One Monday morning, a member of the marketing team pitched an idea: she wanted to build an online store to sell company swag. Under the traditional process—filing a request with the dev team, waiting for it to surface in the backlog, negotiating specs, running QA—it would have taken two to three weeks at best, and might easily have slipped into the next quarter. Instead, she had a live swag store up and running by Tuesday morning. No engineering resources, inside 24 hours. It's one of the AI-adoption stories Sendbird CEO John Kim has shared publicly.
What makes this scene remarkable isn't just the speed. The fact that a marketer shipped to production without a developer is itself evidence of how AI actually operates inside that organization.
Culture campaigns change nothing
Many organizations start their AI journey the same way: a company-wide training session, a new Slack channel, a handful of designated "AI champions," a monthly newsletter. Ask people six months later what's changed, and the answer comes back: "I use ChatGPT sometimes."
John Kim took a different approach. He designed AI adoption itself as an internal product. The company introduced a mission system called quests, awards tokens for completed missions, and posts cumulative token counts on a company-wide leaderboard. Employees who reach a certain tier earn the right to register as in-house instructors in a "skills marketplace" or run sessions for colleagues on other teams. The structure borrows directly from game mechanics.
What separates this from a simple incentive scheme is that the learning cycles back into the organization. Power users don't merely get recognition—they become teachers. Learning doesn't stay locked up as individual skill; it compounds into team capability.
Sendbird is a B2B SaaS company that provides chat, voice, and video APIs, with a headcount in the hundreds. Getting an organization of that size to a state where marketers deploy their own stores, engineering teams build their own automation tools, and every team runs custom AI workflows—the methodology John Kim has laid out holds lessons that apply just as well to solo entrepreneurs and small teams.
The gap between power users and everyone else isn't about tool access
Here's the interesting part. Every Sendbird employee has access to the same AI tools. The company pays for the subscriptions. Yet the gap in how people use them is extreme. One employee launches a store in a day; another, six months in, is still at the level of "I use it for drafts sometimes."
The common explanation for this gap is "tech fluency" or "digital literacy." But that explanation is only half right. The other half is attitude.
Sort out what AI can and cannot take over, and you'll find that much of skill and knowledge is already handled. How to write code, how to write copy, how to analyze data—you no longer need to master these to produce results, as long as AI is working alongside you. But the attitude of "I'll own this outcome," the drive to find a better way than the current one, the willingness to fail and try again—AI can't supply any of that for you.
That's precisely what Sendbird's quest system is actually designed to build. Completing missions nudges employees into uncomfortable experiments: using tools they normally wouldn't touch, solving problems in unfamiliar ways, and putting the results in front of other people. A transparent, public leaderboard means the attempts themselves become visible.
A solo entrepreneur can't copy this structure wholesale—there's no team to fill a leaderboard. But the core of the mechanism John Kim designed—building a structure that forces learning—works even for one person.
It helps to recall the three components of competence: skill, knowledge, and attitude. In the AI era, the barriers to entry for skill and knowledge have collapsed. Anyone can generate code, run market research, or summarize documents with a single prompt. Under those conditions, attitude is what makes the difference: the will to use the tools better, the openness to try new approaches, the habit of continually redesigning how you work. No org chart, no training session, and no AI tool can instill those things. That's why Sendbird gamified them—to turn them into repeated experience.
What this structure means for a business of one
For a solo entrepreneur with no quests and no leaderboard, the most practical takeaway from this case is this: stop treating AI adoption as a nice-to-have, and force it into your actual workflow.
A few ways to do that come to mind.
First, start logging how long tasks take without AI. Client proposals, content drafts, post-meeting summaries, contract reviews. Track the time for a month, and you'll see exactly where AI makes a real difference. Habits form not from a vague "I should use this," but when the time savings become concrete and visible.
Next, hand over what you're bad at first. Using AI in your areas of strength produces little felt impact. Instead, deploy it first on the work that takes you forever, that you're unsure about, that you keep putting off. If financial reports have always felt impenetrable, have AI structure them. If a proposal needs cold-eyed feedback, cast AI as a critical reviewer. Through this process, you'll develop your own criteria for what to delegate and what to keep doing yourself.
Third, once a month—even just once—write a single line recording a new way you used AI that month. If Sendbird's quests are a structure for making attempts visible, the solo-founder version is a routine of recording your own attempts for yourself. It doesn't need to be elaborate; one Notion page or a single tag in your notes app is enough.
As for tooling, going deep on one tool beats assembling a complex stack. Rather than rotating among ChatGPT, Claude, and Gemini, pick one and build your own repeatable way of handling a specific task with it—that path leads more directly to productivity gains. The marketer who launched a store in a day didn't do it by knowing the entire AI tool landscape; she did it because the flow of handling a specific task with a specific tool had become second nature.
One final question worth checking yourself against: among the tasks you do regularly, is there anything you're still doing without AI for no reason other than "this is how it's done"? Habitual inefficiency is hard to see. The team at Sendbird that designed the first quests almost certainly started from that same question.
The heart of AI adoption isn't handing people tools—it's designing a structure that makes people want to use them. Sendbird solved it with a game, and a marketer opened a store in a day. If you're a business of one, you have to be both the architect of the structure and the player inside it. That sounds hard, but in practice it starts with one simple routine. The attitude you bring when you sit down in front of the tool—that comes first.
