Last updated: 13 November 2026 • By the Global Remote Services marketing team

1. “We will answer faster” becomes “we will resolve more”

Speed is the resolution everyone writes down, because it is the one that is easy to measure. It is also the one easiest to improve without improving anything: a faster first reply that does not answer the question simply moves the work to the second contact, where it costs more.

Rewrite it as resolution. How many contacts ended without the customer coming back, and how much effort did the whole thing cost them. Keep speed as a secondary measure, because a customer waiting three days still has a bad experience even if the answer is eventually right. But lead with the question of whether the problem went away.

2. “We will automate customer service” becomes “we will sort the contacts first”

Automation programmes that get cut back usually went wrong before any software was chosen. They were aimed at a channel, as in “automate chat”, rather than at a set of contact reasons. Chat contains both the password reset and the furious complaint about a missing order, and the same automation cannot hold both.

The useful January exercise costs nothing: export the last three months of contact reasons, sort them into three piles, and write a sentence next to each explaining the choice. Automate the repetitive, low-risk requests that can be answered from a single system. Assist the ones where an agent is faster with a suggested answer in front of them. Keep the rest with a person: money, health, complaints, anything where two systems disagree and somebody has to decide which is right.

3. “We will cut cost per contact” becomes “we will count the whole cost”

Cost per contact is a clean number that leaves most of the cost outside it. Onboarding time at the start of a programme, the hours your own managers spend steering it, the cost of replacing people who leave, and the rework created by contacts that were closed but not resolved: none of that appears in the rate, and all of it appears in the budget.

The better resolution is to compare providers and in-house options on the same categories: what it costs to produce the outcome, what the integration between systems costs, what proving data security costs, what training costs, and how much room the contract leaves to move volumes without reopening it. A rate that looks cheap in January is often the one that produces the awkward conversation in June.

4. “We will expand into new markets” becomes “we will staff the languages”

Entering a market is usually planned as a marketing decision and discovered as a staffing one. The campaign runs, the orders arrive, and then the questions arrive in a language nobody on the team speaks. Machine translation has narrowed the gap for short written exchanges and has not closed it for a complaint, a refund argument, or a phone call where someone switches language halfway through a sentence.

So plan capacity per language and not only per channel, and decide before launch which languages get a native speaker, which get assisted translation, and which you are honestly not ready to serve yet. Saying “English only for now” on the site is a better customer experience than answering slowly in a language you cannot support.

5. “We will use more AI” becomes “we will use AI where it is allowed and useful”

There is a legal floor here that is easy to miss when the resolution is written as ambition. Under the EU AI Act, people must be told clearly that they are interacting with an AI system, at or before the first exchange, unless it is obvious from the context, and anyone exposed to emotion recognition must be informed as well. Those transparency duties have applied since 2 August 2026. A deferral agreed in 2026 moved some high-risk obligations later, to December 2027 and August 2028, but left the transparency rules untouched.

In practice that means the disclosure wording has to exist in every language you serve, and sound natural in each, which is the part most often forgotten. Treat this as a planning summary rather than legal advice and confirm your own position with your legal team. Above the floor, the most reliable gains are quieter than full call handling: automatic summaries, call tagging and note-taking give time back on every single contact rather than on a subset.

6. “We will do better this year” becomes “we will write down the January baseline”

This is the resolution that makes the other five checkable. A target without a starting number cannot be assessed, so in March the conversation becomes a matter of opinion and the programme quietly loses its budget.

Pick three numbers in the first week of January, record them somewhere both sides can see, and put a date in the calendar to look at them again. Resolution rate, repeat contact rate and customer effort are enough. If part of your volume is automated, record the automated and the human path separately as well as together, otherwise one hides the other and the average tells you nothing.

What to do in the first two weeks of January

• Export the last three months of contact reasons and sort them into automate, assist and keep human, with a reason written beside each.
• Record three numbers as your baseline: resolution rate, repeat contact rate and customer effort.
• Test your own escalation path the way a customer would, on every channel and outside business hours.
• Check that every automated touchpoint says clearly that it is automated, in each language you serve.
• List the languages you support, and mark the ones currently covered by goodwill rather than by staffing.
• Put the March review in the calendar now, with the three numbers attached to the invitation.

None of this requires a budget decision. All of it makes the budget decision easier when it comes.

How Global Remote Services approaches this

Global Remote Services (GRS) has operated as an independent contact centre in Bucharest since January 2004. Around 500 people work across two sites in Romania, delivering multilingual customer experience services in more than 20 European and Asian languages. GRS is part of the Be Ys group.

Our approach is humans empowered, AI amplified. Automation takes the repetitive requests and helps agents answer faster, while trained people keep the conversations that need judgement and empathy. When a company brings us a January plan, we start from its contact mix and its languages rather than from a tool, agree what is automated and what is not, and fix the measures before anything goes live, so the March review has something real to look at.

Frequently asked questions

What is a realistic customer service resolution for 2027?

One that names a number and a date. “Raise the resolution rate and review it in March against the January figure” is a resolution. “Improve customer service” is a sentiment. The second cannot be checked, so it cannot be defended when budgets are reviewed.

Should we automate customer service in January?

Sort the contacts in January and automate after that. January is a poor month to launch anything into live traffic anyway, because post-season volume is still unwinding and the baseline is distorted. Use the month to decide, and start changing things when the numbers are stable again.

How do we know by March whether a resolution is working?

Compare the same three numbers you recorded in week one, and read them together rather than separately. Handle time rising while repeat contacts fall is usually a good sign: the easy contacts have gone to automation and the ones left are harder by design.

Do we have to tell customers they are talking to an AI?

In the EU, yes. Since 2 August 2026 people must be told clearly, at or before the first interaction, that they are interacting with an AI system, unless that is obvious from the context, and anyone exposed to emotion recognition must be informed too. Confirm the detail for your own setup with your legal team.

Is outsourcing customer support cheaper than hiring in-house?

It depends on what you count, which is the point of resolution three. Outsourcing tends to win on languages you cannot staff yourself, on cover outside business hours and on absorbing seasonal peaks without hiring for the peak all year. It is rarely the cheaper option for a small, stable, single-language queue.

When should we start if we want a partner running before the next peak?

Work backwards from the peak and leave room for recruitment, product training and a pilot period before volume arrives. Starting the conversation in the first quarter for a programme that has to be stable by autumn is comfortable. Starting it in September is not.

Note: the regulatory dates are given for planning and are not legal advice. The practical recommendations are those of Global Remote Services.