Quick Answer: Choosing a customer service outsourcing company is a structured risk decision, not a procurement exercise. The evidence is unusually clear on what works: in Lacity and Willcocks’ study of UK and US sourcing decisions, selective outsourcing achieved expected cost savings in 85% of cases against 29% for total outsourcing. Scope discipline, verified due diligence and contract completeness predict outcomes far better than any provider’s pitch.
Key Takeaways
- Selective beats total. Outsourcing more than 80% of a function realises expected savings in only 29% of cases, against 85% for selective scope.
- Failure is common and structural. Research on long term service relationships finds 20% to 25% fail within two years and around 50% within five, usually from cost-first deals with weak governance.
- Verify certificates, do not accept logos. UKAS CertCheck confirms whether an ISO 27001 certificate is genuine, current and, critically, whether its scope covers the operation you are buying.
- Ask for attrition data by site and programme. Wage compression that drives churn is the single most predictive commercial red flag.
- Realistic SLAs beat aspirational ones. UK benchmarks sit at 80% of calls answered in 20 seconds, 5% to 8% abandonment and roughly 70% first contact resolution.
- Technology pass-throughs can add 30% to 50% to an apparent per-seat price. Get the full one-time and recurring fee schedule before comparing rates.
What Actually Predicts Outsourcing Success
Decades of outsourcing research point to the same four clusters: strategy clarity, evaluation quality, contract completeness and ongoing governance. None of them are about the provider’s brand.
The most useful single finding for a mid market buyer concerns scope. Lacity and Willcocks’ empirical work across UK and US firms found that total outsourcing, defined as more than 80% of the function, realised expected cost savings in only 29% of cases, against 85% for selective outsourcing and 67% for keeping it in house. A related synthesis rated only 38% of total outsourcing decisions successful against 77% for selective sourcing. That data is now older, drawn from 1990s and 2000s sourcing decisions, and should be read as a structural pattern rather than a current benchmark, but it is consistently corroborated in later reviews.
Beyond scope, the predictors are these. Buyers who define success concretely before selection, choosing explicitly between cost, quality, flexibility and capacity, do better than those who leave it implicit. Research on BPO governance published in Sustainability links process standardisation, detailed contracts covering scope, service levels and exit, and genuine coordination between client and supplier directly to success. Client side capability matters too: organisations with real internal vendor management and process management functions get better quality, cost and timeliness outcomes than those that sign and step back.
Relationship quality, meaning trust, communication and conflict resolution, appears repeatedly as a critical factor, particularly in offshore arrangements. That is not a soft consideration; it is the mechanism by which problems get surfaced early rather than at renewal.
Knowing the predictors is useful. Knowing the specific ways these deals fail is more useful still, because the failure modes are consistent enough to design against.
How Often These Engagements Fail, and Why
Failure rates in outsourcing are high enough that a buyer should plan for the possibility rather than assume success. A synthesis of IS and IT outsourcing research found that 20% to 25% of relationships fail within two years and around 50% within five, with roughly 70% of clients unhappy with at least some aspect of vendor involvement. Broader research into inter-organisational relationships puts failure rates as high as 70%.
These figures come largely from IT and BPO studies conducted between the 1990s and 2010s, and no equally rigorous contemporary dataset exists specifically for contact centre outsourcing. Treat them as evidence that the risk is structural, not as a precise current probability.
The causes are more actionable than the rates. Research synthesising outsourcing failures identifies a consistent set:
- Overemphasis on short term cost reduction. Where cost savings dominate the motive, quality, innovation and sustainability suffer. This is cited as a primary cause of outsourcing failure.
- Poor relationship management. Miscommunication, lack of transparency and failure to involve affected stakeholders in governance.
- Inadequate risk and project management. Ineffective leadership, scope creep and weak risk management, particularly visible in public sector case analyses.
- Weak or incomplete contracts. Vague service levels and thin exit clauses produce disputes and lock-in.
- Insufficient due diligence. Selecting a partner with weak process competence or data protection posture.
- Cultural and language mismatch. Documented cases of companies repatriating contact centres over language barriers and service quality show how this undermines offshore engagements. Our analysis of cultural alignment in outsourced customer support covers how to assess this properly rather than by accent alone.
Read together, the pattern is that cost-first deals with long terms, vague service levels and weak governance are the ones that come back in house. Everything below is designed to prevent that specific outcome.
The Due Diligence That Matters
Structured due diligence is repeatedly identified as a success factor in its own right. Six areas carry most of the weight.
Financial stability. Financially fragile providers cut corners on staffing and defer investment in tooling. Obtain audited accounts and key liquidity, leverage and profitability ratios across three to five years, and check for county court judgments and insolvency events via Companies House and UK credit agencies. For offshore providers, request equivalent national filings plus a bank or auditor letter of good standing.
References, verified properly. Ask for at least three references from UK or European clients of similar size, vertical, volume and complexity, not the provider’s largest global logos. Look for evidence of relationships lasting three to five years or more with specific outcomes attached. Then speak to operational leaders rather than procurement contacts, and probe specifically for transition problems, quality dips, attrition and responsiveness when something went wrong. Our guide to verifying offshore BPO provider references sets out the questions that produce useful answers.
Site visits and live demonstrations. Tour the operation, in person or by live video, and look at floor layout, supervisor ratios, real time dashboards, QA stations and training rooms. Ask for a live walkthrough of the quality assurance workflow, including how non compliant calls are detected and how coaching gets triggered. Observe shrinkage and occupancy reporting and the escalation path for performance problems.
Security certification, independently verified. Covered in detail below.
Agent attrition and pay. Ask for annual agent attrition by site and programme for the last two to three years, average tenure of frontline agents and team leaders on comparable accounts, and starting wages and progression bands against local living wage benchmarks. The purpose of that last item is to establish whether the commercial model depends on wage compression, which reliably produces churn. See offshore call centre attrition rates for the benchmarks to hold this against.
Subcontracting transparency. Require a complete list of sub-processors with locations, roles and certifications, plus contractual prior authorisation rights over any new ones. Opaque delivery chains are both a compliance problem and a quality risk, and they are a recognised warning sign covered in our piece on offshore BPO bait and switch.
Certifications deserve their own treatment, because this is where buyers are most often satisfied by something that means less than they assume.
Verifying Security Credentials Independently
A logo in a pitch deck is not evidence. Each credential covers something specific and excludes a great deal, and all four can be verified independently.
ISO/IEC 27001 certifies an information security management system across a defined scope, audited by a certification body accredited by a national body such as UKAS. Request the certificate showing the certification body, the accreditation mark and reference number, the scope statement and the expiry date, then verify it through UKAS CertCheck, the free service confirming that a certificate is genuine, current and issued by a UKAS accredited body. The scope statement is the part that matters most: a certificate may cover only corporate headquarters rather than the contact centre operation you are buying. ISO 27001 addresses the management system, not specific technical controls, and does not imply PCI DSS or UK GDPR compliance.
SOC 2 Type II is an AICPA attestation framework covering security, availability, processing integrity, confidentiality and privacy. Insist on Type II, which tests operating effectiveness over a three to twelve month period, rather than Type I, which only assesses design at a point in time. Review the audit period, the trust services criteria covered, and whether the opinion carries any qualifications. SOC 2 is deliberately flexible and scoped, and is neither GDPR nor PCI specific.
PCI DSS applies where card payments are handled. Request the current Attestation of Compliance, and the Report on Compliance for Level 1 service providers, and check that the assessment scope includes your contact centre payment flows and that it falls inside the annual validity window. PCI DSS covers cardholder data only and says nothing about personal data in non payment channels.
Cyber Essentials is the UK government backed baseline scheme overseen by the National Cyber Security Centre, covering five technical controls: firewalls, secure configuration, security update management, user access control and malware protection. Cyber Essentials Plus adds independent technical testing. Certification runs on a twelve month cycle, so check the date. It is a baseline and does not substitute for ISO 27001 or SOC 2 in a complex multi-site environment.
Certification establishes control maturity. The contract is what establishes your legal position, and UK GDPR sets a large part of it for you.
UK GDPR Terms Your Contract Must Contain
In a customer service outsourcing arrangement you are almost always the controller and the provider is the processor. The ICO defines the controller as the party determining the purposes and means of processing, and the processor as the party acting only on the controller’s documented instructions. If a provider processes for its own purposes, it becomes a controller for that processing and takes on controller level liability.
Article 28 sets mandatory contract terms, and the ICO specifies exactly what must be included: processing only on documented instructions including on international transfers, confidentiality obligations, appropriate technical and organisational security measures, prior written authorisation for sub-processors with equivalent obligations imposed on them, assistance with data subject rights, breach notification and impact assessments, deletion or return of all data at contract end, and audit and inspection rights.
Where any processing happens outside the UK in a country without an adequacy decision, you additionally need the International Data Transfer Agreement or the UK Addendum to EU standard contractual clauses, attached as an annex and populated with transfer details and extra protection clauses, plus a transfer risk assessment. Government procurement guidance PPN 02/20 requires an appropriate IDTA form for restricted transfers on contracts concluded after 22 September 2022.
Identify every non-UK processing location in the contract and classify each one against adequacy, IDTA or another mechanism. Our guide to outsourcing data security and compliance covers the wider verification picture.
With the legal floor established, the commercial terms determine whether the arrangement actually performs.
Which Pricing Model Fits Your Volume Profile?
Four pricing models dominate UK-facing delivery, and the choice matters more for risk alloca tion than for headline cost. Onshore UK dedicated agents typically run £2,200 to £4,000 per month per seat, or £18 to £40 per hour, with specialised and regulated programmes higher.
| Pricing model | Cost predictability | Volume suitability | Where the risk sits |
| Per FTE (monthly seat) | High. Fixed per agent, easy to budget | Stable or steadily growing volumes; supports deep training and account ownership | Buyer carries volume risk, quiet periods cost the same. Provider carries performance risk via SLAs |
| Per hour | Medium. Rate fixed, total varies with hours | Fluctuating volumes, pilots, overflow and out of hours | Buyer carries volume and some efficiency risk. Provider has no direct incentive to reduce handle time |
| Per minute or per contact | Lower. Unit rate fixed, total depends on volume and handle time | Low, volatile or seasonal volumes; simple queries with short handle time | Provider shares volume risk. Watch for incentives to shorten calls at the expense of resolution unless FCR and CSAT are contracted |
| Outcome-based | Variable. Spend tied to defined outcomes | Discrete measurable outcomes such as sales or simple case closure; poor fit for complex multi-step journeys | Provider carries most performance risk. Buyer must define outcomes tightly to prevent gaming of what counts as resolved |
The trap in the per contact and outcome-based models is the same: any unit of payment becomes a target, so resolution quality has to be contracted explicitly or it will quietly degrade.
Whichever model you choose, the service levels attached to it are what make it enforceable.
Service Levels and Contract Structure
Set service levels against published benchmarks rather than provider proposals. The traditional UK standard is 80% of calls answered within 20 seconds, still the common reference point even where some centres target tighter. ContactBabel’s UK benchmarking reports a mean average speed of answer of 116 seconds against a median of 48, and mean abandonment of 8.4% against a median of 6.0%, with 10% or above generally treated as a warning sign. On resolution, SQM Group’s long running benchmarks place average first contact resolution around 69% to 71%, with 80% or above reached by only about 5% of centres, though that is a North American sample. For satisfaction context, the UK Customer Satisfaction Index scored 76.1 out of 100 in January 2025, rising to 78.3 by mid 2026.
Set service levels per channel rather than one blended figure, and include first contact resolution and CSAT as core outcome metrics rather than speed of answer alone.
On commercial structure, three provisions do most of the protective work:
Service credits, typically a percentage of monthly fees rebated on SLA misses and tiered by severity, commonly around 2% to 5% for minor misses and 10% or more for major or repeated breaches. Earn-back mechanisms, which let a provider recover credits by exceeding targets subsequently, align incentives toward rapid improvement rather than pure punishment. Credits should be meaningful but not so punitive that the provider prices in a risk premium or games the reporting.
Term length. Research ties contracts of seven years and longer to worse cost outcomes. Aim for an initial term of three to five years with clear renewal options.
Exit provisions. Include termination for cause on repeated SLA failure or material security breach, termination for convenience with defined notice, and a specified transition-out plan covering knowledge transfer, data export, and handover of training materials and process documentation. Document who owns scripts, playbooks and QA frameworks, or you will discover at renewal that you cannot leave. Our guide to BPO contracts and SLAs covers these clauses in detail.
The Costs and Red Flags Buyers Miss
Comparing headline rates without the full fee schedule is how apparently cheap providers turn out not to be.
Transition and onboarding fees. Discovery, training, scripting, IVR design, CRM integration and reporting setup are frequently one-time charges quoted separately or not at all. Require full disclosure before comparing.
Technology and licence pass-throughs. Telephony, CCaaS platform, QA and workforce management licences may be bundled into a seat rate or passed through at cost plus margin, and can add 30% to 50% to an apparent per seat price.
Volume minimums and breakage. Per minute and per call models often carry minimum monthly volumes with unused capacity forfeited, which converts variable pricing into a quasi-fixed cost.
Change request pricing. Amendments for scripts, reporting, integrations or new channels can be billed at professional services rates. Require a published change request rate card at contract stage.
Inflation and wage uplift clauses. Without a transparent index-linked mechanism you get discretionary annual increases, or worse, a provider squeezed into wage compression that drives the attrition you were trying to avoid.
Four red flags correlate reliably with later problems: headline rates far below market without a coherent explanation of how quality and compliance are sustained; opaque subcontracting chains where it is unclear which entity employs the agents; absent or unverifiable security credentials; and reluctance to share historical FCR, CSAT, abandonment or attrition data, all of which are standard reporting in any mature UK operation.
If you are still weighing whether to outsource at all, our comparison of BPO versus in-house teams covers that decision, and measuring BPO ROI sets out how to build the business case. For IT support rather than customer service, see IT outsourcing, which is a distinct discipline with different evaluation criteria.
Frequently Asked Questions
How do I choose a customer service outsourcing company in the UK?
Treat it as a structured risk decision across four areas that research consistently links to success: define what success means before you shortlist, run genuine due diligence on financials, references, operations, certifications and attrition, write a complete contract covering scope, service levels and exit, and resource the governance function on your side. Scope discipline matters most. Selective outsourcing realises expected cost savings in 85% of cases against 29% for total outsourcing.
What proportion of outsourcing engagements fail?
Research into IS and IT outsourcing relationships found 20% to 25% failing within two years and around 50% within five, with roughly 70% of clients dissatisfied with some aspect of the relationship. These figures come mainly from studies conducted between the 1990s and 2010s and no equally rigorous contemporary dataset exists for contact centre outsourcing specifically, so treat them as evidence of structural risk rather than a precise current probability.
How do I verify a provider’s ISO 27001 certificate is genuine?
Request the certificate showing the certification body, the accreditation mark and reference number, the scope statement and the expiry date, then check it through UKAS CertCheck, which confirms whether the certificate is current and whether the issuing body is UKAS accredited for ISO 27001. Read the scope statement carefully: certificates frequently cover corporate headquarters rather than the specific contact centre operation and data processing you are buying.
What service levels should I contract for?
Use published benchmarks rather than provider proposals. The UK reference standard is 80% of calls answered within 20 seconds. UK benchmarking shows mean abandonment of 8.4% and a median of 6.0%, with 10% or above treated as a warning sign, and average first contact resolution around 69% to 71%. Set levels per channel rather than blended, and contract first contact resolution and CSAT as outcome metrics alongside speed of answer.
Which outsourcing pricing model should I use?
Per FTE suits stable or growing volumes and supports deep training, with the buyer carrying volume risk. Per hour suits fluctuating volumes, pilots and overflow. Per minute or per contact suits low, volatile or seasonal volumes with simple queries. Outcome-based suits discrete measurable results and transfers most performance risk to the provider. In the per contact and outcome models, contract resolution quality explicitly, because whatever unit you pay for becomes the target.
What contract length should I agree?
Three to five years with clear renewal options. Research ties contracts of seven years and longer to worse cost outcomes than shorter terms. Include termination for cause on repeated SLA failure or material security breach, termination for convenience with defined notice, and a specified transition-out plan covering knowledge transfer, data export and handover of documentation, with explicit ownership of scripts, playbooks and QA frameworks.
What hidden costs should I expect?
Transition and onboarding fees for discovery, training, scripting, IVR design and CRM integration; technology and licence pass-throughs for telephony, CCaaS, QA and workforce management, which can add 30% to 50% to an apparent seat rate; volume minimums with forfeited unused capacity; change request pricing at professional services rates; and annual inflation or wage uplift clauses. Require the full one-time and recurring fee schedule plus a change request rate card before comparing providers.
Why does agent attrition matter when selecting a provider?
Attrition drives quality problems and training cost, and is largely a function of pay and working conditions. Ask for annual attrition by site and programme over two to three years, average tenure on comparable accounts, and starting wages against local living wage benchmarks. A model that depends on aggressive wage compression will produce churn regardless of what the service levels say, which is why pay data belongs in due diligence rather than being treated as the provider’s internal business.
Afrishore BPO delivers customer support for UK and US businesses from South Africa, with attrition and pay data disclosed during due diligence rather than after signature, certification scope written to cover the delivery operation itself, and contracts that specify transition-out from the start. If you are comparing delivery locations, outsourcing to South Africa for UK businesses sets out the time zone and language position, and call centre outsourcing in South Africa covers voice operations specifically.
Talk to our team about scoping a customer support programme: https://afrishorebpo.com/business-process-outsourcing/
