If a rival can buy the exact same AI you just rolled out, what’s actually left to compete on?
Rachel Orston has a direct answer. “Anyone can buy the technology,” she says. “What they can’t buy is the human relationship, and that’s where trust, connection, and retention are won or lost.”
Orston is Chief Customer Officer at Instructure, the education technology company behind Canvas — the learning platform tens of millions of students and educators log into every day. She’s spent 25 years building customer relationships across startups and the Fortune 100, and she’s watched a lot of software promise to take over the messy human parts of a business. Her read on this moment is that the software should absorb the drudgery, and only the drudgery.
So where is it safe to let the machines run? Orston is specific. Administrative drag and busywork — “surfacing information, routing routine requests, and reducing repetitive tasks.” Hand those off. What you buy back is room for the things people are uniquely good at. She frames it as a north star: “we elevate human potential by designing systems that make room for the work only humans can do.” Free someone from data entry and ticket triage, and you get back their deepest human capacities, which she names as “creativity, empathy, discernment, and connection.” In education especially, that trade matters, because the whole enterprise still runs on one person handing knowledge to another.
There’s a catch she keeps returning to, though. None of this happens on its own. “We can’t talk about AI in the abstract without acknowledging that it will take humans to implement these changes,” she says. That gap is where a lot of the current hype falls apart. Talking about AI as a standalone fix is, in her view, exactly the thing to resist. What’s durable is far less glamorous. “This era requires massive change management and a level of ‘unlearning’ from all of us,” Orston says. “We are all starting with a beginner’s mindset right now.”
Customer expectations have moved fast — she points to a real shift in just the last six months — and she’s wary of teams that chase every one of those movements. Customers aren’t asking for trend-chasing. They’re asking for AI pointed at a real pain point, “whether that means saving time, reducing friction, or helping them make better decisions.” The rest is noise. And the teams that earn durable loyalty, she argues, are the ones customers trust to steer through the change with them.
“Customers expect us to navigate these shifts alongside them and on their behalf, proving that we understand how these big industry trends manifest in their day-to-day work.”
That trust is the real product. Which raises a harder question from the CCO seat: how do you prove any of it is working, when the payoff is retention and expansion rather than a single sale?
Orston watches the usual instruments — adoption, time to value, customer health, retention, expansion. But she’s clear that the numbers only mean something when they ladder up to what a customer actually set out to achieve.
“The strongest measure of customer experience is whether customers achieve meaningful outcomes and continue to see value over time.”
What grounds those metrics for her is oddly personal. Orston is the parent of two longtime Canvas students about to enter the workforce, and she says she thinks constantly about whether they’re leaving school with the judgment and adaptability an AI-shaped economy will demand. That’s the outcome she’s really measuring toward. “Our CX investments are paying off when we help institutions remove barriers, support educators as they rethink teaching and create better experiences for learners,” she says. Retention and expansion, in that telling, aren’t the goal at all. They’re what happens when the value is real and it keeps showing up, year after year.
For any team smaller than Instructure staring at the same shelf of AI tools, Orston’s takeaway is almost a relief. You don’t win by owning better software than the company down the street — you can’t, not for long. You win on the part of the work that never shows up on an invoice. Automate the busywork. Spend the time you get back on the customer. And treat this whole moment less as a race to deploy and more as a shared education, where nobody has it fully figured out yet, including you.
