Outbound campaign planning

Forecasting · Scheduling · ~7 minute read

A different rhythm from inbound

Almost every guide to contact centre planning is implicitly about inbound. Customers arrive, queue, get answered; the planner forecasts arrival rates and schedules to cover them. Outbound is structurally different. The contacts the operation makes aren’t a function of customer behaviour; they’re a function of how many agents are working, what list they’re working from, what dialler mode they’re using, and the regulatory and ethical constraints on when they can call. The planning craft has the same building blocks — demand, supply, capacity, productivity — but the way they fit together is different enough that an inbound planner moving into outbound for the first time needs to recalibrate. This article walks through what outbound campaign planning involves, the dialler modes that shape it, list management, contact-rate forecasting, and the regulatory constraints that increasingly define the operation.

The three dialler modes

Three dialling modes dominate outbound work. Predictive dialling places more calls than there are available agents, predicting how many will connect and routing connected calls to whichever agent is free first. It maximises agent talk time but produces some abandonment of connected calls when no agent is available, which is heavily regulated in most jurisdictions. Progressive dialling places one call per available agent: the dialler waits for an agent to be ready, then dials the next number. No abandonment, lower agent talk time, simpler to plan. Preview dialling shows the agent the customer record before placing the call, gives them a moment to prepare, then places the call. The lowest agent productivity of the three, used where the conversation needs context (debt collection, retention, complex sales).

Each mode produces a different relationship between agent-hours and customer contacts. A planner moving between modes needs to recalibrate the productivity assumption: a predictive operation might produce 25–35 connected calls per agent-hour; a progressive operation 12–20; a preview operation 6–12. Conflating them produces planning that misses by a factor of two or three.

List management as a planning input

The list is to outbound what volume is to inbound. A 100,000-record list with a 30% connect rate and a 15% conversation rate produces a very different staffing requirement from a 50,000-record list with a 50% connect rate and a 25% conversation rate, even if both are described as “a campaign.” The planner’s job is to convert list size and quality into expected agent-hours, and this means understanding the metrics that describe list quality: contact rate (the proportion of numbers that someone answers), right-party-contact rate (of those answers, the proportion that reach the intended person), conversation rate (the proportion of right-party contacts that produce a meaningful conversation), and conversion rate (the proportion of conversations that produce the desired outcome). Each of these is forecastable from prior campaigns; together they determine how many agent-hours the list will absorb.

List quality decays as the list is worked. The first pass produces the best contact rates; subsequent passes (retries on no-answer, re-attempts on different days and times) produce lower rates each time. A planner who treats a list as a uniform pool will under-staff later passes and miss the campaign deadline; a planner who models the decay produces a realistic staffing curve.

Talk-time distributions and AHT

Outbound AHT distributions are usually more skewed than inbound. A typical outbound call is short — a sale pitch, a confirmation, a polite refusal — but a meaningful tail of calls runs long (objections, complex closes, accidental friend-of-the-business chats). Modelling the average AHT alone misses the shape of the distribution, which matters for capacity planning. The 90th-percentile call length usually drives the staffing more than the median, because the long calls produce the agent unavailability that constrains the dialler.

Forecasting outbound

Outbound forecasting works backwards from campaign objectives. The campaign needs to reach X customers, convert Y of them, by date Z. Working backwards: Y conversions divided by conversion rate gives required conversations; required conversations divided by right-party-contact rate gives required connects; required connects divided by contact rate gives required dial attempts; required dial attempts divided by attempts-per-agent-hour gives required agent-hours; required agent-hours divided by working hours and shrinkage gives required FTE.

Each conversion factor is a forecast, and each is informed by prior campaign data. A new campaign with no prior data needs the same analogue-and-recalibrate approach as a new inbound operation (see forecasting a new operation).

Regulatory constraints

Outbound is more heavily regulated than inbound, and the constraints materially shape planning. In the UK, Ofcom rules limit the abandonment rate of predictive dialling to 3% over any 24-hour period, with abandoned calls requiring a recorded informative message. Telephone Preference Service (TPS) registration means certain numbers cannot be called for marketing purposes; failure to scrub the list against TPS attracts ICO fines. Call hours are conventionally restricted to working hours (in practice, 9am to 9pm in the UK, with some operations narrower); calling outside this window produces complaints and risks regulatory action. GDPR and the UK GDPR impose lawful-basis requirements on outbound marketing: consent or legitimate interest must be documented, and right-to-object must be honoured promptly.

The planner’s role is to bake these constraints into the operational plan rather than discover them as problems mid-campaign. The list-scrub, the consent-record, the call-window discipline, and the abandonment monitoring all need to be planning-ready before the first dial.

Capacity planning for outbound campaigns

An outbound campaign typically has a finite life: hit the target by the deadline, then close. This is different from inbound steady-state planning. The capacity profile usually peaks in the middle and tapers off as the list is exhausted, with a deliberate “runway” in the final days to chase late records. Operations that staff outbound campaigns the way they staff inbound queues miss this shape and either under-staff the peak or over-staff the tail.

The agent experience

Outbound work is emotionally different from inbound. Agents face refusals, hostility, and rejection that inbound agents rarely encounter. Talk time per agent is often lower; non-productive time (waiting for dials, handling no-answers, navigating the dialler) can be high. Attrition in outbound operations tends to run higher than in equivalent inbound operations, and the planner’s capacity model needs to reflect this. Building in genuine recovery time, varied work where possible, and a clear progression path are part of the operational plan, not optional add-ons.

Common mistakes

Three patterns recur. Applying inbound assumptions to outbound. The Erlang calculation is the wrong model for outbound; the right model is closer to a manufacturing throughput calculation. Ignoring list decay. Treating the second-pass as productive as the first-pass produces capacity plans that miss the campaign deadline. Underestimating regulatory risk. An outbound operation that breaches Ofcom abandonment rules or ICO TPS rules faces consequences much larger than any inbound SL miss.

Conclusion

Outbound campaign planning shares the building blocks of inbound planning but assembles them differently. The dialler mode shapes the productivity assumption, the list quality drives the staffing requirement, the regulatory framework defines what is operationally possible, and the campaign-finite lifecycle produces a capacity profile unlike steady-state inbound work. Planners moving between the two need to recalibrate; operations that run both should keep the planning functions either deliberately integrated or deliberately separate, but never accidentally blended.

Pair this with forecasting a new operation for the analogue-and-recalibrate pattern that suits new outbound campaigns, and planning with an outsourcer alongside if outbound work is being contracted out.