How a Lean HR Team Built a Full Pay Structure in Nine Months Using AI as a Force Multiplier
With AI, I think it's possible to do things with a limited staff that you couldn't do before, and do them much more quickly.
Mark McCann
senior HR executive
TRC Talent Solutions
When an established equipment manufacturer with nearly 3,000 employees decided its pay practices had fallen behind the market, it had no job architecture, no career pathways, and almost no one to build them. The person hired to lead the work had already left. Instead of staffing up, the company paired one experienced practitioner with AI and compressed a project that usually takes years into nine months.
Mark McCann is a senior HR executive specializing in total rewards and performance systems who took the engagement through TRC Talent Solutions. He has led compensation and performance work across organizations of every size, from small companies to global SAP SuccessFactors and Workday implementations. He came into this one as a rescue.
“They needed somebody who had the expertise, who had seen this before and done this work before, to come in and get it back on track by the end of the year,” McCann says. He started in October and had the original scope on track by December. Then the real size of the project revealed itself.
Why the timing caught up with them
For years, the manufacturer operated as what McCann calls an automatic pay shop. It hired good people, trained them well in specific procedures, and moved pay based on tenure and internal habit rather than job value or market data. A raise was a matter of time served, not where a role sat against the market, and for a long time that worked because the company was the dominant employer in its area.
Then the local labor market changed, and competition for talent followed. “They had realized that they weren’t really matching the market. They didn’t really have great tools to see whether their jobs were competitive,” McCann says. The company also wanted something its old model never offered: transparency and real pathways for growth. Employees came in as temps, became company members, and moved into other jobs, but nothing was written down. No published routes for blue-collar or white-collar workers. With nearly 3,000 people on the payroll, a large share of whom were exempt professionals, the informal approach had run out of room.
Where AI filled the gap
McCann’s real constraint was people. The organization was lean, and there was no room to staff a full project team the way this work usually demands. “It was me and the executive team in HR, and principally me and one other person who paired up to drive this forward,” McCann says. A build of this scope, including new job descriptions, core and functional competency frameworks across job families, matching performance management forms, and both exempt and non-exempt tracks, would typically require a full project team over many months. He ran it with two people between October and a July handoff, using AI to cover the seats he did not have.
Job descriptions were the first test. He fed AI what managers and subject-matter experts gave him, plus the specifics of the company and its location, and drafted new ones far faster than he could alone. The output improved the more he did. He used the same approach to build the competency framework the company had never had. Getting job descriptions right matters well beyond paperwork. Accurate job content is the foundation for benchmarking roles against market surveys.
Where his judgment came in
Pay decisions leave no room for guesswork, so McCann never treated AI output as finished. He leaned on his own experience to catch what the model missed, then ran drafts past his project partner and the managers who would have to live with them, the kind of human review that catches what automation alone lets through.
“I never assume that AI is always completely right. In many cases it did have some hiccups and some hallucinations, and I could tell from my experience that they missed something,” McCann says. The test was always whether the result fit the company, down to whether a competency definition sounded like how a manager would actually describe it.
The workflow inverted the old build cycle. Instead of a team writing and editing in endless rounds, McCann interviewed the reviewers first, used AI to build a framework, applied his judgment, and delivered something concrete for them to tweak. The company still owns the finishing work, revisiting the functional competency definitions and the role-specific examples, but it started from a structure rather than a blank page.
“If I hadn’t leveraged AI to get us a framework, I think it could possibly never have gotten done, or it could have spun and spun for weeks,” McCann says.
When a smaller company should start
The manufacturer is also a case for starting this work earlier. The questions employees ask, what is my job, how do I grow, how do I make more money, do not change with headcount. What changes is how painful it becomes to answer them once an organization reaches thousands of employees without a system.
“With AI, I think it’s possible to do things with a limited staff that you couldn’t do before, and do them much more quickly,” McCann says. He believes in building something workable and iterating rather than waiting for a perfect system. “Everybody who goes into a company has those questions. To the extent that the company can answer them definitively, that’s to the company’s advantage,” McCann says. Smaller organizations, he adds, have employees no less important than those at any blue-chip firm, and plenty of blue-chip firms never build this at all.
The cost of waiting
For companies weighing whether to invest, McCann starts with turnover in the roles they most need to protect. In his experience, being able to show employees how they can grow their value and their pay can help keep turnover down. Succession is the second measure. Clear levels and pathways create a stronger pool of internal candidates.
The foundation under all of it is job content. The manufacturer had never maintained accurate job descriptions, because it never had to. Without them, there is no reliable way to benchmark a role or show an employee the distance between where they are and where they want to go.
“You can’t really draw that map out unless you have job descriptions with competencies and with levels,” McCann says. Map the gap, and you can build the development plan that turns an employee into a candidate. He handed off the major pieces on July 10, and the company will keep iterating from there. His next move is toward a smaller company where he can do this kind of work from the ground up. “I’m a builder. It was fun to create something that didn’t exist before that’s going to have real value for the entire organization,” McCann says.
Related articles
TL;DR
Mark McCann
TRC Talent Solutions
senior HR executive
senior HR executive
When an established equipment manufacturer with nearly 3,000 employees decided its pay practices had fallen behind the market, it had no job architecture, no career pathways, and almost no one to build them. The person hired to lead the work had already left. Instead of staffing up, the company paired one experienced practitioner with AI and compressed a project that usually takes years into nine months.
Mark McCann is a senior HR executive specializing in total rewards and performance systems who took the engagement through TRC Talent Solutions. He has led compensation and performance work across organizations of every size, from small companies to global SAP SuccessFactors and Workday implementations. He came into this one as a rescue.
“They needed somebody who had the expertise, who had seen this before and done this work before, to come in and get it back on track by the end of the year,” McCann says. He started in October and had the original scope on track by December. Then the real size of the project revealed itself.
Why the timing caught up with them
For years, the manufacturer operated as what McCann calls an automatic pay shop. It hired good people, trained them well in specific procedures, and moved pay based on tenure and internal habit rather than job value or market data. A raise was a matter of time served, not where a role sat against the market, and for a long time that worked because the company was the dominant employer in its area.
Then the local labor market changed, and competition for talent followed. “They had realized that they weren’t really matching the market. They didn’t really have great tools to see whether their jobs were competitive,” McCann says. The company also wanted something its old model never offered: transparency and real pathways for growth. Employees came in as temps, became company members, and moved into other jobs, but nothing was written down. No published routes for blue-collar or white-collar workers. With nearly 3,000 people on the payroll, a large share of whom were exempt professionals, the informal approach had run out of room.
Where AI filled the gap
McCann’s real constraint was people. The organization was lean, and there was no room to staff a full project team the way this work usually demands. “It was me and the executive team in HR, and principally me and one other person who paired up to drive this forward,” McCann says. A build of this scope, including new job descriptions, core and functional competency frameworks across job families, matching performance management forms, and both exempt and non-exempt tracks, would typically require a full project team over many months. He ran it with two people between October and a July handoff, using AI to cover the seats he did not have.
Job descriptions were the first test. He fed AI what managers and subject-matter experts gave him, plus the specifics of the company and its location, and drafted new ones far faster than he could alone. The output improved the more he did. He used the same approach to build the competency framework the company had never had. Getting job descriptions right matters well beyond paperwork. Accurate job content is the foundation for benchmarking roles against market surveys.
Where his judgment came in
Pay decisions leave no room for guesswork, so McCann never treated AI output as finished. He leaned on his own experience to catch what the model missed, then ran drafts past his project partner and the managers who would have to live with them, the kind of human review that catches what automation alone lets through.
“I never assume that AI is always completely right. In many cases it did have some hiccups and some hallucinations, and I could tell from my experience that they missed something,” McCann says. The test was always whether the result fit the company, down to whether a competency definition sounded like how a manager would actually describe it.
The workflow inverted the old build cycle. Instead of a team writing and editing in endless rounds, McCann interviewed the reviewers first, used AI to build a framework, applied his judgment, and delivered something concrete for them to tweak. The company still owns the finishing work, revisiting the functional competency definitions and the role-specific examples, but it started from a structure rather than a blank page.
“If I hadn’t leveraged AI to get us a framework, I think it could possibly never have gotten done, or it could have spun and spun for weeks,” McCann says.
When a smaller company should start
The manufacturer is also a case for starting this work earlier. The questions employees ask, what is my job, how do I grow, how do I make more money, do not change with headcount. What changes is how painful it becomes to answer them once an organization reaches thousands of employees without a system.
“With AI, I think it’s possible to do things with a limited staff that you couldn’t do before, and do them much more quickly,” McCann says. He believes in building something workable and iterating rather than waiting for a perfect system. “Everybody who goes into a company has those questions. To the extent that the company can answer them definitively, that’s to the company’s advantage,” McCann says. Smaller organizations, he adds, have employees no less important than those at any blue-chip firm, and plenty of blue-chip firms never build this at all.
The cost of waiting
For companies weighing whether to invest, McCann starts with turnover in the roles they most need to protect. In his experience, being able to show employees how they can grow their value and their pay can help keep turnover down. Succession is the second measure. Clear levels and pathways create a stronger pool of internal candidates.
The foundation under all of it is job content. The manufacturer had never maintained accurate job descriptions, because it never had to. Without them, there is no reliable way to benchmark a role or show an employee the distance between where they are and where they want to go.
“You can’t really draw that map out unless you have job descriptions with competencies and with levels,” McCann says. Map the gap, and you can build the development plan that turns an employee into a candidate. He handed off the major pieces on July 10, and the company will keep iterating from there. His next move is toward a smaller company where he can do this kind of work from the ground up. “I’m a builder. It was fun to create something that didn’t exist before that’s going to have real value for the entire organization,” McCann says.