Procurement gets the rate down four percent and everyone celebrates. Eighteen months later the program costs more than the in-house team it replaced, and no single decision explains why. The rate was never the variable. Outsourcing total cost of ownership is decided almost entirely by lines that do not appear in the proposal — and the proposal is the only document most buyers actually compare.
This is not a new observation, which is part of what makes it strange that it keeps happening. A study by Jérôme Barthélemy in MIT Sloan Management Review in April 2001, drawing on 50 outsourcing efforts, identified four categories of hidden cost — vendor search and contracting, transition, ongoing management of the relationship, and eventual switching or reintegration — and found companies largely unaware of them, and therefore taking no measures to reduce them.
The context has shifted since, but not in the direction that would make this easier. Deloitte’s 2024 Global Outsourcing Survey of more than 500 executives found cost reduction had fallen to 34% as the primary outsourcing driver, down from 70% in 2020, while 70% of executives described their own vendor management function as not fully mature. Buyers are outsourcing for reasons other than cost, using governance they admit is underbuilt, and still comparing vendors on price per hour.
The total cost of ownership lines that never reach the comparison sheet
Nine categories account for most of the gap between quoted cost and actual cost. Where they land matters as much as their size, because several of them are absorbed by the client’s own budget and therefore never appear as an outsourcing cost at all.
| Cost category | Whose budget it hits | Typically visible in the proposal? |
|---|---|---|
| Hourly or per-seat rate | Vendor invoice | Yes — the only line most buyers negotiate |
| Vendor search, RFP and legal review | Client, pre-contract | No |
| Transition and parallel running | Both, simultaneously | Rarely, and never in full |
| Ramp-period productivity gap | Vendor invoice at full rate, client in service quality | No |
| Client-side program management | Client headcount | No |
| Technology, licences and integration | Usually client | Partially |
| Requalification after vendor attrition | Absorbed silently by both | No |
| Escalation and rework caused by gaps | Client, as repeat contacts | No |
| Exit, knowledge recovery and reintegration | Client, at end of term | Almost never |
Two of these decide most outcomes. The rest are real, but rarely large enough on their own to reverse a decision.
Transition and parallel running
For a period, the buyer pays twice. The incumbent team is still handling volume because the new team cannot yet, and the new team is billing while it learns. Barthélemy’s work put average transition duration at around a year for the IT engagements studied; contact center transitions are usually shorter, but the mechanism is identical and the duration is systematically underestimated in planning.
The reason it gets underestimated is that transition is defined as complete when the last cohort certifies. It is actually complete when the new team resolves at the rate the old team did — a different date, and never the one in the plan. Anyone treating ramp timelines as a real cost rather than a formality will size this correctly; almost nobody does on the first engagement.
Client-side management
Someone has to run the relationship. Weekly calls, QA review, escalation handling, forecast reconciliation, quarterly business reviews, and the two or three people who now spend a meaningful share of their week on vendor governance rather than on the work.
This cost is invisible because it lands in an existing salary line. It does not show up as outsourcing spend anywhere, which is exactly why Deloitte’s finding about immature vendor management functions matters commercially rather than administratively: an underbuilt governance function does not save money, it moves the cost somewhere it cannot be measured.
Where geography changes the arithmetic
The categories above behave differently depending on where the team sits, which is the honest case for evaluating a bpo mexico option against both a domestic build and a distant offshore one. Shared working hours compress the management line, because governance conducted inside one business day takes fewer calendar days. Proximity makes transition oversight cheap enough to actually do. And a labour market where support is a durable career shortens the requalification cycle.
None of that shows up as a lower rate. Nearshore quotes usually sit above the cheapest offshore ones, which is precisely why the comparison has to be run on total cost rather than unit price — and why teams that only ever compare support costs at the line-item level reach conclusions that do not survive contact with the second year.
The practical test is a single number: fully loaded cost per resolved contact, over a full year, including the ramp months. Vendors will not volunteer it because it is worse for them than the rate. Ask anyway, and build your own version from the categories in the table.
FAQ: Evaluating Outsourcing Total Cost of Ownership
It is the full cost of an outsourcing arrangement across its life — search and contracting, transition, the ongoing rate, client-side management, technology, and eventual exit — rather than the vendor’s quoted price alone. Research on hidden outsourcing costs has identified these categories consistently since at least 2001.
Transition and parallel running, followed by client-side management. The first is concentrated in the first few months and is easy to underestimate; the second is permanent and rarely counted at all because it lives inside existing salaries.
There is no reliable universal multiplier, and any figure quoted without reference to your scope should be treated with suspicion. Build the estimate from your own categories: name each cost, assign it to a budget owner, and put a number on it even if the number is rough.
Sometimes, for simple, high-volume, low-judgment work with stable requirements. The rate advantage erodes as complexity, escalation rates and turnover rise, because those are exactly the conditions that inflate the categories the rate does not cover.
Convert both to cost per resolved contact over twelve months, including ramp, and require each vendor to state their assumed attrition and time-to-proficiency. Two quotes that look four percent apart on rate frequently sit twenty percent apart on that basis.
Leave a Reply