AI Transformation Has to Serve the Customer

Internal AI programs can become strangely self-referential: more copilots, more dashboards, more “hours saved,” and somehow less clarity on whether customers feel any of it. As COO of Box, Olivia Nottebohm owns customer-facing teams—marketing, sales, success, support, partnerships—and her wrap-up advice lands on a first principle: AI also has to work for the customer.

Impact for impact’s sake is not the journey

When teams run “Box on Box,” they still have to remember they are a business. Customer stays at the center. AI impact pursued only to look modern—or only to pad an internal scorecard—is a weak destination. The better questions are ordinary and commercial: Are we serving customers better? Are we more responsive? Are we getting technology out to them faster?

B2B or B2C, the test is the same

Industry labels do not change the anchor. Whether you sell to companies or consumers, the customer should sit at the center of the redesign. Internal efficiency can fund better service—but efficiency that never reaches the outside world is a private hobby with a corporate budget.

• Ask of every AI project: who outside the company feels this?
• Prefer responsiveness and delivery speed over vanity automation.
• Keep cultural change tied to customer outcomes, not tool fashion.

Don’t get caught looking inward

Operators are especially prone to inward-facing programs: process maps, enablement tracks, model bake-offs. Those can be necessary scaffolding. They are not the point. Whatever process you build, whatever transformation you drive, whatever cultural change you try inside the organization, end the day with one check: is this better for the customer, or will the customer have a better experience? Returning to that business first principle keeps AI from becoming an expensive hall of mirrors.