Ask a mid-market company about AI right now and you’ll hear about pilots. A tool in finance, a proof of concept in operations. Nobody has figured it out, and every conversation is about the technology: the platform, the use case, the ROI. The people who currently do the work show up as one line at the bottom of the slide that says “workforce reduction”
Here’s why that catches up with you. Pilots don’t stay pilots. One of them works, the business case gets real, and leadership is suddenly looking at a function that needs seven people instead of twelve. Add cost pressure, a restructuring, or a merger, and the people decisions arrive all at once, made in a hurry by whoever holds the spreadsheet.
The Playbook Pays Twice
Here’s how workforce reduction normally works. Leadership counts roles. A function that needed twelve people needs five once the automation lands. Seven people get severance, and the plan moves on.
Meanwhile, down the hall, HR has requisitions that have been open for months. The roles every company is hunting right now, compliance and risk specialists, and the rare analysts who can sit between the business and the technology, are scarce, expensive, and slow to land. Mid-market companies feel this worse than anyone, because they’re competing for the same talent as enterprises with deeper pockets and bigger brands, and they usually lose the bidding war.
Nobody connects the two spreadsheets. The company pays severance to walk experienced people out the front door, then pays recruiters to search for similar people to bring in the back door. Twice the cost. And the person leaving knew your systems, your customers, and your quirks. The person arriving, if you ever find them, knows none of it.
Crack Open the Role
A job title is a container. Inside it is a bundle of skills, and automation doesn’t take the whole bundle. It takes some skills and leaves others, which means the useful question is what’s inside the role and where each piece goes.
The analyst’s job breaks down into five or six distinct skills: document verification, data processing, exception investigation, working regulatory knowledge, and pattern recognition built from years of messy files.
AI takes the first two. Document verification and data processing were most of the hours, so on a role-based spreadsheet, the job looks 60 percent gone and like an obvious cut.
But look at what AI didn’t take. Exception investigation gets more important, not less, because someone has to handle everything the machine flags and can’t resolve. And that working regulatory knowledge is the exact foundation of the compliance role this same bank has failed to fill for five months. Set the two jobs side by side, skill by skill, and several of the analysts headed for the reduction list already hold most of what the unfilled seat requires.
The part they’re missing has a name: the skills delta. It’s simply the gap between what a person already has and what the destination role needs. For the analyst moving into compliance, the delta might be a certification and case management training. Call it twelve weeks.
That delta is a useful little number because it’s three things at once. It’s the training plan, since closing the gap is literally the curriculum. It’s the price tag, since a small delta means a fast, affordable transition and a big delta means you’re forcing the wrong move.
The Person You’re Cutting is the Person You’re Recruiting
That’s the sentence to bring to the CFO, because this argument wins on the business case, and the right thing for your people happens right alongside it.
Growing that specialist from a displaced analyst costs a defined training investment over a few months. Buying one on the open market costs a recruiter fee, a premium salary, a six-month search, and a year of onboarding before they know where anything is buried. The homegrown one already understands your company. And people a company visibly invests in stick around, which matters when your competitors are poaching.
For a mid-market company, there’s a second cost hiding in the standard playbook, and it bites harder here than anywhere. In a large enterprise, knowledge is spread across hundreds of people and layers of documentation. In a mid-market company, it’s concentrated. The way things actually get done often lives in a handful of heads, and the reduction list can’t see which heads. The analyst with fifteen years of pattern recognition and the one hired last spring look identical on a spreadsheet. Cut by title and you find out which one you lost after something breaks. Cut by skills and you knew all along.
None of this requires a moral argument, though the moral outcome comes along for free. People whose work was automated get a real path forward instead of a severance packet and a job board. The company keeps knowledge it can’t buy back. And the cost target still gets hit, because redeploying is cheaper than the sever-and-rehire cycle it replaces.
Why Mid-market Companies Can Actually Win Here
It’s tempting to assume skills-based workforce planning is an enterprise game, something you need a nine-figure learning budget and a workforce analytics team to attempt. The opposite is true.
A 50,000-person enterprise trying to map skills is a multi-year program. A mid-market company redesigning a few functions can do it in weeks, because the raw material already exists. Any serious look at where AI creates value in your workflows already inventories the tasks: what people do, how long it takes, what requires judgment. That’s most of a skills map. The work is running a second lens over it.
In practice, that means four moves.
- Decompose the roles being redesigned into their component skills, and sort each skill into what automates, what transfers, and what stays.
- Decompose the roles you can’t fill the same way.
- Match the two lists and calculate the deltas, so every realistic pathway from a shrinking role to a growing one is visible before reduction decisions get made.
- Build the pathways with real training behind them, and measure success by people successfully moved, not courses completed.
Sequence matters more than anything. This analysis has to run alongside the workflow redesign, not after it. Once the reduction list is final and the severance letters are out, the option to redeploy has already walked out the door.
Know What Your People are Made Of
AI will change what the work looks like. Cost pressure will make the organization leaner. Neither force decides what happens to the people caught between them. That decision belongs to whoever is willing to look past the job titles and see the skills underneath, while there is still time to choose.
The companies that get this right won’t be the ones that cut the least. They’ll be the ones that knew exactly what they were cutting, kept what they were about to go shopping for, and turned a workforce reduction into the fastest talent pipeline they’ve ever built.
Both Sides of the Equation
At ttcInnovations, we work both sides of this equation. Our AI transformation assessments show you where automation creates real lift and ROI in your workflows. Our people practice takes the same data and maps what happens to the humans in those workflows: the skills, the deltas, and the pathways from shrinking roles into the ones you can’t fill.
One looks at the work, one looks at the people, one plan. If your pilots are about to become decisions, let’s talk before the spreadsheet does the thinking.
