The only way to get good at AI is to fail at it, says AWS’s chief marketing officer Julia White

When Amazon Web Services’ marketing chief first started experimenting with artificial intelligence agents, she was doubtful about her future.

"I myself had a moment of like, oh my gosh, am I useful anymore?" Julia White told Euronews Next at the VivaTech conference in Paris. After testing it out, she later felt relieved that the technology might strip away the tedious parts of her job.

This realisation is exactly why leaders need to get their hands dirty with AI, she said.

Full workflow reinvention

Experimentation is key, White said, adding that early AI adoption delivered modest but real productivity gains of 10 to 30% as AI was layered onto existing workflows. But the bigger breakthroughs only came after throwing out the old playbook.

"To get really big effects — like 5x effectiveness — we actually had to step back and rewrite how our processes work," she said.

The results were striking: creating a new webpage, a task that once required around three hours and multiple team members, now takes 30 minutes, with AI agents handling much of the heavy lifting. AWS publishes more than 5,000 new webpages a year.

What AI can’t do

Despite the efficiency gains, AI cannot replace human creativity.

"AI isn't very good at beautiful storytelling that really connects with the human experience," she said. "It's a wonderful thought partner — but it's not a tastemaker," White said.

That distinction has shaped how the team uses Amazon's own internal assistant, Amazon Q.

Rather than outsourcing creative judgment to the technology, marketers use it as a sounding board. The approach recently produced what she described as a breakthrough brand narrative for AWS, which she said even moved colleagues to tears.

"That core storytelling, that unique human insight, is a person," she said. "Knowing that and working with it in that same way is what has got us the best outcomes."

Celebrating failure

But to truly get good at AI, failure is the only clear option, White said. She has launched a “Be Brave” award within her team, which honours the efforts that did not work out — and makes a point of sharing her own missteps with her team.

"Failure is necessary on the path to mastery," she said. "We're never going to get great at AI if we don't try and fail."

She also said that finding time to experiment at all is a challenge for many companies. White has gotten around that by introducing dedicated training days with no meetings, reserved purely for learning new tools.

The personalised AI experience

The most exciting aspect of the AI age for her is the resurrection of ideas she had once written off as impractical. Chief among them: truly personalised marketing at scale, tailored to each individual customer.

"I've always dreamed of how I can have a truly personalised experience for every single customer," she said. "This was not practical before, but now it's suddenly practical."

Her advice to other leaders is simple: start using it.

"If you just read about it or hear about it but you don't actually use it, you're going to miss it," she said. "And if you don't, you're going to not be leading your teams well."

Video editor • Roselyne Min