Business process outsourcing is the alternative most mid-market companies weigh up first, and for good reason. For high-volume, standardised work it is genuinely effective, and the industry is enormous precisely because it solves a real problem. Where it tends to disappoint is when a company hands over a process that is still evolving, or one that turns out to be closer to the core of the business than it looked on the day the contract was signed.
This guide sets out what BPO actually is, what it costs, where it works well, and where the structure creates friction. It then compares it with an embedded offshore team, so you can match the model to the process rather than to the sales pitch.
In short: Use BPO when you want to hand a high-volume, standardised process to a specialist and stop managing it. Use an embedded offshore team when the process is bespoke or still changing, when client experience depends on it, or when you want the institutional knowledge to accumulate inside your business rather than inside a vendor. The deciding question is not cost, it is who should own the process in three years' time.
Business process outsourcing is the practice of contracting an entire business process to a third-party provider, who then runs it using their own people, their own operating procedures and usually their own technology. You govern the relationship through a service-level agreement and periodic reviews rather than by managing the individuals doing the work.
Three structural features define the model. The provider owns the process, including how it is performed and how it changes. Delivery usually happens in a multi-tenant centre, where the same floor and often the same agents serve several clients. And pricing is transactional, typically per seat, per full-time equivalent or per transaction. Those three things are what make BPO efficient at scale, and they are also the source of every trade-off discussed below.
The scale is substantial. Global business process services spend approaches 300 billion US dollars a year according to Everest Group, and the customer-experience segment alone was valued at around 113 billion dollars in 2025, rising to roughly 126 billion in 2026. The UK is one of the most open markets in Europe to offshore outsourcing, and the broader UK outsourcing market accounts for over a quarter of European activity.
Two shifts are worth noting. Growth among the largest providers has slowed markedly, from 7.5 to 8.5 per cent down to 4.5 to 5.5 per cent, which suggests the straightforward labour-arbitrage proposition is maturing. And the industry is moving toward outcome-based pricing and AI-augmented delivery, both of which are genuine improvements on pure per-seat models. Deloitte's global outsourcing research points the same way: buyers increasingly want measurable results rather than bodies on seats.
There are clear situations where BPO is not just acceptable but the best available answer, and it would be dishonest to pretend otherwise.
Choose BPO when
The process is high-volume, standardised and unlikely to change much. You genuinely want to hand it over and stop thinking about it. The provider's existing platform, accreditation or scale is itself the reason for going, and rebuilding it in-house would be uneconomic. Volumes are seasonal or highly variable and you want that variability to sit on someone else's balance sheet. Or the function is genuinely non-core, where good-enough execution at low cost beats excellence.
Classic examples that fit well: high-volume claims processing, payroll administration, standardised accounts payable, and first-line support for a stable product with a mature knowledge base. If your process looks like that, a competent BPO will very likely serve you better and cheaper than building anything yourself.
The difficulties are structural rather than a matter of provider quality. A good BPO run well will still produce them, because they follow from the model.
Change costs money and time. Because the provider owns the process and the contract specifies it, every material change becomes a change request with commercial negotiation attached. That is manageable for a stable process and genuinely painful for one that is still evolving. Many mid-market firms discover their process was less settled than they assumed.
Attrition erodes knowledge. Call centre turnover typically runs at 40 to 45 per cent a year, reaching 55 to 60 per cent in high-stress environments, with UK figures around 30 to 45 per cent. Replacing an agent costs somewhere between 10,000 and 20,000 dollars once recruitment, training and lost productivity are counted. You do not see that cost directly, because it sits inside the provider's rate, but you do feel it every time someone who understood your business is replaced by someone who does not.
Agents are shared. In a multi-tenant centre, the people handling your work often handle other clients' too. That is what makes the economics work, and it caps how deeply anyone learns your particular business.
Knowledge accumulates in the wrong place. This is the one that matters most over time. Every process improvement, every hard-won lesson about your customers, every workaround for the awkward edge case, all of it builds inside the vendor's operation. Three years in, they understand how your process works better than you do. That is comfortable until the day you want to change provider or bring it back in-house.
Pricing takes three common shapes. Per hour, with fully-loaded onshore contact centre rates around 30 to 45 US dollars, rising to 45 to 65 for licensed or certified roles. Per full-time equivalent, with offshore FTEs commonly 1,200 to 2,500 dollars a month. Or per transaction, often 1 to 5 dollars per ticket. All three include the provider's margin, their overhead, their recruitment costs and the cost of their attrition.
Judging the models on the headline rate alone is where most comparisons go wrong. Over a three-year horizon you also need to count transition and onboarding costs at the start, change requests as the process evolves, your own management and governance time, and the exit or knowledge-transfer cost at the end. Those are real and they are rarely in the initial business case.
An embedded offshore team model inverts the ownership question. A partner recruits, employs and houses a dedicated team for you, and provides the office, local management, HR and compliance. You keep the process. The team works in your systems, follows your procedures, joins your performance cycles and reports through your management. Nobody else's work goes through them.
Commercially you are paying for headcount and tools rather than a per-seat rate carrying vendor margin, which typically lands 30 to 60 per cent below equivalent UK hires on a fully-loaded basis. But the structural difference matters more than the price. Because the process stays yours, changing it is a Tuesday morning conversation rather than a change request. Because the team is dedicated and stable, knowledge compounds inside your business. And because there is no vendor owning the operation, there is nothing to be locked into.
| Dimension | Traditional BPO | Embedded offshore team |
|---|---|---|
| Who owns the process | The provider, to an agreed specification | You, entirely |
| Changing how it works | Change request, negotiated and often chargeable | You just change it |
| Who the people work for | Often shared across several clients | Dedicated to you alone |
| Pricing | Per seat, per FTE or per transaction, including margin | Headcount plus tools, no vendor margin |
| Where knowledge accumulates | Inside the provider's operation | Inside your business |
| Governance | SLAs and service reviews | Normal line management |
| Exit | Contractual, with knowledge-transfer risk | The team and its knowledge are already yours |
This deserves more attention than it usually gets at signature. Exiting a BPO arrangement means transferring back a process that has been run by someone else, often for years, using their procedures and their documentation, staffed by people who do not work for you and will not be moving. Well-drafted agreements include exit assistance, termination rights and change-of-control provisions, and these are worth negotiating properly at the start rather than discovering later.
The practical risk is not the contract, it is the knowledge. If the only people who fully understand your process work for your outgoing provider, the transition is expensive and disruptive however good the paperwork. This is the specific problem the embedded model avoids, because there is nothing to hand back.
Worth checking before you sign
Ask what exit assistance is included and for how long, who owns the process documentation, how change requests are priced, whether your agents are dedicated or shared, and what the provider's agent attrition rate actually is. The answers to those five questions tell you more about the next three years than the headline rate does.
| If your process is | Best model | Why |
|---|---|---|
| High-volume, standardised, stable, non-core | BPO | Scale and specialisation you would not rebuild economically |
| Seasonal or highly variable in volume | BPO | Variability sits with the provider, not your payroll |
| Bespoke, evolving, or differentiating | Embedded team | Change without change requests, knowledge stays with you |
| Client-facing, where experience is the product | Embedded team | Dedicated people, your standards, lower churn |
| A whole function you intend to keep and grow | Embedded team | You are building capability, not buying throughput |
Not sure which side your process falls on?
If it is genuinely commoditised, a BPO will serve you well and we will say so. If it is closer to your core than that, see how Potentiam builds embedded offshore teams across four global hubs.
Explore Our SolutionsNot a BPO, by design
Potentiam is not a business process outsourcer, a recruiter or a platform. We design, build and run embedded offshore teams across four hubs in South Africa, Romania, India and Brazil, providing offices, local management, an embedded HR business partner and structured onboarding, while you keep the process and the people. Our founders scaled an energy procurement business to more than 300 employees, around 60 per cent of them offshore, before its acquisition by Accenture in 2015. You can read more about our story, or compare the other routes in our guides to recruitment and staffing agencies and Employer of Record platforms.
What is the difference between BPO and an embedded offshore team?
With BPO, the provider owns the process and runs it their way to a service-level agreement, usually with staff shared across several clients. With an embedded offshore team, you own the process and the team is dedicated to you, working in your systems and reporting through your management, while a partner provides the office, employment, HR and compliance. BPO buys you throughput; an embedded team builds you capability.
Is BPO cheaper than an embedded offshore team?
On headline rate it often looks cheaper, because the price is quoted per seat or per transaction. Over three years the comparison usually narrows once you include change requests as the process evolves, your own governance time, and transition costs at both ends. An embedded team is priced on headcount and tools without vendor margin, typically 30 to 60 per cent below equivalent UK hires, and it does not carry per-change pricing.
When should we choose BPO over an embedded team?
When the process is high-volume, standardised, stable and genuinely non-core, when volumes are seasonal and you want that variability off your payroll, or when the provider's platform, accreditation or scale is itself the reason for going. In those cases BPO is the better answer and usually the cheaper one.
What are typical BPO pricing models?
Three are common. Per hour, with fully-loaded onshore contact centre rates around 30 to 45 US dollars and 45 to 65 for licensed roles. Per full-time equivalent, with offshore FTEs commonly 1,200 to 2,500 dollars a month. And per transaction, often 1 to 5 dollars a ticket. The industry is also moving toward outcome-based pricing, where you pay for results rather than time.
Why does agent attrition matter if it is the provider's problem?
Because you pay for it twice. It is priced into the rate you are quoted, and you feel it in quality every time someone who understood your business is replaced. With call centre turnover typically 40 to 45 per cent a year and replacement costs of 10,000 to 20,000 dollars per agent, it is one of the largest hidden costs in the model. Dedicated, embedded teams generally hold on to people longer because they have career paths and belong to one organisation.
How hard is it to exit a BPO contract?
The contractual side is manageable if you negotiated exit assistance, termination rights and documentation ownership at the outset. The practical difficulty is knowledge: after a few years the people who fully understand your process work for the provider and will not be transferring. Budget properly for transition and insist on process documentation you own throughout, not just at the end.
Keep the process, lose the cost
If your outsourced process has become harder to change than it should be, there is another structure. Talk to us about an embedded offshore team at 30 to 60 per cent lower cost, with the process and the knowledge staying with you.
Book a Discovery CallSources: Everest Group, BPS Top 50 (2024); Deloitte, Global Outsourcing Survey; Grand View Research, Customer Experience BPO market; Insignia Resource, call centre turnover rates; CBI, IT outsourcing trade statistics.