Email customer service outsourcing: the channel nobody staffs properly
Paul Szerszen, SVP of Professional Services & Customer Operations
In most contact centers, email is what agents do when the phones are quiet. It has no dedicated forecast, no interval-level staffing, often no separate quality review, and a response time target that was set once and is quietly missed.
This works until volume grows. Then the backlog becomes visible, someone assigns a group of agents to clear it, the queue empties, and the same pattern rebuilds over the next quarter. The underlying problem is that email was never planned as a channel. If you are considering email customer service outsourcing, that planning gap is the thing to fix in the contract rather than inherit.
Email is not filler work
The logic behind treating email as overflow is that it is asynchronous, so it can wait. That is true for minutes and false for the rest.
Async work fills the gaps in a schedule only if the gaps are real and predictable. In practice, the quiet periods when agents are supposed to work email are also the periods most likely to be reassigned when something happens elsewhere. Email is the first thing sacrificed and the last thing measured, so it absorbs every other problem in the operation.
The fix is unexciting: treat email as a queue with its own forecast, its own staffing requirement, and its own service target, in the same way a genuine omnichannel customer service model treats every channel. Agents can still be cross-trained across channels. The capacity has to be planned rather than assumed.
Forecast email separately, because it arrives differently
Email volume has an arrival pattern, and it does not look like your phone pattern.
Email arrives more heavily outside business hours, because customers send it when it suits them. It spikes the morning after an incident, a billing run, or a marketing send. And it has a different weekly shape, often with a Monday peak reflecting weekend accumulation.
Build the curve. Pull email arrivals by interval and day of week and compare it to your phone curve. The two will diverge, and where they diverge is where blended staffing has been failing silently.
There is a second input that phone forecasting does not have: the backlog carried into each day. Email staffing has to cover today's arrivals plus whatever is aging from yesterday, and a plan that only covers arrivals will never catch up.
Set a response target you can actually hold, and measure the right thing
Most email targets are set as a first-response time and left there. Two refinements make the target useful.
Measure time to resolution alongside time to first response. A fast acknowledgment followed by four days of silence is worse for the customer than a considered reply in one day, and a first-response metric rewards the former. Mpathic's published work on a statewide education benefits program reported an eight-hour email response standard alongside a 12-second average speed to answer on voice (State agency contact center case study). Commit to a number you can hold on your worst day, not your average one.
Then measure the exchange count. The number of back-and-forth messages required to resolve one issue is the email equivalent of first call resolution, and it is the single most useful email metric most teams do not track. A rising exchange count means replies are incomplete, and incomplete replies multiply your volume.
Written quality is a different skill and needs its own scorecard
A quality scorecard written for voice does not transfer to email. Tone, pacing, and active listening either do not apply or apply differently, and the things that make a written reply good are not on a voice scorecard at all.
Score email on its own terms. Did the reply answer every question the customer asked, including the second one buried in the third paragraph? Is it comprehensible to someone who does not work at your company? Does it anticipate the obvious follow-up and answer it in advance, which is what reduces exchange count? Is it structured so the answer is visible without reading the whole message? And is it accurate, which matters more in writing because it is a record the customer keeps.
This is also a hiring signal. Strong voice agents are not automatically strong writers, and the reverse holds too. If you are outsourcing email, ask how the partner assesses writing during hiring and how written contacts are reviewed. If the answer is that the same scorecard covers both channels, email quality is not being managed.
Templates help until they replace thinking
Templates and saved replies are necessary at volume and become a problem in a specific way: they answer the question the template was written for rather than the question the customer asked.
Two rules keep them useful. Templates are starting points that must be edited to the specific contact, never sent unmodified, and quality scoring should catch unmodified sends. And templates need an owner and a review cycle, because a template with outdated information gets sent hundreds of times before anyone notices, which is the difference between a bad call and a bad template. The same ownership question applies to the knowledge base agents draw from.
Track which templates are used most and review those first. The highest-volume template is the single piece of writing in your operation that carries the most weight.
What to require from a partner
Five requirements make email outsourcing work. Email forecast and staffing reported at the interval level, separately from voice. Both first-response time and time to resolution, with the target set for the worst day. Exchange count per resolved issue, tracked and trended. A written-contact quality scorecard distinct from the voice scorecard, with sampling rates stated. And a named owner for template and knowledge content, with a review cadence.
ISO 18295-1:2017 sets out requirements for customer contact centers across the channels they operate rather than for voice alone (International Organization for Standardization). If a provider's reporting only describes voice, ask what happens to the rest.
Email customer service outsourcing works when email is treated as a channel with its own forecast, staffing, targets, and quality standard. It fails when it is defined as what agents do between calls.
Pull one report this week: email exchange count per resolved issue for the last quarter. If it is rising, your replies are incomplete and your volume is being generated internally. That is the most fixable problem in the channel.
Frequently asked questions
What is email customer service outsourcing?+
Email customer service outsourcing means engaging an external team to handle inbound customer email and other asynchronous contacts. Done properly it includes a separate email forecast, dedicated staffing by interval, response and resolution targets, and a quality scorecard built for written contacts rather than borrowed from voice.
How fast should you respond to customer service emails?+
Set a target you can hold on your busiest day rather than on an average one, and measure time to resolution alongside time to first response. A quick acknowledgment followed by days of silence scores well on first response and serves the customer badly, which is why the second measure matters.
How do you forecast email volume?+
Build an arrival curve by interval and day of week, as you would for voice, then add the backlog carried in from previous days. Email arrival patterns differ from phone patterns, typically with heavier out-of-hours submission and a Monday peak, so blended staffing built on the phone curve tends to under-cover email.
How do you measure email support quality?+
Use a scorecard built for written contacts: whether every question asked was answered, whether the reply is comprehensible to someone outside your company, whether it anticipates the obvious follow-up, whether it is structured so the answer is easy to find, and whether it is accurate. Voice scorecards do not transfer.
What is a good email support metric to track?+
Exchange count per resolved issue, meaning how many back-and-forth messages it takes to close one customer question. It is the email equivalent of first call resolution and it is rarely tracked. A rising exchange count indicates incomplete replies, which generate additional inbound volume from your own responses.
