Inbound
Inbound is customer contact the customer initiates: a chat message, a phone call, or an email the customer sends because they have a question or a problem, as opposed to contact the business starts.
It is the traditional shape of customer service: a customer has an issue, they reach out, and the business responds. Inbound contact volume is typically driven by triggers the business does not control directly, such as product issues, seasonal patterns, or external events, which makes it harder to plan staffing or capacity around than outbound contact, which the business initiates on its own schedule.
How does inbound support differ from outbound contact?
Inbound support reacts to a customer's initiative: the business only learns about a problem once the customer has already noticed it and decided to reach out. Outbound contact, including proactive outreach and agent-initiated outreach, starts from the business side, flagging something to a customer before they ask. The two usually run through the same team and the same AI agent in practice, but they carry different pressures: inbound volume is largely reactive and variable, while outbound volume is planned and, in principle, controllable.
Where AI agents fit into inbound customer service today
Inbound is where most AI agent deployments in customer operations start, because the volume is highest and the entry point, a chat widget or phone line, is well understood. An AI agent handling inbound needs to triage accurately, recognising what the customer actually needs, which is often different from their first message, resolve what it can end to end, and escalate the rest cleanly. As trust in an AI agent's inbound performance grows, firms typically extend it from simple queries into more complex case types, following the same gradual expansion pattern that applies to autonomous agents generally. Inbound rarely stays separate for long, which is the case for treating support, back-office, and outbound as one workload.
Measuring inbound support: volume, resolution, and cost per contact
Inbound is usually measured on volume, how many contacts arrive, by channel and reason, resolution rate, how many are actually solved, not just answered, and cost per contact, the total cost of handling divided by contact volume. These three interact: an AI agent that resolves more of the inbound volume directly reduces cost per contact, but only if resolution holds up under scrutiny, since a repeat-contact spike on the same issue erodes the apparent savings once it shows up in the numbers a few weeks later.
Frequently asked questions
Is inbound the same as reactive support?
Yes, they describe the same pattern: the business responds after the customer initiates contact. Inbound is the more common term for the channel and volume side of this; reactive is more common when contrasting with proactive strategy.
Does every business get more inbound than outbound contact?
Historically yes, because most customer service was built around responding to contact rather than initiating it. As proactive outreach and agent-initiated outreach mature, the balance shifts in some case types, such as collections or onboarding, where reaching out first reduces the inbound volume that would otherwise follow.
Can an AI agent handle inbound contact across every channel?
Coverage depends on the agent's build, not a fixed limit. Some AI agents handle chat only; others extend to voice, email, and SMS. The channels a buyer needs should be confirmed against a vendor's actual coverage rather than assumed.
Why is inbound volume harder to plan for than outbound?
Inbound is triggered by the customer, often in response to something outside the business's direct control, a product issue, a market event, a billing cycle, so volume can spike unpredictably. Outbound volume is planned by the business itself, which makes staffing and capacity easier to forecast around it.