How Snowflake's Marketing AI Council Helped Turn a Global Org into an AI-Native Team

In a large meeting room, dozens of people watch a live demo of new AI features in Snowflake, their laptops open. Many more are doing the same over Zoom. The presenter walks them through a model context protocol (MCP) installation with Google, shows how the new tools can analyze marketing-qualified lead data and create formatted Google Sheets without moving between interfaces. 

These demos aren’t from a user conference or a customer-facing hands-on lab, and the people at the keyboards aren’t all devs or professional data scientists. They’re Snowflake marketers — ops managers, account-based specialists, brand wranglers, designers, PR professionals — almost all "nontechnical personas” more comfortable with a campaign slogan than a SQL statement. They’re learning about Snowflake CoCo at the company’s most recent Marketing AI Day, one of the quarterly internal enablement events run by the marketing org’s AI council. 

On this day, many from the audience confidently respond to questions in the chat with the AI use cases they’ve been spinning up. By the end of the two-hour workshop, more than 10 presenters have tackled amplification, self-service funnel intelligence, account-based marketing acquisition and more. Folks leave with new capabilities installed and a novel set of tips to deploy in their workflows. Marketing leadership knows they’ll use them, too — 93% of Snowflake’s marketers use AI tools daily in their work;¹ 70% of that weekly use across the nearly-600-person global marketing team is done with Snowflake’s own AI tooling.²

But a little more than a year before, that wasn’t the case. 

Before the spark

In early 2025, Snowflake’s marketing organization probably looked pretty similar to its peers in the tech space in terms of its AI usage. Some people on some teams were experimenting. Enterprisewide, tools like Google Gemini were available, as well as early versions of Snowflake Cortex AI that offered an LLM playground to less code-savvy employees. Like many companies trying to balance security, speed and innovation, there were a lot of questions as to how marketers could use the AI that was available. At the time, an internal survey that the council ran for North American marketers found that only 11.6% of respondents felt "very confident" using AI in their role.³

But the tools on their own weren’t the issue; those were becoming more available day by day. What was missing was something harder to roll out: trust, fluency and a shared framework for what responsible AI use actually looked like.

The Marketing AI Council formed in January 2025 — eight people spanning demand generation, content marketing, marketing analytics, web experience, marketing operations and account-based marketing. Its mission was to drive internal AI adoption for marketers through education, advocacy and inspiration. 

What they did first was not run a training session. They spent six weeks writing content guidelines, navigating review and compliance processes and answering the foundational questions that had been stopping marketers cold: which tools were approved, what data could go where, what good usage looked like in practice. The infrastructure — trust, rules, shared vocabulary — came before anything else.

When curiosity became adoption

In April 2025, Snowflake hosted its first Marketing AI Day in person in Menlo Park, California, and streamed globally. In structure, it looked a lot like the sessions that would follow, with live demos, AI workflows, maybe an external speaker. The difference was the depth; this was the start of AI experimentation for many people in the room. So council members walked through AI tips for improving prompt engineering and shared simple workflows for better content generation and more efficient data entry. 

Seeing AI in action on the work people were doing daily felt a little bit like magic. The energy in the room was palpable. “The first AI Day was the spark the marketing org needed. It was the moment people realized AI was much more than chatbots,” says Marc Nixon, a founding council member and senior manager for web engineering on the brand and creative team. “The real unlock came from bringing the right context into our workflows. From there, the buzz around what we could do with AI spread quickly across the organization.”

The event’s impact was immediately apparent. That single week, enterprise technology logged more than 1,200 Gemini interactions, six times the previous monthly peak. Over the next 90 days, usage climbed 418%.⁴ Not even a month later, Ryan Green, Senior Director of Program Content, built a Gemini Gem workflow directly from a live demo at the event that compressed weeks of interview preparation into days, producing AI-built outlines for more than 40 interviews.

By mid-summer 2025, 87% of the North America marketing org were active AI users.⁴ The shift was real. But it was still just the beginning.

The builders arrive

After the success of the first org-wide event, more enablement workshops hit Snowflake’s calendars. The council started a quarterly internal newsletter to keep the team abreast of AI updates, which were coming at breakneck pace. By August 2025, the question had changed from “do you use AI?” to “what are you building with AI?” Hands-on sessions for prompt templates, live demonstrations of Snowflake CoWork (at the time Snowflake Intelligence) for natural language campaign reporting and the first AI marketing days ensued across the regions. 

For example, at an event for marketers in the Europe, Middle East and Africa (EMEA) region that boasted 80% team participation, use cases included a workflow to automate personalization at scale for account-based marketing, an AI-powered dashboard for meeting note quality to help coach sales development representatives and a knowledge graph construction tool for SEO that maps topic structures across 7,100 concepts, surfacing 10.5 million monthly searches worth of content opportunity in the U.S. market alone. 

Even hackathons, building-focused deep dives typically reserved for developers or engineers, became part of marketers’ vernacular. The council led the first Marketing AI Hackathon, with events spread from winter to spring. Rather than those being organizationwide, the hackathons were smaller groups of specific teams so everyone could hone in on use cases that made most sense for their remit. 

And then the game changed.

Marketing meets Snowflake CoCo

In late 2025, an advancement to Cortex Code — Snowflake's AI-native coding agent, now called Snowflake CoCo — arrived on Snowflake marketers' desktops as part of a research preview. Only eight North America marketers started out using it weekly,² but the potential was enormous. With CoCo, marketers could connect Snowflake data to other tools through MCP without wasting time toggling between contexts. There was the potential to create Skills for repetitive work, much like a Gem or custom GPT, but with the flexibility to select frontier LLMs in the Snowflake interface. 

About two months after the “Loco for CoCo”-themed workshop in early April (where this story began), weekly usage for Snowflake CoCo by its own marketing team had climbed to 70% globally. 

Since the AI literacy foundation had been laid so well, marketers could take a powerful tool like CoCo and run with it — in every direction. Timmy Beckmann, a brand motion graphics designer, used CoCo to create a motion graphics export pipeline that turned a whole week of work for Summit, Snowflake’s biggest and most visible event of the year, into 20 minutes. “For the first time, I'm actually ahead of the content instead of scrambling behind it,” he says. 

Senior competitive intelligence manager Spencer Hong has developed a Streamlit app on CoCo that surfaces, prioritizes and schedules win-loss interviews, then generates a polished deck for quarterly reporting to leadership. What once took him three hours a month identifying candidates and a full week of deck work is now about 90% automated. 

Vinoti Desai, a senior specialist on the marketing operations team, partnered with the enterprise technology team to build an inbound automation engine that processes 2,000 to 3,000 daily email replies to global campaigns with zero human intervention, replacing a planned external software purchase. 

The same dynamic is playing out 12 time zones away. In the Asia-Pacific region, marketing and sales development representatives came together for a 150-person hackathon in May 2026 that produced multiple tools now running in production. 

Multiply those examples across every function and region, and the aggregate numbers tell the same story. Since CoCo's rollout, Snowflake marketing users have created more than 600 personal and project-level skills. Between 100 and 200 have gone further, building their own from scratch; the median user runs seven skills a week.⁵

"AI has closed the gap between thinking and doing," says Lindsey Sole, Senior ABM Manager. "Strategy and execution feel like one fluid motion."

Every team in every region was seeing the real impact of AI on their roles, and CoCo and CoWork were key to the transformation. As community marketing manager Amilee Alesna put it, “AI has fundamentally shifted the way I work so significantly that my professional life is now broken up into two distinct eras — BC (before CoCo) and AC (after CoCo).”

The AC era

Today, Snowflake’s marketing team is tracking reach to 100% AI usage by the end of the quarter. Implementation has matured so much that the original council, recently renamed the AI Peer Committee, is now the education wing of a multi-team AI Accelerator that inspires individual adoption, equips marketing employees with appropriate technical resources and holds the organization accountable for scaled AI impact. 

As it turns out, the approach the AI Peer Committee took — hosting events that provided low-pressure chances to brainstorm, create and share rather than just sit back and watch — correlated to more sustained tool usage globally. Having the space to ideate with adequate technical support also encouraged marketers to think beyond simply implementing tools already built. Marketers have become a part of the building process, and they have seen their ideas come to life as skills they can use in their day-to-day roles.  

Earlier this month, the council ran an internal survey to capture some of the shifts people were seeing now that they were more than a year into their education journey. The outcomes were stark: 92% reported that AI enabled work that would have been delayed or impossible, and 60% of respondents completed a specific task in half the time or faster with AI. Those are the efficiency numbers. But the qualitative shift may be the more important story: Respondents described moving up the value chain, from execution to strategy, as AI absorbed the mechanical layer of their work. Ninety-three percent rated AI tools as “essential” or “frequently used for key tasks.” People who may have been inexperienced at the start are now writing some of the most compelling use cases. That, more than any statistic, is the measure of what the council helped build at Snowflake. Not an AI program. An AI era. 

Learn more about setting up your own AI council, or explore how leaders can use Snowflake’s AI Data Cloud for marketing in this year’s Modern Marketing Data Stack.

¹ AI Peer Committee Internal AI Tools Impact Survey, June 2026.

² Snowflake internal AI tool usage data, as of May 26, 2026. 

³ Marketing AI Council internal survey, North American marketing organization, February 2025.

⁴ Internal Google Gemini usage data from North America marketing organization, March 7, 2025-June 5, 2025. 

⁵ Snowflake CoCo skills usage data, North America scope, January-May 2026.