The question usually comes when something has grown too big: tickets are growing faster than the team, evenings and weekends are unstaffed, or a campaign is launching in three new markets. Then someone in management turns up with a quote from a contact centre, and the head of customer service has to say whether it is a good idea. I have sat on both sides of that table, as a buyer of outsourced customer service and as a provider of it. This article is about what decides.
Why do companies outsource customer service in 2026?
Not primarily to save money, even though that is how it usually starts. In Deloitte's global outsourcing survey 2024 34 percent of companies named cost as the main reason, against 70 percent in the 2020 survey. The most common reason was instead rising customer expectations, named by 35 percent. Companies outsource to get opening hours, languages, volume peaks and a ready-made organisation they do not have time to build themselves.
That matches what I have seen in Swedish companies. What drives the question is rarely the wage cost per hour, but three practical things:
- Volume swings. An online store with 300 tickets a week in March and 1,200 in December cannot hire for December.
- Opening hours fall short. Customers shop in the evenings and at weekends, the team works office hours.
- Languages multiply. Norwegian, Danish and Finnish customers expect answers in their own language, and three part-time posts are hard to staff.
What do you hand over when you outsource?
You hand over the execution, and if you are not careful you also hand over the knowledge and the customer insight. The execution, answering the tickets, is what the provider is good at. The knowledge, what the answers should be and why, is what you are good at. The problem arises when that knowledge exists only in the provider's heads and templates, and walks out the door when the contract expires or the agents are replaced.
Three things should therefore always stay with you, whoever answers:
- The knowledge base. The answers to customers' questions, reviewed by whoever owns the subject, in a system you own. The provider answers from it, not from its own documents. What such a knowledge base is and how it differs from an FAQ is described in the article on knowledge bases.
- The exceptions. Decisions on compensation, exceptions to policy and complaints that put the customer relationship at risk. The provider should have a clear route to hand them over, and you should have someone who receives them.
- The insight. Which questions come in, which answers are missing and what customers say in the comments. That is the input for product, logistics and the website, and it should land with you every week, not in a quarterly report.
The Swedish industry association Kontakta has a guide for buyers that walks through the outsourcing process in ten steps, from starting point and goals via requirements and provider selection to contract. Much of the work sits with the buyer before a provider is chosen: whoever cannot describe ticket types, volumes per hour and what a good answer is gets a quote for the wrong thing.
What does it cost, and what does the part not in the quote cost?
The quote is priced per hour, per ticket or per staffed seat, and the hidden cost is the work you must do yourselves so the provider can answer correctly. The pricing models have different logic:
| Pricing model | You pay for | Fits when | Risk |
|---|---|---|---|
| Per hour or per staffed seat | Time, regardless of ticket count | Steady volume, complex tickets | You pay for idle time in low season |
| Per ticket | Every handled ticket | Swinging volume, simple tickets | The provider earns from more tickets, not fewer |
| Per outcome, such as resolved ticket or customer satisfaction | Results | Mature relationship with good measurement | Requires that you can measure, otherwise it becomes an argument |
The per-ticket model deserves a comment. It gives the provider revenue per contact, while you want fewer contacts through better self-service and clearer information. That is not a reason to avoid the model, but a reason to write into the contract that the provider must report recurring questions and propose how to remove them, and that the fee does not fall when they succeed.
Then come the costs that are not in the quote: writing and maintaining the knowledge base, training at every staff change at the provider, quality-checking a sample of tickets every week, and having a person at your end who owns the relationship. Without that person outsourcing does not get cheaper, only less visible.
How does AI change the calculation?
AI makes the question less about "who answers" and more about "who owns the knowledge the answers are built on". According to a Gartner survey of 321 customer service leaders in autumn 2025, 91 percent feel pressure from management to implement AI, and more than 80 percent expect to reduce agent headcount within 18 months, mainly through attrition and hiring freezes. At the same time Gartner predicts that more than half of customer service organisations will double their technology spend by 2028, without a corresponding reduction in staff.
For the choice between in-house and outsourcing that means two things. The first is that the simple, recurring tickets, the ones that have traditionally made outsourcing profitable, are the same tickets that self-service and AI answers take first. The second is that AI only works on source material: Deloitte reports that 83 percent of companies use AI in their outsourced services, but only a quarter see lower cost or higher quality yet. Whoever lacks a reviewed knowledge base gets neither good AI nor good outsourcing, since both answer from the same source. How AI drafts reviewed by a person work in practice, and why that is the right start, is covered in the article on AI drafts in customer service.
What does GDPR require when a provider handles your customers?
A data processing agreement, documented instructions and control over sub-processors. A provider answering your customers' tickets processes personal data on your behalf and is therefore a processor, while you remain the controller. According to the Swedish Authority for Privacy Protection the processing must be regulated in a binding agreement that among other things states that the processor may only process the data according to your documented instructions, and the processor may not engage sub-processors without your written authorisation.
In practice that means three questions to ask before the contract is signed: where the agents sit and where ticket data is stored, which systems the provider uses and which sub-processors sit behind them, and what happens to the data when the contract ends. Simplest is when the provider works in your system, so that the tickets, the knowledge and the history stay with you and access can be closed the day the collaboration ends; that is how we built permissions and data storage in Supportifier.
When does in-house, outsourcing or hybrid fit?
Hybrid fits most companies with a few hundred tickets a week: you own the knowledge and the difficult tickets, a partner takes peaks, evenings or languages. This is how the trade-off looks:
| Situation | In-house | Outsourced | Hybrid |
|---|---|---|---|
| Steady volume, office hours | Fits well | Rarely justified | Unnecessary |
| Strong seasonal peaks | Hard to staff | Fits for the peaks | Fits: own base, partner for peaks |
| Evenings, weekends, several languages | Expensive at low volume | Fits well | Fits: partner takes hours and languages |
| Complex tickets, many exceptions | Fits well | Risk of wrong decisions | Fits: exceptions handed over to you |
| Customer insight important for product and site | Fits well | Lost without a reporting clause | Fits if insight is reported weekly |
| Small organisation without a head of customer service | Hard to build | Fits, but someone must own the relationship | Still requires an owner at your end |
The last row is the most underestimated. Outsourcing does not remove the need for someone who owns customer service; the job changes from answering to steering, which is the same shift we describe in the article on the knowledge manager role.
What to do
- Count volume per hour and weekday for the past three months, per channel. That is the input for both staffing and the quote.
- Sort the ticket types into three piles: lookup without judgement, standard answer with some judgement, and exceptions. Only the first two can be handed out.
- Write the knowledge base before you ask anyone. One reviewed answer per ticket type in the first two piles.
- Decide the metrics in advance: response time, first-contact resolution, customer satisfaction and share of handed-over tickets, defined as in the article on metrics. Write them into the contract with weekly reporting.
- Require in the contract that the knowledge is yours, that the provider works in your system or exports all history, that recurring questions are reported, and that the data processing agreement covers sub-processors.
- Pilot a limited part, such as evenings and weekends in one channel, for three months. Compare the metrics with your own base before expanding.
Common questions
Is outsourcing customer service cheaper?
Not automatically. The hourly rate may be lower, but you also pay for writing and maintaining the answers, training at staff changes, checking quality and owning the relationship. According to Deloitte's 2024 survey only 34 percent of companies name cost as the main reason, against 70 percent in 2020. Calculate the total cost per resolved ticket, including your own time, and compare with what self-service and AI answers remove from the volume.
Can we outsource only evenings and weekends?
Yes, and it is often the best first step. The partner takes a limited time slot in one or two channels, answers from your knowledge base and hands exceptions over to you the next morning. The requirement is that the partner works in the same system and with the same answers as you, otherwise the customer gets different answers depending on the time of day.
Who owns the knowledge base when customer service is outsourced?
You, always. The provider executes, you decide what the answers should be. Write into the contract that the knowledge base sits in a system you own or control, that the provider proposes changes but the subject owner at your end approves, and that all history comes with you if the collaboration ends. A provider that wants to keep the knowledge in its own templates makes you dependent on them.
How do we know whether the provider is doing a good job?
By measuring the same things you would measure on yourselves: response time, first-contact resolution, customer satisfaction per ticket and share of tickets handed over. Add your own quality review of a sample of tickets every week, read by someone who knows the subject. The metrics should be in the contract from the start, with weekly reporting, so a deviation is visible before it has become customer churn.
Sources
- Global Outsourcing Survey 2024 — Deloitte, 2024
- Outsourcing: så gör du som uppdragsgivare — Kontakta, read 2026
- Gartner Survey Finds 91% of Customer Service Leaders Under Pressure to Implement AI in 2026 — Gartner, 2026
- Gartner Predicts Over 50% of Customer Service Organizations Will Double Their Technology Spend By 2028 — Gartner, 2026
- Personuppgiftsbiträdesavtal — Swedish Authority for Privacy Protection, read 2026
Rickard Collander