Partway through Good Karma Brands' company-wide AI rollout, employees started asking Ty Bauschek a question he didn't expect:
Had their names reached the alphabetical cutoff yet?
Bauschek, the company’s Senior Director of Innovation and Sales Development, was helping lead Good Karma’s AI transformation...
And what employees meant was:
They were eager to know if this Tuesday was their Tuesday to get a ChatGPT Enterprise account. (The company was rolling out licenses to 50 people a week at a time.)
Needless to say…
Boundless enthusiasm is not your typical AI adoption problem.
But boy, is it a good “problem” to have. And Good Karma has it in spades.
In less than a year, Good Karma Brands has 65% or more of their entire company using AI daily, with usage sometimes reaching 75%. The sports media and marketing company has more than 550 employees working across audio, video, digital, and live events. It did not reach that level of adoption by handing everyone a tool and hoping curiosity would take care of the rest.
Good Karma made AI a visible company priority. It rolled access out in cohorts. It gave employees repeated chances to learn from one another. And, once usage grew, it changed the question from Are people using AI? to Is AI helping them do better work?
Good Karma's transformation started with an unusually strong signal from the top.
Founder and CEO Craig Karmazin told employees that becoming an AI-forward company would be one of the three biggest initiatives in Good Karma's history. He placed it alongside the company's decision to become a sports company in 2002 and its move into ESPN Digital work in 2015.
That framing did two things immediately. It created urgency, and it made clear that AI was not a side experiment owned by one technical team.
Karmazin personally tested tools including Microsoft Copilot, ChatGPT, and Claude. The company hired two senior software engineers. Around Labor Day 2025, Good Karma also began giving every full-time teammate access to ChatGPT Enterprise.
The message did not disappear after the announcement. Leaders repeated it on company-wide calls, connected it to the future of the business, and showed employees that they were learning too.
"You often need to have executive buy-in in order for change to really happen," Bauschek says. "It's almost the opposite here, where the executive was telling us we need to have change."
Leadership set the direction. But direction alone would not get hundreds of people from first login to daily use.
Good Karma resisted the temptation to release hundreds of licenses to hundreds of employees at once.
It started with a pilot group of about 50 teammates on the Tuesday after Labor Day. The group had weekly check-ins and a Microsoft Teams chat where employees could share what worked, surface problems, and compare notes.
After a month, the company decided to expand access across the organization. For seven consecutive Tuesdays, another 50 employees received accounts in alphabetical order.
Each group joined a kickoff call. Karmazin reinforced why AI mattered. Teammates demonstrated early successes. Leaders reviewed company policies, practical guardrails, and setup steps. Then each cohort entered its own group chat for support.
The staged rollout created a useful compounding effect. Early cohorts found the rough edges and lingering issues. Later cohorts received answers to common problems before they encountered them. By the fifth or sixth session, Good Karma could teach teammates about recurring roadblocks (like setting up connectors between ChatGPT and their company systems) using lessons gathered from the groups that came before.
The initial rollout ended. But the learning cadence just kept growing.
AI now appears in company-wide, sales, marketing, and content meetings. Teams highlight how an employee used AI that week, demonstrate useful workflows, and make successful approaches more visible across the business.
Good Karma also uses an alternating Tuesday rhythm. One Tuesday, employees receive a newsletter with new features, practical tips, and examples from their peers. The next Tuesday, they can join a live training session built around real tools and real work.
This frequency solves a basic adoption problem: Most employees do not need one heroic training session. They need regular reminders, relevant examples, and help at the moment a tool becomes useful to them.
It also makes adoption feel less like a mandate from a distant transformation team. Employees see colleagues doing recognizable work, then learn how to apply the same ideas in their own roles.
Good Karma's broad access model let useful ideas emerge from anywhere in the company.
A sales support teammate automated a manual reporting workflow that had required employees to pull files, update spreadsheets, and send recurring emails. A marketing teammate built a campaign assistant that reduced a process from roughly three hours to 10 minutes. Content teams began using daily show transcripts to create AI projects that could help with topic selection and show preparation.
The point was not that every employee had to build an advanced system. The point was that every employee could see AI solving problems that looked like their own.
Bauschek often acted as a connector. When someone asked a question, he could point that person to a colleague in another market who had already solved a similar problem. A local discovery could then become a company-wide capability.
"That culture of sharing and collaboration is what made it successful so far," says Bauschek.
Good Karma reinforced that culture by giving employees permission to experiment and fail. It then amplified the solutions that proved useful until they became part of the process.
Early in the rollout, Good Karma tracked signs of activity: messages, custom GPTs, connectors, tokens, and power-user scores.
Those numbers helped the company see whether employees were engaging with the tools. But they could not show whether the work itself was improving.
As Bauschek puts it, "Maybe we just have some chatty teammates, right?"
So Good Karma shifted its attention to impact. Are employees spending more time with partners? Are they removing repetitive steps? Are they automating workflows that keep people behind a computer? Are they able to do more of the work they do best?
"We really shifted that mindset to what impact is AI having with us, and then how can we accelerate that impact rather than just usage," Bauschek said.
This is also why he does not treat 100% daily usage as the obvious goal. If AI helps a marketing consultant automate routine work and spend the day meeting with partners, that employee may not log in. The lower activity number could reflect a better business outcome.
Good Karma's results came from a sequence of choices that reinforced one another:
Good Karma is already moving into that next phase by creating dedicated innovation roles to find, improve, and scale high-value workflows.
But the foundation came first: clear leadership, broad access, structured support, constant sharing, and enough repetition for AI to become part of how the company works.
The 65% figure is impressive, no doubt. But the operational system that produced it is the more useful result.
Did you enjoy this transformation story? Go deeper on how AI is reshaping work and business with The Artificial Intelligence Show. Each week, we break down what matters in AI and what leaders should do about it.