Audience preferences are constantly evolving. So why not ads that evolve in real time, too?
No, really.
Media agency Multiply claims to have “self-learning ads” and proprietary AI software that tracks real-time signals to understand performance and brand sentiment. It develops new versions of ads for B2B companies based on the most recent data about customers, according to CEO and Co-Founder Matt Jayson.
On Wednesday, the company came out of stealth mode and announced a $9.5 million seed-funding round, led by early-stage VC firm Mayfield, with participation from Sorenson Capital. Max Mullen, the co-founder of Instacart, also contributed, alongside Josh Woodward, Google’s head of Gemini and Google Labs, and executives from HubSpot and Braze, among others.
Striking the balance
With only 13 people, Multiply’s team is lean. A portion of the funding will go toward hiring, with a focus on strategists.
For starters, Multiply is only helping brands advertise across LinkedIn and Google Ads, but plans to use the funding to expand its product to more ad channels, with Meta and Reddit on the horizon.
Notably, Multiply has been “in talks with OpenAI” to introduce ads on ChatGPT “as early as possible,” Jayson told AdExchanger.
But despite being an AI-powered startup with plans to advertise on chatbots, he emphasized that Multiply isn’t asking its customers to “surrender to the AI.”
A lot of B2B companies feel like they “need to adopt” AI, said Jayson, but concerns about cost and brand safety hold them back. Multiply aims to address those concerns by prioritizing human involvement throughout the process.
Clients want humans to be on calls with them and to know that their marketing strategy was given “real human attention,” said Jayson, who added that people review all versions of ads before they get sent to a client.
Live and learn
That human touch is especially important in B2B marketing, which faces a set of challenges distinct from B2C.
For example, a lot of the data that’s most valuable to B2B marketers lives “in the head of the sales team,” Jayson said, and comes from direct conversations between customers and sellers.
Multiply taps into that data through direct integrations with sales call recording platforms. The AI can determine the “hot spots” of a given conversation, said Jayson, like a comment about why a customer chose this or that brand over a competitor.
Multiply also integrates with CRM systems, collects customer testimonials to get a fuller picture of a B2B customer’s own clients and pulls in data from that client’s recent customers. This information is often found on their LinkedIn page, like specific skill sets and job titles.
The AI then generates sometimes several and sometimes dozens of versions of ads, which are approved by Multiply’s human team and pushed to the customer for further approval or revisions. The ads might feature a specific word that was mentioned frequently in positive sales calls.
And this is where the “self-learning” aspect comes in.
The platform continuously tests new audiences, messages and creatives so results keep improving over time instead of decaying. Different audience segments get slightly different ads.
For instance, Jayson said, a salesperson and a CFO would probably be persuaded by pretty different language. As ads run, Multiply tracks which versions perform best and which messaging leads to the highest click-through rate.
Each iteration informs the next, which creates a feedback loop focused on performance, he said.
Multiply’s platform is powered by “insatiable ad agents,” Jayson said, which are constantly optimizing and improving campaigns. For the agents, he added, there’s “no such thing” as being satisfied.
