The work behind the numbers.
Two case studies. Both from environments where the work was directly led, not advised on. The first is about driving AI adoption. The second is about scaling operations 10x without losing quality.
Driving 86% AI adoption building-wide
15-month engagement. Role: building principal. Industry: K-12 education (the skills transfer; the context is incidental).
The starting point
A school building in the bottom 5% of state performance ranking. Staff with no exposure to AI tools beyond having heard of ChatGPT. Existing technology stack was outdated. Leadership turnover had been high. By every reasonable measure, this was the kind of operation where new initiatives go to die.
What we did
The turnaround had several pieces. AI was one of them, not the whole thing. The non-AI work mattered: rebuilt operating cadence, tighter intervention systems, a refreshed performance focus. The AI work was layered on top to give the staff capacity to actually do all of that without burning out.
Specifically, on the AI side:
- Structured training across the staff. Multi-session, role-specific. Not a keynote. The same model now offered as AI Training for Teams.
- Workflow library by role. Specific saved prompts and processes for each role. Shared across the organization. Not generic templates.
- Coaching the leads. Department leads got deeper training so they could carry the work after major training pushes.
- Tool selection discipline. Free tier where it worked, paid where the integration mattered. No vendor relationships clouded the recommendation.
- Outcome focus. Every training session ended with "what's the workflow you'll change this week." Adoption was tracked weekly, not surveyed quarterly.
What didn't work
A first attempt at staff-wide AI training was a single half-day workshop. Three months later, only about 20% of staff were using anything regularly. The takeaway: half-day workshops feel like value, but they don't change practice. The pivot to multi-session structured training with practice between sessions was where adoption actually moved. Same lesson now baked into every Palmetto Group AI training engagement.
A second early misstep was buying into a tool that promised "AI-powered" features and didn't survive contact with actual workflows. The lesson is part of why implementation work always starts with an audit and why the recommendation isn't tied to vendor relationships.
Why this matters for small businesses
The clearest takeaway: AI adoption is a change management problem, not a tools problem. Small businesses fail at AI adoption for the same reasons big organizations do: nobody defines success, the training is too generic to stick, and nobody follows up at the 90-day mark to make sure it's still happening.
A small business with 30 people facing the same dynamics will see results faster than a school building did, because the structure is simpler and the lines are shorter. But the work that produces results is the same.
Scaling an operation 10x without losing quality
6-year engagement, 2017-2023. Role: Virtual Learning Academy Coordinator. Operation grew from 100 users to 1,000.
Not an AI engagement, but the process behind Done-for-You AI Workflows and Implementation Consulting as services. Across six years, a small operation grew 10x in users while the underlying systems were redesigned end-to-end.
What it looked like
- Operating model redesigned three times across the growth curve, because what works at 100 users breaks at 400 and breaks again at 1,000
- Tool stack consolidated repeatedly, with quarterly review of whether, something was getting added or retired based on whether it earned its line item
- Onboarding processes built and rebuilt as the scale changed what was possible
- Data infrastructure shifted from "people manually track things" to "systems track things and people interpret"
- Quality metrics held throughout, even as user-to-staff ratio increased substantially
Why this matters for small businesses
Most growing small businesses face the same problem at smaller scale: the operation that worked at 5 people doesn't work at 15, doesn't work at 30. AI is one of the tools that can make those transitions less painful, but it has to be deployed inside a clear operational vision, not in spite of one.
The discipline of scaling operations, knowing what to automate, what to consolidate, what to keep human, applies directly to how AI should be introduced into a growing business. That's the operational frame we bring to every implementation engagement.
More coming.
Palmetto Group AI is new (2025) but the operator work behind it isn't. As current small business engagements wrap, written case studies will appear here. Active engagements are not published until they close out and clients approve.