About Greg
40 years putting technology to work. Never once selling it.
I look at AI the way an owner does: as a way to run the business better, not a product to sell you.
I started on a Commodore 64 in 1982. Since then I’ve put mainframes, PCs, the internet and the cloud to work inside running businesses. Never as a vendor. Always as the person who had to make it pay off.
For 25 years I’ve been an owner of Flexo Concepts, a manufacturer in Plymouth, Massachusetts. Sales, marketing, engineering, manufacturing, accounting, supply chain: every part of the business runs on technology I helped choose and roll out, including ten years on Salesforce. Today Flexo is fully paperless, and every workflow is digital.
Here’s what that taught me: software alone rarely fixes anything. The gains come from watching how the work really gets done, then redesigning it. AI changes who gets to do that redesign. The people closest to the work can now build the better version themselves. That’s what I teach owners.
“My goal is to help businesses take control of their tech and data. AI is the unlock we’ve been waiting for, giving small businesses the chance to do this on their own.”
Why I’m different
A Chief AI Officer has to work across the whole business.
Most tech consultants know the tools. Most business consultants know the strategy. AI needs someone who knows both, in every department: what each part of the business is trying to achieve, and how technology actually gets it there.
I’ve done the job in each of those areas, not just advised on it. And through all of it I’ve been a change manager, because improving a business is about the people as much as the systems.
Finance
P&L is my first language
My degree is in finance, and I’ve carried P&L responsibility for 25 years. It’s the part of the business I know best, so every AI idea gets the same test: does it pay?
Finance degree · 25 years of P&L ownership
Technology
Thirty years as the technology leader
Networks, infrastructure, hardware and vendors. I went from DOS to Windows, Windows to the cloud, and the cloud to AI, hands-on the whole way.
30 years leading IT · every platform shift since DOS
Marketing
I’ve built the marketing machine
Direct mail, complex digital ad campaigns, social media, a website and content management system I built, and marketing automation. I know what each one costs and what it returns.
Marketing tech, campaigns, web and automation
Sales & CRM
I’ve led sales teams and rolled out the CRM
I’ve implemented several CRMs, including Salesforce across an entire organization, and I’ve led the sales teams that had to live in them.
Multiple CRM rollouts · sales leadership
Operations
Lean is in my DNA
Manufacturing is my roots. I practiced lean when it was still called the Toyota Production System. It’s the best model I know for finding the work worth improving with technology, then managing the change.
Process flow · continuous improvement · manufacturing
People & culture
I’ve built the culture, not just the org chart
I’ve run HR, hired and ended hundreds of employment relationships, and handled the details of employee benefits. Ten years ago I led the work to create Flexo’s vision, mission and values, and I’ve led the company in living them since.
Six-time Great Place to Work · three-time best small manufacturer to work for
Your Chief AI Officer could also sit in the CMO, CFO, COO or head of people seat.
That’s the difference. I don’t just understand technology and business. I know how to apply one to the other, department by department.
How I think about it
What I’ve learned putting technology to work in a small business.
Featured article
My work is as much about culture as it is about technology.
Most owners start with the tools. I start with the business: how the work actually gets done, and the people doing it.
Ask ten people what “AI” means and you’ll get ten answers, and the definition keeps moving. So before anything else, we agree on what we’re talking about.
Read the article →Own your website
Most small business sites are rented, and every change is a ticket. Move to plain HTML on Cloudflare and make changes yourself with Claude, for close to nothing a month.
What doesn’t exist, you build
At Flexo Concepts we ran the whole business on one system we built in Salesforce: sales, ERP, accounting, shipping, manufacturing, quality and HR. Today you can do that with AI, faster and for far less.
Own it, even if someone else helps you maintain it
You used to pay someone to support your PCs and network. Next, you’ll pay someone to help support your code base and knowledge vault. Done right, you own it. They just help you maintain it.
AI delivers the digital transformation we were promised
We were told the systems would just integrate. They didn’t. The result is high software costs, disconnected systems and ROI that never showed up. That is what AI fixes.
The people doing the work know how to make it better
The people closest to the work know where it breaks. Fix their problems first, and they will start fixing yours. That is how a culture of continuous improvement begins.
You can’t implement AI by directive
AI changes how people work, so it behaves like culture. You can’t change culture with a memo, and you can’t roll out AI that way either. Fix real problems and habits follow.
Two meetings to start with: the Claude jam and the AI committee
Alternate two recurring meetings. The jam is everyone with a license, sharing work and solving problems. The committee is three to five people who own governance, architecture and the roadmap.
AI is a skill you practice, not a product you buy
AI is new to everyone and it can be learned by all. Like an instrument, people get good by practicing in small bits, often. Plan a trip, summarize a messy email thread, or ask the AI to write your prompt for you. Then do it again tomorrow.
Start with the problem, not the tool
Pick a real problem first, then work out how tech and AI can fix it. Start small. A web form that captures data, sends invites and books a calendar is a project for later, not a first one.
Context is everything
AI models remember nothing. They only know what you give them. The same model can return a generic answer, a specific one or a rich one, depending on the documents and connections you hand it.
MCP and the end of middleware
An MCP is an AI-enabled version of an API. It translates in natural language and adapts when data changes, so a small company can ask questions across its systems in plain English instead of paying for custom integrations.
Capture what your people know
Every lesson one person learns with AI should help the next. A company wiki or knowledge base of markdown files, searchable by meaning, turns tribal knowledge that lived in people’s heads into something the whole team can use.
Governance comes before scale
Adoption outruns the rules. Decide early who sees which data, what can be changed or deleted, how activity is logged and how people are trained. Moving fast is fine. Moving fast without guardrails is the risk.
Local models: cost will matter before security
Local models cost nothing per token, but you buy and run the hardware. Privacy gets the headlines, but I expect the cost of token usage to push more businesses toward them than data security will.
Free · 20-minute call with Greg
Want to talk about your business?
Twenty minutes to talk about your people, your processes and where AI fits. Operator to operator.
Questions people ask
Who is Greg Howell?
Greg Howell is the founder of SMB AI and co-founder of Flexo Concepts, a Plymouth, Massachusetts manufacturer. He has owned a business for 25 years and implemented technology inside running businesses for 40 years, from mainframes to a ten-year Salesforce rollout.
Why work with an operator instead of a technology vendor?
Greg has never sold technology. He has only implemented it, hands-on, as an owner leading sales, marketing, finance, engineering, HR, manufacturing and IT. His advice comes from running a business on AI every day, not from selling a product.
How is Greg different from other consultants or fractional executives?
A Chief AI Officer has to work across the whole business. Greg has hands-on experience in each key area: a finance degree and 25 years of P&L responsibility, 30 years as a technology leader, building marketing technology and campaigns, rolling out CRMs including Salesforce and leading sales teams, lean manufacturing and process improvement, and running HR and company culture. He could fill a fractional CMO, CFO, COO or people role as well as the CAIO role.
Is this about technology or about the work?
Both, but the work comes first. Greg starts by seeing how work actually gets done in the business, how much is automated or captured digitally, and then where AI and technology fit. With AI, the redesigned workflow can often be built by the people who do the work.
Does Greg use AI in his own business?
Yes. Flexo Concepts runs a paperless operation with all digital workflows, and AI is part of how that work gets done every day.