AI isn’t coming to the workplace. AI is here, and we have questions.
Leaders want to know: how do I get employees to use AI tools safely? How do I price our AI features?
And how do other companies keep customer data secure?
Thankfully, there are companies—like Zoom, Asana, and Okta—leading the way to learn from. We gathered intel from leaders at these organizations to give you the knowledge you need to price accordingly, build responsibly, and engage effectively with AI tools.
To help you make the most of the insights, take a moment to see where you stack up among the stages of organizational AI maturity, developed by Anthropic and Asana.
Wherever you fall, the following insights will help you feel more confident about your position and maybe be just what you need to move to the next stage.
On engaging employees to use AI
Asana takes an education-first approach to engage employees by teaching everyone in the organization (including non-technical roles) how to best use AI tools and chatbots, including prompt engineering courses. They’ve also built an internal AI community with a mix of AI experts and enthusiasts. “Building communities and leaning on each other to ask questions has been incredible,” says Ethan DeWaal, AI GTM PM at Asana.
Meanwhile, Okta aims to enable employees to use AI safely and securely. “People will use it anyway, so we must help them use it responsibly,” Ron Piovesan, Sr Director, Technology Partnerships at Okta, explains.
They, too, have an internal community, an AI and machine learning guild, that proactively shares AI advances in a Slack group and bi-weekly Lunch and Learns.
“We try to balance being aware of all current developments and being mindful about selecting the right pilot programs to deploy rather than having people off in corners,” adds Piovesan, who also points to Okta’s use of AI in internal hackathons, encouraging employees to think creatively about opportunities to make people’s lives—customers and Okta employees—easier.
Zoom takes another approach to engaging employees: leveraging their insights as beta testers. “Any product Zoom launches to our customers first gets tested by Zoomies, aka those of us who work at Zoom,” says Robin Bunevich, lead AI product marketing manager at Zoom. “You quickly become an excellent beta tester as an employee at Zoom.”
On keeping customer data secure
Data security is a hot topic industry-wide and requires a new level of trust.
Bunevich’s first-hand experience shows that companies have many AI options and will not adopt tools if they don’t trust the company, how it’s built, or the people building the tools.
To build trust, Zoom doesn’t use any customer communications to train its AI models, says Bunevich. They also take an organizational approach because “if you aren’t adopting an organizational-wide AI tool, it can be harder to manage the security risk, but if you have a sanctioned tool everybody uses, and it’s integrated with an existing product you already have, it’s a lot more manageable,” Bunevich explains.
Similarly, Okta ensures that training data going into any AI model is relevant only to the customers and owned only by the customers who get the output from that model, explains Piovesan.
On pricing AI products and features
Many founders have experienced the 1,000-piece puzzle that is pricing, and while there’s no clear answer, these companies have seen success with their strategies.
DeWaal explains that Asana believes that AI shouldn’t be an add-on, but rather is something that will be baked into the core experience of technology. However, they only include some of the more complex/powerful AI features in higher tiers, incentivizing up-tiering within the seat-based model.
Meanwhile, Zoom chooses to offer its generative AI assistant, Zoom AI Companion, at no additional cost with eligible paid plans. “We did that because we wanted to make sure there weren’t employees at companies who got to have AI superpowers with gen AI and others didn’t just because of cost,” explains Bunevich. Zoom does, however, offer other monetizable AI products, particularly verticalized AI products.
For startups trying to determine their pricing, DeWaal advises keeping it simple. “Determine what value you add. Charge them 10% of that. Then, be really confident in your value proposition,” he says.
Pro tip: Avoid simply saying your product is AI-powered when expressing the value of features. “Using ‘AI-powered’ carries little weight in a world where so much is already powered by AI,” notes Bunevich. “Instead, be really clear about what problem your AI solution solves.”
On the future of AI
Where will AI go from here?
Bunevich foresees a world where AI works on your behalf, where you delegate tasks to it after it learns your skill set and fine-tunes how you work best.
Piovesan sees AI as a tool that’ll empower anybody to become an expert in any product or domain and vastly accelerate the time it takes to ramp up against any challenge—if you know how to use the tools available.
To DeWaal, the future of AI is all about more collaborative workflows, and he cautions those who bank their AI strategy entirely on chat. “Chat is great as a testing ground, but it’s hard to operationalize,” he adds.
As for how to get to that AI future yourself, DeWaal’s advice is this: “Try things, and don’t follow market leaders unwittingly because they, too, are still trying to figure it out.”
Figure out AI at WeWork
Right now, a vibrant, growing community of tech innovators and investors gather at WeWorks worldwide to figure out what’s next in their AI journey. And you can be one of them.
Discover how you can learn from and work alongside leaders in AI and other emerging industries to put AI into action at your organization.
