Hightouch today introduced Hightouch Agents, designed to accelerate and automate a wide array of slow, manual tasks that hinder marketing efforts.
Hightouch Agents connect directly to a company’s data warehouse and marketing stack, offering visibility into the data behind marketing decisions, customer transactions, inventory levels, creative performance and more.
The initial release of Hightouch Agents focuses on the inefficient workflows that plague marketers and span teammates, agencies and channels. The agents are designed to answer complex questions with precision, Tejas Manohar, co-founder and co-CEO of Hightouch, told MarTech. That precision stems from their focus on context.
Hightouch added what it calls the “marketing context layer” to its agents. It consists of customer data, campaign details, including creative and spending, and brand information such as brand guidelines.
Hightouch integrates with 300 marketing and data platforms. Manohar said the new agents do more than activate data from those platforms — they consume context.
Purpose-built agents trained in marketing
He described Hightouch Agents as purpose-built for marketers, giving them advantages over more generic AI tools. The company embedded specific skills in its agents that marketers need and enabled them to collaborate.
For example, Manohar said that most marketers would be disappointed if they uploaded 10 creative images from a campaign, along with metrics, into ChatGPT and asked for analysis and recommendations.
“It has a hazy understanding from its pre-training of being trained on the internet of how to think about ads, but it’s very hazy,” he said. “It doesn’t have the hard skills on how to identify creative fatigue, what to look for week over week.”
Hightouch collaborated with customers and marketing teams to develop its agents, asking them to outline all the steps involved in launching a campaign. These steps include research, creating briefs, campaign planning, content creation, gathering feedback, obtaining approvals, planning distribution and analyzing results.
While many AI efforts in marketing focus on content creation, Manohar estimates content is only 10%–20% of campaign activity, and it’s often siloed by role. As AI models improve at content creation, that’s not an area where Hightouch — or other vendors beyond foundational models like OpenAI and Perplexity — are likely to win.
Instead, Manohar sees the value of AI in marketing coming from systems that augment and automate the other tasks.
“I think what’s really interesting about marketing is marketing covers such a breadth of dimensions of what AI makes possible,” Manohar said. That includes reasoning, reading unstructured content, multimodal creative generation, coding (for HTML emails and landing pages) and more.
Dig deeper: Oracle debuts AI agents for CX and revenue teams
The post Hightouch enters the fray for agentic AI for marketers appeared first on MarTech.
