Hello, {{hc_sub_firstname | friend}}! Param here.
88% percent of companies now report regular AI use somewhere in the business, up from 78% a year ago. That's not just a small business number – it's every business, from Fortune 500s down. And, even at that scale, most of them are stuck. Two-thirds haven't moved past pilots and experiments to anything resembling scale.
Now shrink the frame to the businesses Lynnfield actually buys. Self-reported AI adoption among small businesses sits closer to 57-58%. Lower than the corporate world, but headed the same direction, fast. And the same McKinsey survey that produced the 88% figure also found the gap by size: nearly half of $5-billion-plus companies have reached the scaling phase, versus 29% of companies under $100 million in revenue.
Read those three numbers together and a pattern emerges. AI adoption is now the default at every size of business. Scaling it into something that moves the numbers is still rare everywhere, and rarer the smaller you get.
For most people, that's a workplace-productivity curiosity. But, for someone about to acquire a small business, it's a diligence problem hiding in plain sight. You're no longer just inheriting a target's trucks, its customers, its receivables. You're also inheriting its AI stack – the half-finished voice agent, the automation nobody documented, the tool that breaks every time a vendor pushes an update.
Some of that stack is an asset. But a lot of it might just be a subscription with no owner.
The Adoption-Value Gap
Here's the number that should worry you more than any of the above: 80% of AI initiatives fail to scale past the pilot stage, according to Accenture's research. A separate 2025 MIT study found something sharper still – 95% of generative AI pilots never produce measurable financial return.
The reason is the foundation underneath it: no owner, no redesigned process, no integration into how work actually gets done. An owner who says "we're AI-forward, we use it for everything" hasn't told you whether that AI creates value or just adds a cost line.
That sentence should trigger diligence, not comfort.
Four Ways "They Use AI" Hides Risk Instead of Value
There are four patterns worth knowing before you sign an LOI.
Tools without process – a subscription layered on an unchanged workflow, cost on the P&L, benefit never materializing.
Half-implemented automations – the plumbing bot above, or a scheduling tool that double-books, quietly destroying value while looking like progress.
Vendor lock-in – a brittle integration bolted to one vendor's ecosystem, so the acquirer inherits switching costs and renewal cliffs the seller never tracked.
Key-person AI – the one employee who built clever workflows in a personal account; when they leave at close, the capability leaves with them.
AI adoption shows up on the cost side of a target's financials immediately, but it shows up on the benefit side rarely. Treating "they use AI" as a positive without verifying the value is a mispriced deal.
CARVE-OUTS
I’m going be at the American Tamil Entrepreneurs Association (ATEA) meetup in Milpitas, on Friday, talking about how to acquire SMBs, how to structure the deal, and how to use retirement accounts to participate.
If you're nearby and thinking about building permanent wealth through small business ownership, come say hi and we can talk about it.
Continue reading the main story below ⬇️
The Five-Step AI Line-Item
Add this to the existing due diligence checklist:
One: inventory every AI and automation tool the business pays for – contracts, monthly cost, renewal date, administrator. If the seller can't produce this list, that's itself a finding.
Two: trace each tool to a Key Performance Indicator (KPI) – booked calls, speed-to-lead, hours saved. No traceable metric, assume zero value and a removable expense.
Three: test the customer-facing automations directly – mystery-shop the voice agent. A bot that mishandles an after-hours call (like in ‘The SMB Campfire’ horror story above) is a leak you're buying.
Four: map the dependencies – what breaks if a vendor raises prices or the one employee who understands the setup walks out.
Five: score the net position as ‘Asset’, ‘Neutral’, or ‘Liability’, and let that number flow into the model as a premium, a savings line, or a remediation reserve.
The output is a line item, underwritten like any other.
Why This Is the Same Discipline as Everything Else We Do
Lynnfield's whole approach is underwriting what's real and refusing to pay for what's narrative. AI is just the newest place where the two diverge – "we're AI-forward" versus an unowned pile of subscriptions. Look at what almost no other buyer is checking yet, and the edge is available to anyone disciplined enough to add five questions to their process.
It also connects forward. Once you know precisely what's running, you know what to fix and what to deploy next – the highest-ROI moves become a sequenced plan instead of a guess, the same way we approached it a few editions back when we walked through our own AI value-creation framework.
CARVE-OUTS
The Lynnfield Investor Program
At Lynnfield, we acquire cash-flowing businesses in the $1M–$10M EBITDA range and offer co-investment opportunities to qualified investors. Join our investor list to receive deal flow as we evaluate new acquisitions.
Continue reading the main story below ⬇️
Thanks for reading!
In 2026, "Do they use AI?" is a settled question for almost every business, large or small. The question that separates a disciplined acquirer from an overpaying one is sharper: is the AI already in this target creating value, sitting idle, or quietly costing money?
Four times out of five, you'll find subscriptions to cut, automations to fix, or key-person risk to price. Occasionally, you'll find something worth paying for.
Either way, you won't know until you look.
Talk soon,
Param
P.S. In case you’re joining us late, check out our previous editions.