Somewhere in the last few years, a large share of British consumers stopped treating messaging as a novelty and started treating it as the default way to contact a business. Not email, which they associate with waiting. Not the telephone, which they associate with queueing. Messaging — asynchronous, interruptible, and conducted in the same app they use for everything else.
Most UK contact operations have not restructured for this. They have added the channel and staffed it with whoever was spare.
Why messaging breaks conventional workforce planning
A telephone call is a bounded event. It starts, it occupies one agent exclusively, it ends. Planning around it is a solved problem.
A messaging conversation is not bounded. It can run for eleven minutes or three days. The customer replies when convenient, which may be during the meeting they left to send the first message. An agent can hold several concurrently, but not an unlimited number, and the point at which quality degrades varies by complexity rather than by count.
Accent is irrelevant. Written English can be quality-assured directly and improved systematically in a way conversational handling cannot. The asynchronous nature means a time-zone offset is an advantage rather than a constraint — a message sent at eleven at night in Britain can be answered properly rather than queued until morning. That combination is why messaging work has moved quickly to call centres in the Philippines, including in operations that kept their voice channel at home.
This defeats the standard planning model in three ways. Occupancy becomes ambiguous — is an agent with four dormant conversations busy? Handle time becomes meaningless as a productivity measure. And service level needs redefining entirely, because “answered within twenty seconds” describes nothing a messaging customer experiences.
“Teams try to run messaging on telephone maths and it falls apart within a month,” said Ralf Ellspermann, founder and chief strategy officer at PITON-Global. “Then they decide messaging is a poor channel, when the real fault was measuring an asynchronous conversation with a synchronous yardstick.”
What to measure instead
Three replacements do most of the work, and specialists who assess written-channel quality tend to ask for all three before they ask about headcount.
Time to first meaningful response, not time to acknowledgement. An automated “thanks, we’ll be with you shortly” is not a response and customers do not count it as one.
Conversation resolution time, measured from first message to resolution, with explicit exclusion of time waiting on the customer. Without that exclusion the figure measures the customer’s responsiveness rather than the operation’s.
Concurrency with a quality floor — the number of simultaneous conversations an agent can hold before quality measurably drops, established empirically for your own contact mix rather than taken from a vendor’s benchmark.
The channel mix a UK operation actually needs
“Nobody fails at messaging because of grammar,” said John Maczynski, chief executive of PITON-Global. “They fail because the reply is correct, complete and reads like a terms-and-conditions page. British customers notice instantly.”
Not every channel, and not uniformly resourced.
Telephone remains essential for complex, emotional and urgent contacts, and for the substantial minority of UK consumers who prefer it — a group skewed older and toward exactly the situations where getting it wrong costs most.
Messaging suits status queries, account changes, straightforward problems and anything the customer wants to handle while doing something else. It also suits customers who find telephone conversation difficult, which is a larger population than most operations allow for.
Email persists mainly where the customer wants a record. Live chat on the website has largely been absorbed into the same operational model as messaging.
Social media is the one most often mishandled. Public complaints on a brand’s own feed are not really a service channel; they are a reputational event with a service component, and routing them into the standard queue produces both slow responses and inappropriate ones.
Where offshore delivery fits
On balance, messaging travels offshore more easily than voice does.
What it demands is written register. British customers read tone in text with considerable sensitivity, and the gap between correct English and natural English is where offshore messaging operations most often fall short. “We apologise for the inconvenience caused” is impeccable and lands badly. “Sorry about that — let me sort it out” does not.
That is a training and quality-assurance problem rather than a capability one, and it is the single most useful thing to examine when evaluating any provider for messaging work: ask to see how written tone is coached, not whether it is.
Automation sits differently here
The automated layer in messaging is less visible to customers than in voice and therefore easier to over-extend.
In a telephone call the handover from system to person is audible. In a messaging thread it can be seamless, which is useful when the automation is working and damaging when it is not — a customer may spend several exchanges with a system before realising, and the realisation itself becomes a complaint.
Two practices manage this. Declare the transition explicitly, which costs a line of text and prevents the customer feeling deceived. And set the escalation threshold lower in messaging than in voice, because the cost of a wrong automated answer is higher when it arrives in writing and can be screenshotted.
Getting the operating model right
Four decisions, made deliberately rather than inherited.
Decide whether messaging agents also take calls. Blended working sounds efficient and produces poor outcomes in both channels, because the cognitive models differ and switching between them is expensive.
Set concurrency empirically and revisit it as the contact mix changes.
Build a written quality framework with its own criteria — tone, clarity, structure, appropriate brevity — rather than adapting the voice scorecard.
And define what happens to a conversation at the end of a shift. Handover in messaging is where context dies, and an operation without an explicit handover standard will hand customers a new agent and no memory.
The thing worth saying plainly
Messaging is not a cheaper telephone. It is a different service with different economics, different planning mathematics and different quality criteria, which happens to be where a growing share of British customers now prefer to start.
Operations that treat it as a bolt-on get a bolt-on’s results. The ones that restructured around it tend to find it becomes their highest-satisfaction channel within a year — which is not an argument that it is easy, only that it repays being taken seriously.
