One of the biggest promises of AI is: companies can do more with lower costs. For leaders facing pressure to improve margins, that’s an attractive proposition. And with AI becoming increasingly capable, reducing labor costs is no longer just a theoretical benefit.
In September 2025, Salesforce CEO Marc Benioff described reducing the company’s customer support workforce from 9,000 to 5,000 as AI agents started doing more customer service. In February 2026, Block CEO Jack Dorsey announced plans to reduce workforce from 10,000 people to fewer than 6,000, citing AI-enabled changes in how the business operates.
Those were huge reductions and I’m not suggesting they were necessarily wrong. Businesses have every reason to pursue efficiency and there are situations where reducing staffing makes sense.
But I worry that companies treating workforce reduction as the primary measure of AI’s success may be overlooking a much larger commercial opportunity. They may also be eliminating the people best positioned to help them capture it.
What happens when you eliminate the people who know how things work?
I’ve spent much of my career working with teams responsible for complicated projects, client relationships, and business operations. One thing I’ve learned is that an employee’s value rarely fits neatly into a list of assigned tasks.
Consider an experienced employee in a proposal department. AI may be able to review a solicitation, identify requirements, prepare a compliance matrix, and produce a respectable first draft. Work that once took days might take hours. On paper, that creates an obvious opportunity to reduce staffing. But the experienced proposal manager knows which requirements deserve extra attention because of a previous client experience. She knows when the technical team’s proposed approach is unrealistic, which commitments the company can safely make, and where a seemingly minor ambiguity could create trouble during contract performance. Those insights don’t necessarily appear in the documents an AI system processes. Much of that knowledge has accumulated through years of experience, relationships, and difficult decisions.
If the company eliminates that employee because AI can now perform much of her routine work, it may save a substantial salary. It may also lose someone who prevents expensive mistakes and helps the company win business. The financial consequences of losing that expertise can be difficult to quantify until something goes wrong.
This becomes especially important when AI systems encounter unusual circumstances. Someone still needs to recognize when an answer is plausible but incorrect, understand why a process has failed, and determine what to do next. A company can automate a task without fully replacing the business knowledge that made the task successful.
Before eliminating a position, leadership should understand exactly which parts of that employee’s contribution have become unnecessary and which remain valuable.
The problem with competing on lower costs
There’s another reason I question the wisdom of building an AI strategy primarily around workforce reduction: competitors have access to much of the same technology.
Suppose two consulting firms adopt similar AI capabilities.
The first uses AI to reduce its staffing requirements. It can now produce the same reports and recommendations with fewer employees. Its operating costs decline, and it gains flexibility to improve margins or offer more competitive pricing.
The second firm takes a different approach. It keeps its experienced consultants and uses AI to reduce the time they spend gathering information, preparing routine analysis, and producing first drafts. Those consultants now have more time to investigate clients’ problems, explore solutions they previously couldn’t justify researching, and develop additional services. The firm can pursue more engagements without expanding its workforce at the same rate.
Both companies benefit from AI. Both may become more profitable. The difference is what they’re building their competitive advantage around.
The first firm has improved its cost structure. That’s valuable, but other firms can adopt similar tools and pursue comparable savings. Over time, the initial advantage may narrow.
The second firm has an opportunity to offer something clients find more valuable. Its consultants can apply their experience to more substantial work, deepen client relationships, and potentially generate additional revenue.
There’s no guarantee that the second approach will be more profitable. The firm still has to turn its additional capacity into work customers will pay for. But its leadership is pursuing a broader set of opportunities than reducing the cost of delivering what it already sells. And that’s where I think many businesses should be concentrating their attention.
IKEA found another use for the time AI saved
IKEA offers an interesting example. In 2021, IKEA retailer introduced an AI customer service chatbot called Billie. By 2023, the chatbot had resolved approximately 47% of the inquiries it received, saving the company nearly €13 million.
That would have been a meaningful accomplishment on its own. But IKEA also retrained approximately 8,500 customer service employees to perform different work. Employees developed capabilities in remote interior design, digital sales, and handling complicated customer needs that required more personal attention. Instead of viewing every automated customer interaction as an opportunity to eliminate labor, IKEA invested in having its employees contribute elsewhere in the business.
The company’s remote sales channel generated approximately €1.3 billion in fiscal 2022. By July 2026, Fortune reported that IKEA’s remote sales centers had been its fastest-growing sales channel over the preceding three years, with annual growth of roughly 15% to 20%.
We shouldn’t attribute all that revenue or growth to AI. IKEA’s remote sales business involves other investments and commercial decisions, and the available reporting doesn’t establish how much additional revenue resulted directly from retraining. Still, the business decision is worth examining.
IKEA automated millions of routine interactions and preserved the opportunity to use experienced employees in work that could contribute to sales. It captured operating savings while developing another way for its people to create value. That’s a much more impactful management decision than simply calculating how many positions a chatbot could replace.
A particularly significant opportunity for smaller businesses
For small and midsize organizations, I think this issue deserves even more attention.
Historically, larger companies have enjoyed advantages that come with having more people and specialized expertise. They can devote entire teams to market research, pursue numerous business opportunities simultaneously, and invest in analysis that smaller competitors simply don’t have the capacity to perform. AI is beginning to change that equation.
A company with a modest business development team may be able to evaluate more opportunities and prepare stronger proposals. A small consulting practice may be able to conduct research that previously required additional staff or outside specialists. An operations team may finally have time to address recurring problems that have been ignored because everyone was occupied with immediate demands. These are possibilities worth investigating before assuming that the best use of AI is to reduce payroll.
Of course, additional capacity doesn’t automatically produce additional business. Someone has to decide where that capacity will go, whether customers will value the resulting improvements, and how employees should be prepared to do the work. That’s a leadership responsibility.
I believe businesses should begin by examining the work they want to improve and the opportunities they haven’t had the resources to pursue. Then they can decide where AI belongs. If an employee can recover ten hours a week through AI, it’s reasonable to calculate the labor savings. But I’d also want to know what those ten hours could produce if we kept them in the business. Could that employee help win another client? Develop an offering customers have been asking for? Resolve an expensive service problem that nobody has had time to address? Sometimes the right decision will still be to reduce costs. I don’t believe retaining every position indefinitely is sound management. But I would be reluctant to eliminate experienced employees until I understood what their newly available time could be worth.
AI can create capacity. Turning that capacity into business growth requires management to recognize the opportunity and act on it. Companies that focus exclusively on reducing labor costs may discover they’ve saved money while giving up something considerably more valuable: the experience and expertise that could have helped them build a stronger business.