Article Series · Labor Optimization
The previous article in this series made the strategic case for salary arbitrage and Global Capability Centers as the highest-return levers in labor optimization. This article moves from strategy to operations — the practical diagnostic tools I use to audit an existing labor force and find the inefficiencies that are hiding in plain sight. These tools apply whether your organization is onshore, offshore, or a combination of both.
Contractor vs. Staff: The Conversion Rule Most Organizations Ignore
Many organizations have a persistent overuse of contractor resources. A practical rule of thumb: if you are using a contractor in the same role for longer than 12 to 18 months, that position should be converted to — or replaced by — a staff member. Contractors are the right instrument for shorter-term, project-specific, or variable-demand needs. When a significant percentage of the organization is made up of contractors, it often signals an overuse of the model.
The overuse of contractors is frequently correlated with an organizational culture that avoids reductions in force. When managers cannot reduce staff, they create contractor pools as a parallel workforce that can be cut without triggering the same internal processes. The problem is that contractors cost more per hour than equivalent staff, so this approach carries a persistent premium.
One important caveat: be careful about separately governed pools of resources. In one organization I worked for, senior management declared a hiring freeze to control costs — and then watched, surprised, as the contractor pool grew by 20% over the following year. The freeze covered staff positions, governed through one process. Contractors were governed through a separate process that nobody had locked down. Any governance mechanism for headcount control must cover all resource types simultaneously.
When negotiating new vendor contracts, include clauses that allow for the conversion of contractors to staff — ideally at no cost. This is difficult to obtain on a renewal where the contractor already has domain knowledge and the vendor has leverage, but much easier when establishing a new relationship.
Skill Set Analysis and Squad Composition Benchmarking
One of the most powerful — and most underused — diagnostic tools is a systematic review of the skill set mix across your organization. Examine the numbers of certain roles, the ratios between them, and how they compare to market norms. The results should be treated as red flags that warrant investigation, not as automatic mandates for action.
Some years ago, I took over a team of 2,800 people covering a wide range of IT services. My analysis showed 40 staff and 210 contractors executing QA functions — approximately 9% of the entire organization. Based on experience, the right number for this function should have been half that at most. After a two-hour session with the staff and the vendor, it was clear the team was significantly overstaffed. The solution was to bring the function in-house within our Global Capability Center. The end result: 15 staff and 108 contractors performing the same work to the same service level agreements.
IT Squad Composition: The Ratios That Matter
The composition of IT development squads is a particularly fertile area for analysis. Producing code is the natural bottleneck for these teams — it takes substantially more time to write code than to test it, design it, or manage it. The following ratios represent baseline expectations for a well-functioning squad:
- QA resources: one for every 4 to 7 developers, depending on the nature of the work
- Application or solution architects: one for every 10 to 50 developers (lower ratio when building new solutions, higher when executing enhancements or maintenance)
- Program managers: one PM overseeing project teams of 15 to 40 total staff
Note: These ratios are changing. AI code generation tools are beginning to shift where the bottleneck sits in software development. As they mature and are adopted more broadly, the developer-to-QA and architect-to-developer ratios will change substantially. This is addressed in detail in the third article of this series.
Management Layers: Where Bloat Hides in Plain Sight
I have worked in too many organizations where — half-jokingly — the first thing a new manager does is hire someone to do the job they were hired to do. The result is a bloated middle management layer that generates overhead without generating proportional value.
I once managed a department with nine levels of management. I eliminated two layers and nothing changed — the remaining managers absorbed the work without issue. That was followed by the elimination of a third layer, again with no negative operational impact.
The metric I use is straightforward: in the middle manager ranks, each manager should have 7 to 14 direct reports, assuming their primary function is people management. Create a list of every middle manager in the organization and the number of people they directly support. Any manager with fewer than 10 direct reports is a candidate for team consolidation — doubling or tripling teams where needed, retaining only the strongest managers.
One nuance worth flagging: first-level managers at the bottom of IT organizations often blur the line between manager and practitioner. Many are effectively team leads who spend the majority of their time as hands-on developers or architects, with a small management component. These should be assessed separately, as the standard span-of-control metric does not apply in the same way.
Combining Teams Across Functions
When consolidating management layers, the ideal approach is to combine functionally or conceptually related teams under a single manager. But there is no operational reason why unrelated teams cannot be combined under a single manager, and I am consistently surprised by the resistance this idea encounters.
At the World Bank, I was managing data center, infrastructure, cloud, and middleware teams when the CIO assigned me responsibility for a new rapid application development program — a function that had nothing to do with infrastructure. The teams operated normally. The management overhead dropped.
Vendor Labor: Apply the Same Discipline
Everything described above applies equally to vendor-supplied labor. Outsourcing vendors are heavily incentivized to propose larger teams than necessary, and the hourly rates they quote carry significant profit margins. When reviewing an outsourced team, benchmark each role against market rates, identify any roles that are mismatched to the actual work being done, and negotiate the basket of roles holistically rather than one rate at a time.
Periodic RFPs on contract labor — every three to four years — are also a powerful tool for resetting rates to market. The competitive pressure alone keeps incumbent vendors honest.