Inbound Call Centre Services

An inbound queue is a maths problem before it is a people problem

Most inbound call centres are understaffed at specific hours rather than overall. The monthly average looks adequate, the Tuesday 6pm queue collapses, and adding headcount across the board fixes it expensively while leaving the pattern intact.

Getting inbound right means understanding four numbers and how they interact: how calls arrive, how long they take, how much of an agent's paid time is actually available, and what waiting time you are willing to accept. Everything else - hiring, rostering, cost - follows from those.

RHI runs dedicated inbound teams sized from your arrival data, not from a headcount you nominated.

Arrival patterns, not averages

Calls do not arrive evenly and they do not arrive predictably within an hour. They arrive in bursts, and a team sized for the hourly average will fail inside that hour while still hitting the average.

We forecast in short intervals rather than by day, using your historical arrival data, and we look for the patterns that matter: the daily peak, the weekly shape, the seasonal or event-driven spikes, and the recurring anomalies - the Monday after a bill run, the morning after a marketing email, the day a delivery promise expires.

If you do not have interval-level data, that is common and worth fixing early. It is the single most useful input to sizing a team, and its absence is usually why the current team feels wrong without anyone being able to say why.

Occupancy: the number that quietly decides quality

Occupancy is the proportion of an agent's available time spent on calls. It sounds like something to maximise. It is not.

Push occupancy too high and two things happen: quality falls because agents have no recovery time between difficult calls, and attrition rises because the job becomes relentless. Both are more expensive than the idle time you saved. Run it too low and you are paying for people to wait.

We plan to a sustainable band rather than a maximum, and we report actual occupancy alongside service level so you can see the trade-off rather than only its consequences. A supplier reporting excellent service level and unusually high occupancy is running a team that will not hold.

Shrinkage, stated honestly

Shrinkage is paid time an agent is not available to take calls: breaks, training, coaching, meetings, holiday, sickness, system downtime. It is real, substantial, and the most common reason a team that looks correctly sized on paper misses its targets.

We plan with shrinkage explicit rather than hidden in a rounded-up headcount, and we tell you what we have assumed. If a supplier's staffing model does not show a shrinkage assumption, ask for it - the number is either absent or embarrassing.

Service level, not average speed of answer

An average speed of answer conceals its own tail. Half the callers answered instantly and half waiting several minutes produces a respectable average and a bad experience for a lot of people.

Service level - the proportion answered inside a target time - is the more honest measure, and we report both alongside abandonment rate. Abandonment is the clearest signal that coverage does not match arrival, and it is worth looking at by time slot rather than in aggregate: abandonment concentrated in one window is a rostering problem, not a headcount problem, and it is much less expensive to fix. See the measures we report.

Routing and queue design

Calls arrive against a queue and a skill group. Menus stay shallow, because every extra layer loses callers and a caller who abandons a menu rings back less patient than before.

Skill-based routing improves resolution when the skill groups reflect real differences in competence, and damages it when they fragment the pool so far that each group is too small to absorb a burst. That trade-off is a design decision we make explicitly with you rather than inheriting from a platform default. Where volume genuinely exceeds capacity, an announced callback offer beats an unannounced queue - see callback handling.

How calls are handled

Routing, then identification against your verification rules, then context from every prior contact across channels, then resolution using your knowledge base with a plain statement of what happens next, then disposition against defined reason codes. Where a promise was made, a follow-up is scheduled and owned. Phone support covers the handling workflow in more detail.

Escalation

Agreed before launch: what the agent resolves, what goes to a senior agent, what reaches you, and what triggers an immediate alert regardless of hour. Vulnerability, safety issues, regulatory complaints and data protection requests never queue.

Languages, coverage, systems and quality

From India we staff Hindi, Hindi-English bilingual, English, Bengali, Nepali and Punjabi voice support, with further Indian languages on request. For international voice we staff neutral international English, recruited for comprehension and clarity rather than accent imitation; UK onshore and nearshore options are available - see the UK market page.

Coverage is built from shift units. Agents work in your telephony and CRM wherever possible - see technology. Calls are sampled and scored against an agreed scorecard with reviewers calibrated against each other, and average handling time is reported for capacity planning but never used as a target: pushed as a goal it produces rushed calls and repeat contact. See quality assurance.

Getting started, and what it costs

Priced per dedicated agent per month with a five-agent minimum - published rates and what moves them. Voice sits above written channels because handling time is longer and the recruitment bar is higher.

The most useful thing you can bring is interval-level arrival data, average handling time and abandonment by time slot. From those three we can size a team properly and tell you whether your real problem is headcount, rostering or routing. Talk to us.