Pick, told CosmeticsDesign-Europe.
“The goal is to enable a brand, manufacturer or retailer to actually measure how much demand consumers are showing for products before those products actually exist. From a holy grail standpoint, it’s like ‘before I allocate all these resources, is that actually even a product category that this specific target audience would want to buy?’”
The aim was to create “predictive” and “demand-driven decision making” that responded to real-time consumer reactions across the market, he said.
Beauty bytes: consumer, category, product
The three stages of Cherry Pick’s AI analysis enabled businesses to identify: the archetypal consumer; the categories suited to this consumer; and the product that should be built for them.
“Where this gets incredibly exciting and valuable is the fact that you can get down to measuring demand on product level, SKU level,” Stewart said.
“…There may be 500 mascaras that are currently on sale, for example, and actually all of those mascaras have different product attributes, different price points, different shades, different sizes, different claims. So, by looking at many, many consumer demand signals across every mascara that exists, you can see that demand and funnel that through to find out what are the attributes of those mascaras responsible for driving demand.”
Stewart said this was even more relevant in today’s beauty world, where lead times had become shorter and launch costs considerably lower.
“The cost of launching a new brand has gone down and what’s risen in importance, in terms of building a defensible brand, is getting distribution as quickly as possible to consumers,” he said. “This is doubly important in a world where brick and mortar is struggling.”
Time-sensitive consumer insights?
Asked if Cherry Pick’s insight could be irrelevant by the time a company developed and launched a product, Stewart said no, particularly for broader insights and considering target clients were start-ups.
“Something high-level like ‘this audience shows lots of demand for hair care products’ – these are very stable and forward-looking, longer trends we’re confident predicting further out,” he said.
Of course, with deeper insights like identifying the best finish for a specific product, at a certain price point, there was a faster turnaround needed and, in these instances, he said Cherry Pick recommended brands tracked preference movements closely. Alternatively, these finer insights could be used for things like on-pack claims or marketing messages – decisions made later in the development stage, he said.
“The way people are thinking about how to use Cherry Pick AI is an insurance policy for brand launches,” he said.
Building AI-driven brands
Stewart said while Cherry Pick would continue to work with existing brands, manufacturers and retailers, another side of its business was developing new brands from scratch – an aspect it would invest more in over the coming years.
“There’s two parts of the business: one actually building brands and two doing this data insights layer.”