Squadhelp adds AI to help invent successful brand names
With AI, the Chicago-based platform is adding visual search, performance tracking and name generation to its crowd-sourced naming platform.
Naming a product or a company can make or break the effort. Traditionally, it has involved branding agencies engaging a variety of experts and test groups.
Chicago-based Squadhelp employs a crowd-sourced platform to find exactly the right moniker, after which it employs various verification and testing tools. It describes itself as the largest such naming platform, and it now says it is the first to employ AI.
Since it was founded in 2011, founder Darpan Munjal told me, the platform’s naming process has involved a brand launching a naming contest after supplying info on the company and, if appropriate, on the product. About 75,000 users are registered on the platform, and the number of proposed names they can each submit is dependent on how many of their previous submissions have been accepted into higher contest rounds.
The brand rates the names, with a typical contest going through 500 to 1,000 names and costing $199 to $999. The person who submitted the core name, even if it goes through several generations, receives the winning payment from the brand.
Now, with AI, customers can first employ a visual name search. It is a quiz of various possible names that could be appropriate for that industry, offered with a font selection and a background image. Those names –– from a collection of about 5,000, shown sequentially in groups of 50, up to about 300 — come from staff- and platform-curated categorizing of the entire 5 million previous entries.
The AI learns about the brand’s likes and dislikes as it responds to the visual name search, and then AI makes suggestions from those names and others in the total collection. If those suggestions don’t hit the spot, the brand can then conduct a contest with the crowd.
Alternatively, the AI can use the reactions to the visual search and create a name from scratch, with Squadhelp’s new Name Generator.
Munjal said his company has conducted about 17,000 contests since its founding, with a 90 percent success rate of finding a name a brand wants. Before launching AI, he said, the platform only ranked the quality of participants in finding usable names, but now it is tracking the performance of individual names. Each contest in the past, essentially, started over.
With a month of beta testing, he said, the AI-powered platform now appears to be successful, through one means or another, about 96 percent of the time.