Mariya Lytvynyuk
The First Question Is Always the Same
The first question we get about AI is always the same. How many people can we replace?
It doesn't matter the industry, the company size, or the role of the person asking. The assumption is baked in: AI means fewer people.
We understand why people think that. It's the story everyone's been told: in headlines, in boardrooms, in the pitch decks of every tool promising to do the work of a department. But it's not the story we believe in, and more importantly, it's not the one that matches what we actually see working.
What AI That Actually Works Looks Like
The AI that actually works in ecommerce looks different from the one in the headlines. It's quieter, and it sits next to people rather than in place of them.
It looks like a support team that finally has time to handle complex, emotional, high-stakes customer issues, because the repetitive "where's my order" and "how do I return this" questions are handled automatically, the moment they're asked.
It looks like a merchandising team that spends its day on strategy and storytelling instead of manually tagging thousands of products, writing the same descriptions over and over, and reconciling spreadsheets that no customer will ever see.
It looks like a small brand suddenly able to greet every visitor, answer every product question, and follow up on every abandoned cart: things that used to require a headcount they could never justify.
In each case the team stays the same size. What it can take on grows.
That pattern is measured, not just asserted. When Anthropic asked tens of thousands of people how AI had changed their work, 48% of those reporting a productivity gain described an expansion in what they could take on, against 40% who pointed to speed. The average self-reported productivity effect landed at 5.1 on a 7-point scale, which the researchers read as substantially more productive.
The Team Stays the Same, the Scope Grows
That last line is the whole point, so it's worth slowing down on it. The promise of AI in ecommerce isn't subtraction. It's leverage.
When you take the predictable, repetitive, time-consuming work off a person's plate, you don't end up with a person who has nothing to do. You end up with a person who finally has time to do the work that was always too important to rush: the difficult conversation, the creative campaign, the strategic bet that needed a human who wasn't buried in busywork.
Human and Machine: Who Does What
The brands getting this right draw a clear line between the two kinds of work, and they let each side do what it's actually good at.
The human handles what requires judgment, empathy, and creativity. The upset customer who needs to feel heard. The brand decision with no obvious right answer. The campaign idea that's never been tried before.
The AI handles what's repetitive, time-consuming, and predictable. The thousandth identical question. The product tag. The status update. The follow-up that should have happened but nobody had time for.
Drawn this way, AI stops being a threat to the team and becomes the most useful member of it: the one that never sleeps, never gets bored of the routine questions, and never resents the grunt work.
It is worth being honest that this does not happen automatically. In the same research, roughly one in ten people said the time they saved was simply absorbed by more work being asked of them, and early-career workers were markedly less likely than senior ones to feel they had personally benefited. Whether AI creates leverage or just raises the quota is a management decision, not a property of the software.
The Brands Getting Real Results
Look closely at the ecommerce brands pulling ahead with AI right now and a pattern shows up. They aren't the ones who treated it as a layoff plan. They're the ones who treated it as an upgrade for the people they already had.
They gave their support team an AI sales agent that handles the routine questions so the humans can handle the hard ones. They gave their merchandisers tools that automate the tedious catalog work so they can focus on what sells. They gave their marketers systems that handle the repetitive sends so they can focus on the story.
The result isn't a smaller company. It's a company that punches well above its headcount, one that feels, to the customer, like a much bigger and more attentive business than it actually is.
The Real Question
The question isn't whether AI will change ecommerce. It already has. Stanford's AI Index puts adoption at 88% of surveyed organisations, while finding agent deployment still in single digits across nearly every business function. Almost everyone has started; very few have finished.
The real question is what you'll do with it. Whether you'll use it to shrink your team, or to make the team you have better at what they already do.
One of those paths ends with a leaner version of the same business. The other ends with a fundamentally more capable one. The technology is the same either way. The difference is entirely in the question you decide to ask of it.
Frequently Asked Questions
Not in the brands seeing real results. The most successful adopters aren't cutting staff, they're giving their people better tools. AI takes over repetitive, predictable work (routine questions, product tagging, status updates) so the team can focus on what requires judgment, empathy, and creativity. The team stays the same size; the output goes up.
Draw the line by the nature of the work, not the job title. AI handles what's repetitive, time-consuming, and predictable: the thousandth identical question, the product tag, the routine follow-up. People handle what needs judgment, empathy, and creativity: the upset customer, the brand decision with no obvious answer, the campaign that's never been tried.
Use a simple test: does it free your people to do more of what only people can do? If a tool genuinely creates capacity for higher-value work, it's leverage. If it just does the same work with fewer people, you've bought a cost cut, not a new capability, and you'll likely feel the limits of that quickly.
Because the repetitive work was never the valuable part. When you remove it, you don't get a person with nothing to do. You get a person with time for the difficult conversation, the creative campaign, and the strategic bet that busywork used to crowd out. That's where growth actually comes from, and it's why augmented teams tend to outperform leaner, automated-away ones.