Marketers spend a lot of time and energy on manual, lower-level tasks. Even if the individual tasks don’t register as taxing – 10 minutes a day here, an hour per month there – in the aggregate, they’re like holes in a boat, draining marketers’ ability to focus on the creative and decision-making aspects of their roles.
Arcane, a marketing workflow automation startup, announced $5 million in seed funding on Thursday as it throws its hat in the ring to take on the problem.
The VC firm Accel is leading the round. But Arcane’s angel investors include Connor Hayes, Meta’s VP of product; Tom Blomfield, former CEO of British online bank Monzo; Anthony Yazbeck, WeWork’s former president; and Ilkka Paananen, CEO of Finnish mobile game developer Supercell.
Most of the funding will go toward hiring back-end and data engineers to join Arcane’s 14-person team, Ben Hacking, Arcane’s co-founder and CEO, told AdExchanger. The company’s core product, an AI copilot that Hacking described as a “ChatGPT experience plugged into your model marketing data,” is in closed beta. Arcane is planning a public launch of a self-serve offering at the end of Q1, in March or April.
Arcane’s clients have been mostly early to mid-stage startups in the UK and Europe, with a few US companies on a waitlist, Hacking said. They skew more toward B2B marketers thus far and tend to have marketing teams of one to eight people.
Talk to me
Three priority areas for Arcane’s workflow management are competitor research, ad analysis and content creation support.
For competitive research, marketers can fill in fields – company name, geographies where they want to do research and a specific timeframe – to pull in rival marketers’ creative and media. In January, Hacking said Arcane will introduce a chat interface that answers questions (e.g., “What are our competitors running on LinkedIn and Meta?”) in natural language.
Marketers can also ask which assets are performing better or worse based on different metrics. They might ask for a notification “when a metric has spiked in a particular direction that you care about,” Hacking said. If a marketer sees a massive impression dip on TikTok, it’s probably worth looking at their entire line of communication – ad campaigns, organic posts and channels – to see what might have influenced the drop as a “first level of investigation.”
To support content creation, Arcane built out a static world calendar of events with categories like news, sports, business and media. The tool might alert marketers when there’s a relevant sports match or event coming up. It can also tap into Google search traffic and do social listening to find trending topics, such as TV show finales, that relate to a target audience.
Data pull
Arcane relies on OpenAI as well as open-source LLMs to return responses to marketers in human-readable language, Hacking said. That conversational layer is crucial, but it means nothing if Arcane’s AI co-pilot doesn’t understand how to find the right answer to a marketer’s question.
“If you want to ask a question about a metric and a time frame, you have to build a custom tool that can interpret metrics and route that response to a specific part of your database to pull the concepts you’re talking about and the data associated with that,” Hacking said.
Arcane’s team is still building out the workflows, Hacking said, and its areas of focus might change. But, in the near term, “we’ll keep iterating until we capture enough of a marketer’s work that they are far less stressed.”
