Skip to main content
TurnellaBeta
WFM guideForecasting

Contact centre event forecasting

A known event that the WFM team does not know about becomes an unplanned volume spike — handled at emergency staffing cost. A known event the WFM team does know about becomes a scheduled adjustment — handled at normal cost with appropriate lead time. The difference is cross-functional information flow.

Event types and their contact impact patterns

Event typeTypical contact upliftPeak timingDurationAHT impact
Email campaign (B2C)1–5% of email audience contacts within 48h2–4 hours after send for 9am dispatch80% of contacts within 24h; trailing edge to 72hNeutral to low — typically simple transactional queries
Direct mail / letter0.5–2% of recipients contact2–5 days after expected delivery dateSpread over 7–14 days (delivery timing variability)Higher than email — customers read carefully and have more detailed questions
TV advertisingHighly variable — 0.1–1% of audience contacts in 24hWithin 30–90 minutes of ad broadcastShort — 60–70% of contacts within 6 hoursLow — typically new customer enquiries using a short script
Product launchDepends on customer base size — 3–15% contact in first weekLaunch day and the following 48hSustained over 2–4 weeks as customers discover the productHigh — new product contacts require explanation, AHT typically 20–40% above baseline
Price change / billing update1–8% of affected customers contactWhen the statement / bill arrives (2–4 weeks after change notified)Clustered around statement date — 5–10 day windowHigh — customers who call about price changes are more likely to escalate
Bank holiday (return)20–50% volume increase on the first working day after the holidayAll day — sustained above-normal volume1–2 days to clear accumulated demand from closed period+10–15% as post-holiday contacts include accumulated frustration contacts
Regulatory announcement0.5–5% of regulated customers contact — depends on public prominence of announcementWithin 24h of media coverage of the announcement3–7 days if sustained media coverageHigh — regulatory contacts are complex and often involve vulnerable customers

The event register: capturing event intelligence

An event register is a shared calendar or database where all known events that will affect contact centre volume are recorded before they occur. The WFM team uses it to overlay event adjustments on the baseline forecast. Without an event register, known events are either missed or discovered only when the surge arrives.

Event register fields — minimum viable set

Event name

Identifies the event for future reference and post-event analysis

Event date and time

When the event occurs — required for forecast adjustment timing

Affected customer segment

Determines the relevant audience size for contact rate calculation

Contact type expected

Determines which queues are affected and whether AHT will be different from baseline

Estimated volume uplift

The forecast adjustment — requires historical data for similar events

Confidence level

High/Medium/Low — drives how much contingency buffer to add

Source team

Who provided the event notification — for feedback and accountability

Lead time received

Days' notice the WFM team received — used to hold teams to notification SLAs

The notification SLA:The event register is only as useful as the notification lead time. Establish and publish a notification SLA: campaigns affecting >50,000 customers require 10 business days notice to WFM; campaigns affecting >200,000 require 15 business days. Make the notification process low-friction — a short form or a Teams channel notification is easier to comply with than a formal process. Track compliance by team and escalate when teams consistently notify late.

Campaign contact rate calculation

A contact rate is the percentage of a campaign audience that contacts the contact centre within the contact window. It is the primary input to the event volume uplift calculation.

Contact rate calculation — worked example

Event: Email campaign to 250,000 customers, send at 10am Tuesday

Audience size
250,000
Historic email contact rate for similar campaigns
2.3%
Adjustment: time-sensitive offer (+0.5pp)
+0.5%
Adjusted contact rate
2.8%
Expected total contacts within 48h
7,000
Peak hour (12:00–13:00, 2h post-send)
~28% of 24h volume = ~350 contacts/hr
Estimated AHT for these contacts
4.5 minutes (vs. 5.5 baseline)
Additional agent requirement at peak
~7 agents above baseline schedule

Distribution of 7,000 campaign contacts across 48h

Hour 1–4 (10am–2pm Tue)45%3,150
Hour 5–8 (2pm–6pm Tue)20%1,400
Hour 9–24 (6pm–10am Wed)15%1,050
Hour 25–48 (Wed working day)15%1,050
Post-48h trailing contacts5%350

Peak hour is typically 2–3 hours post-send for 10am dispatches due to email open lag. Distribution shifts for earlier sends (6–8am) which peak nearer the morning contact surge.

Two common event forecasting failures

Failure 1: No cross-functional notification process

The WFM team does not receive event notifications from marketing, product, billing, or communications teams until the event causes a volume spike. The contact surge is managed as an emergency rather than a planned event.

Fix: Establish a notification SLA with each team that generates contact-driving events. Make notification easy (a Teams channel, a form). Track notification compliance and escalate to operations director when teams consistently miss the SLA.

Failure 2: Replacing the baseline with the event estimate

The event volume forecast replaces the baseline forecast for the period, rather than being added to it. If the baseline forecast is 1,000 contacts on Tuesday and the campaign adds 500, the correct forecast is 1,500 — not 500. Over-constraining the event adjustment produces under-staffing on top of the event uplift.

Fix: Event forecasts are overlays — additive to the baseline. Build the event adjustment on top of the baseline for each interval, not as a substitution. Apply separately and sum before the final staffing calculation.

Post-event review: building a better model

Every planned event should be reviewed after it occurs. The comparison between actual event contacts and the forecast event contacts provides the data to calibrate the contact rate model for future events of the same type.

1.

Compare actual vs. forecast contacts by hour for the event period

Was the total volume close to forecast? Where did the timing distribution diverge? Was there an unexpected tail?

2.

Calculate the actual contact rate

Divide actual event contacts by audience size. Compare to the contact rate assumption. Update the historic contact rate database for this event type.

3.

Review AHT actuals vs. assumption

Did event contacts take longer or shorter than expected? Was there a different contact mix than predicted?

4.

Log the notification lead time received

How many days notice did WFM receive? Was this within the notification SLA? If not, escalate.

5.

Update the contact rate table

Each post-event review adds a data point to the historic contact rate table for each event type. Over time, this table becomes the primary input to event volume estimates — reducing reliance on assumptions and improving forecast accuracy.

Event forecasting questions

How do you forecast contact volume for a marketing campaign?

Five steps: (1) Obtain campaign details: channel, audience size, send date and time, call-to-action; (2) Apply historical contact rate for similar campaigns: email typically 1–5% within 48h; adjust for time pressure, audience quality, campaign prominence; (3) Distribute contacts across time using the campaign contact curve — email typically peaks 2–4 hours post-send; (4) Add to the baseline forecast — event contacts are additive, not substitutes for the baseline; (5) Adjust AHT assumption if the contact type differs from baseline — complex product contacts run 20–40% higher AHT than simple transactional queries.

Related guides