Gartner predicts AI agents will outnumber human sellers 10 to 1 by 2028. Business growth strategist Kelly Roach says more bots do not mean more sales, and the companies that rushed in are learning that the hard way
Open your inbox on any Monday morning and you will probably find a few sales pitches waiting. One has sentences that break in the wrong places. Another is the same message you got last week from a different company, and a third reads as though a machine wrote it, which it did. Most people delete all three without a second thought, and that habit says a lot about where sales is headed.
Gartner, the research firm, predicts that by 2028 AI agents will outnumber human sellers by 10 times. In the same release, it says fewer than 40 percent of sellers will report that those agents made them more productive. So the number of agents is set to explode, while the payoff for the people using them stays small. Kelly Roach is not surprised by that gap.
Kelly is a former Fortune 500 senior vice president turned 8-figure entrepreneur. She hosts a top 0.5 percent globally ranked podcast with more than 1,100 episodes and has 121,000 Instagram followers. She has spent years teaching sales teams why relationship-driven selling beats automation alone, and she believes the distance between adopting AI and earning from AI is the biggest business story of the moment.
Companies Cheered the Layoffs, Then Hired Everyone Back
Last year, a number of companies announced large layoffs with pride, sure that AI was ready to take over the work of whole divisions. Kelly says many of them have since reversed course. “A lot of companies that were very proud to announce entire division layoffs and massive employee cuts last year have now retracted and are actually having to hire people back,” she says.
Her explanation is that AI is still wildly inconsistent. Anyone who uses it daily has watched it invent facts, defend the wrong answer, and then admit the mistake later when questioned again. That risk grows when a company hands its front line, the people who actually talk to customers, over to a tool that behaves this way. Kelly is clear that she is not against AI. She believes it is here to stay and will keep improving, but she does not think it can replace the front line of a company in 2026. The businesses that treated it that way, in her words, learned quickly that “AI was not quite ready for the promotion that they gave it.”
Trust Takes Years to Build and Seconds to Lose
Buyers today are tired and cautious, and Kelly describes the moment as a trust recession that has been building for years. In a market like that, one mistake stands out above the rest: sending a message that obviously came from AI while a person takes credit for writing it. “Nothing breaks trust faster than when you represent work as your own that someone can clearly see isn’t,” she says.
This is where her advice runs against the noise. Business owners keep hearing that they will be out of business overnight if they do not adopt AI fast enough. Kelly works with more than a thousand businesses a week as a growth strategist, and what she sees is the opposite. The owners who rush to let AI replace the human side of their work are often the ones struggling most, and she believes the risk of losing trust by leaning on AI is far greater than the risk of being slow to start.
Fix the System Before You Add the Software
When asked where she would first put AI to work inside a sales team, Kelly gave a surprising answer: she would not, at least not yet. She would begin by helping the team get clear on its numbers, its system, its offer, and the quality of its conversations, and only after those pieces were working would she look at what could be automated. “I actually wouldn’t bring AI into the mix until we dialed in the system, the process, and got things actually working,” she says.
Her reasoning comes from a pattern she sees again and again. A salesperson who is not getting results gives instructions to an AI agent, and now there is a human and a machine that are both falling short. Owners then wonder why the agent is not effective, but Kelly points out that someone had to train it. Gartner reaches a similar conclusion, warning that agents are only as effective as the data, workflows, and systems around them, and that adding more of them to a messy process creates more activity without more impact.
Busy Is Not the Same as Paid
Kelly also points to a study from MIT, The GenAI Divide: STATE OF AI IN BUSINESS 2025, that she says found 95 percent of company AI projects produced no return at all. She believes part of the reason is what people choose to automate. Saving time sounds like a win, but she says many owners use AI to speed up tasks that never needed to be done in the first place. “It’s saving time on things that probably didn’t need to get done in the beginning,” she says.
Her approach is simple. Start with something that already works, find the repeated tasks inside it that a person does not need to handle, and give those to AI so the working process can be copied and multiplied with fewer hours. Doing it the other way around produces a lot of busy work and very few new clients. From what she has seen personally, more business owners are wasting time with AI right now than making more money with it.
Where Human Sellers Will Still Win in 2028
Kelly does not doubt Gartner’s numbers. She notes that more than half of online activity is already bots, so she is certain agents will outnumber human sellers. What she questions is whether they will be more effective, and she draws a sharp line between the two ideas. “Volume does not necessitate result,” she says.
In her view, the human seller matters most at the moment a buyer decides, especially when the decision is large and the stakes are high. AI can organize conversations, suggest next steps, share helpful resources, and draft follow-up messages, but a person still needs to review and refine that work. The human heart, voice, and touch, she says, are still essential, and buyers can tell when they are missing.
Advice for AI Beginners
Plenty of people feel behind because everyone online seems to be saying they use AI for this or that, without ever explaining what they earned from it. Kelly tells those beginners to slow down and focus on the fundamentals first: winning customers, serving them well, and keeping the ones they already have. Once that foundation is tight, they can start layering AI onto repetitive tasks that do not need a human touch.
The companies that come out ahead in the AI age may not be the ones with the most agents. They may be the ones whose customers still trust them enough to pick up the phone.
About Kelly Roach: Kelly is a business growth strategist, two-time Inc. 5000 honoree and USA Today best-selling author. She helps business owners break founder dependence by building sales systems for scale. Visit KellyRoachInternational.com, follow her on Instagram at Kelly Roach Official, or read her Substack, The Sacred Art of Selling.
