New York, United States – 21 August, 2026 – A 2025 report from MIT’s NANDA initiative, The GenAI Divide, analyzed 300 corporate AI deployments and found that 95 percent produced little to no measurable effect on profit and loss. Companies spent heavily, ran pilots, and had almost nothing to show on the bottom line.
The 5 percent that worked looked different in two ways. They put AI inside the part of the business that produces revenue rather than running it as a side experiment, and they hired specialized builders instead of building in-house. Tools purchased from outside vendors succeeded about 67 percent of the time in the study, roughly three times the rate of internal builds.
The researchers traced the failures to what they called a learning gap. Generic tools like ChatGPT work well for individuals, but they stall inside a business because they never learn the company’s specific knowledge or adapt to how it actually operates. The winners used systems built around one business, trained on its material, and placed where customers could touch them.
A test from the expert services world shows what sitting in the revenue path can mean. Kodara, which builds AI products for coaches, advisors, and consultants, ran one ad to one audience through three different funnels. A free guide roughly broke even. A direct offer returned about three times the ad spend. A low-priced AI product placed between the ad and the sales call returned about nine times. Nothing changed except the middle step.
“Most companies bolt AI onto their business like a gadget,” said Lucas Tyson, founder of Kodara. “The returns show up when the AI is the thing the customer buys and talks to. It has to sit where money changes hands.”
For an independent expert, that means the AI itself becomes the product. Kodara trains it on the expert’s own material and sells it at a low price, so a stranger can spend twenty minutes with the expert’s thinking before any human conversation happens. By the time that person books a call, they already know how the expert works and what working together would look like. The sales conversation starts warm instead of cold.
The survey data suggests why most owners stay stuck on the wrong side of the divide. In a Goldman Sachs survey of small business owners published in March, 73 percent said they would benefit from more training and support to put AI to work. The successful minority in MIT’s data mostly skipped that problem by paying someone else to build. That matches how Kodara’s clients approach it: they treat AI the way they treat accounting or law, as a job for a specialist rather than a skill to learn.
Kodara delivers its builds in six weeks and guarantees each client 1,000 warmed-up leads in the first year or a full refund.
Tyson shares the company’s funnel tests and results on YouTube at youtube.com/@lucaslee-tyson and on Instagram at instagram.com/lucasleetyson.
About Kodara
Kodara is a done-for-you AI building service that turns expert knowledge into custom AI products. The company builds AI systems trained on an expert’s proprietary knowledge, enabling coaches, consultants, advisors, and other professionals to create scalable AI products that generate qualified leads and revenue. Learn more at kodara.com.
Media Contact
Company Name: kodara
Contact Person: Media Relations
Email: Send Email
Country: United States
Website: https://kodara.com