Skip to content

Does AI training pay off? What the latest research says

Illustration of a rising chart icon for AI training results

Short answer: Training reliably gets people using AI. On its own, it does not guarantee they save more time or produce more value. Research published in October 2026 points to three things that make the difference: regular use on real work, changing how the work is done, and deciding what people do with the time they save.

Training drives adoption

A California Management Review article from October 2026 reports on a survey of more than 3,000 employees and executives. 93% of employees who had received AI training were using AI tools, compared with 57% of those without training. Among people using AI for their job, those trained in the past year reported saving around 10 hours a week, roughly double the five hours reported by those without training (California Management Review, October 2, 2026).

The authors also found that company-provided training moved about half of employees to regular use, roughly four times the rate among people with no interest in training. Their explanation is useful for any leader: company training works because it comes with approved tools, colleagues learning together and a clear signal that this is how work gets done.

But training alone does not predict time saved

The same researchers ran a follow-up survey of 859 workers. Most reported saving 0 to 3 hours a week, and that barely changed with more training or more frequent use. In their words, firm training "is a powerful predictor of whether someone uses AI. It is a weak predictor of how much time they save by doing so" (California Management Review).

Workday's monthly tracking points the same way. Employees report saving about 3 hours a week on average, and 33% say they lose at least half of that to fixing or reworking AI output (Workday, October 22, 2026).

Skills fade without continued practice

A randomized trial published in Nature Human Behaviour on October 26, 2026 gave 487 management consultants access to ChatGPT and training to tackle data science tasks outside their skill set. They scored higher than the control group on three tasks. After the experiment, though, the researchers found no evidence they were better at answering technical questions without ChatGPT. The authors describe AI as an "exoskeleton" whose gains depend on continued use (Nature Human Behaviour, October 26, 2026).

For a small business, the practical lesson is simple: a one-off session is a start, not an end. People need to keep using AI on their real tasks, with feedback, for the habits to stick.

Returns come from changing the work, not just the tools

Wharton's 2026 AI Adoption Report, published October 13, studies larger enterprises, but its pattern is relevant to smaller teams. 67% of leaders whose departments fully embedded AI in workflows, alignment and measurement reported strongly positive ROI, compared with 19% of those that had not. Organizations standardized on a single approved AI platform reported strongly positive ROI at 50%, against 9% of those still experimenting. And 35% of leaders tracking at least one growth measure reported strongly positive ROI, against 13% of those tracking efficiency alone (Knowledge at Wharton, October 13, 2026).

Small businesses face the same gap at a different scale. In Goldman Sachs 10,000 Small Businesses' October 2026 survey, 86% of alumni use AI but only 15% have fully integrated it into core operations. 56% are working on AI upskilling for staff, most often by giving people tools to learn on their own (35%), on-the-job training (30%) or encouraging outside training (20%). 77% say they would benefit from more AI training and resources (Goldman Sachs 10,000 Small Businesses, October 2026).

What this means for a small business

Put together, the research suggests a few practical rules:

  1. Train on real work, not demos. Training should use the tasks your people do every week, so the habit forms where the value is.
  2. Make use regular. The CMR authors found that regular users were more likely to reinvest saved time in higher-value work, even when the time saved was small.
  3. Pick your tools and say so. Standardizing on approved tools gives people a clear path and makes training easier.
  4. Decide what saved time is for. The CMR article argues the real question is what happens to the time AI frees. Agree with each team what higher-value work comes first.
  5. Measure more than hours saved. Track one outcome that matters to growth, such as response time to customers or proposals sent, alongside time saved.
  6. Plan for review. Rework eats into time saved. Teach people how to check AI output quickly, and where not to use AI at all.

Workshops, coaching or both?

A workshop is a good way to start a whole team at once and agree on tools and rules. The research above suggests continued practice is what keeps the skills from fading, so plan for both.

At LetsGo, in-person AI workshops run as a half day or a full day around the team's own work. LetsGo for Teams adds one-on-one sessions, where 1 credit equals 1 session, so people can keep practicing on their own tasks after the workshop.

FAQ

Does AI training increase productivity? Recent research shows training strongly increases AI use. Whether it raises productivity depends on regular use, changes to how work is done and what people do with the time they save.

How much time does AI save employees? It varies. In the CMR follow-up survey, most workers reported saving 0 to 3 hours a week. Workday's tracking shows about 3 hours a week on average, with a third of users losing half of that to rework.

Is a one-time AI workshop enough? It is a strong start for alignment and basic skills. The Nature Human Behaviour trial suggests gains depend on continued use, so plan follow-up practice.

How should a small business measure AI training results? Track adoption, time saved and at least one growth outcome. Wharton's report found leaders who measured growth alongside efficiency were more likely to report strong returns.

Next step

If you want your team's AI training to stick, start with real work and plan the practice that follows. See how LetsGo workshops work or learn about LetsGo for Teams.

Free 30-min call

See what your team can achieve

A quick call with one of our founders. Tell us about your team, and we’ll show you what’s possible.

Talk to us