The Problem
How Outsourcing Decisions Get Made Wrong
In most organisations, the insource vs. outsource conversation starts in the wrong place: with a vendor presentation. A BPO provider walks in with a cost model that shows significant savings, a reference list, and a service level commitment. Leadership sees a number that looks lower than the current run rate and asks the ops team to evaluate it.
The problem is that vendor-built cost models are designed to justify the vendor’s offering. They compare the visible cost of the internal operation against a quoted per-seat rate that bundles all the vendor’s own costs invisibly. The internal operation’s full cost structure is understated; the outsourced model’s full cost structure is obscured.
At Tyco/JCI, with a total budget approaching $9M across North America, EMEA, and APAC, the economics of certain support functions warranted genuine analysis. The framework I developed was designed to make that analysis from a neutral starting point — not a vendor deck.
“The vendor will always build a model that favours outsourcing. Your job is to build a model that favours neither — and see what the numbers actually say.”
The Framework
Six Cost Categories Most Analyses Miss
A complete insource vs. outsource comparison requires accounting for six categories of cost. Most analyses miss at least two on the insource side — and at least two on the outsource side. The result is a comparison that looks rigorous but is structurally biased.
01
Direct Labour Cost (Insource)
Base salary plus benefits, employer-side contributions, paid leave, and variable compensation. In a multi-region operation this requires currency conversion — the cost of an L1 agent in Brossard is different from Echt or Shanghai. The Tyco/JCI budget spanned GBP, EUR, USD, and CAD across seven cost centres.
02
Management and Overhead (Insource)
The fully loaded cost of the management layer — regional managers, supervisors, quality leads, trainers — that doesn’t disappear when front-line agents are outsourced. This is the most commonly understated cost in insource modelling.
03
Infrastructure and Tooling (Both Models)
Telephony, CRM/ticketing, hardware, facilities, and software licences. In the outsource model, clarifying which tools stay with the client is essential — the assumption that “the vendor handles everything” is frequently wrong for ticketing systems and knowledge bases that remain under client ownership.
04
Transition and Implementation Cost (Outsource)
The one-time cost of moving to an outsourced model: knowledge transfer, vendor training, process documentation, system integration, and the productivity dip during transition. These costs rarely appear in vendor proposals but are real and significant.
05
Quality and Control Risk Premium (Outsource)
The cost of reduced control over the customer experience. An outsourced support team operates under a contract, not a culture. The alignment between a BPO agent’s incentives and the client’s customer experience objectives is imperfect by design.
06
Institutional Knowledge and Scalability (Both Models)
The long-term cost of knowledge concentration in an external organisation. An outsourced team that accumulates three years of product knowledge represents a significant dependency — and a renegotiation leverage point for the vendor.
The Multi-Currency Challenge
Why This Was Harder Than a Single-Region Analysis
Running this analysis across a $9M multi-region operation required a modelling layer that most single-region analyses don’t need. The Tyco/JCI budget spanned Technical Support UK (GBP), Technical Support Europe (EUR), Kantech and SWH in Brossard (CAD), DSC (USD), and newly onboarded teams in Bangalore and Shanghai.
The analysis distinguished between cost centres where insourcing was clearly more competitive — typically specialised technical support functions where product knowledge depth was the differentiator — and cost centres where the economics of volume-based BPO were more compelling. Not all support functions have the same insource/outsource profile, and treating the entire operation as a single decision typically leads to either over-outsourcing or under-outsourcing.
What We Learned
The Decisions Beyond Cost
Build the model before the vendor builds it for you. An internal model, built before vendor engagement, is the only way to enter a vendor conversation with a neutral baseline. Once the vendor’s model is on the table, it anchors the conversation.
The transition cost is the most ignored variable. The savings from an outsourcing arrangement are typically realised over 18–36 months. The transition costs are incurred in the first 6–12. An analysis that shows positive economics over three years but negative economics over one year is a very different risk profile.
Not all functions have the same outsource profile. The decision to outsource L1 volume-based support and retain L2 specialised support is often more economically rational than treating the entire operation as a single decision.
The strategic knowledge dependency is a real cost. Three years after an outsourcing arrangement begins, the vendor team knows your product and your customers in ways your own organisation may have forgotten. That dependency has a renegotiation cost that should appear in a 5-year model.
Insource vs. OutsourceCost AnalysisShared Services
BPO EvaluationBudget ManagementGlobal OpsMake vs. Buy