Support Reporting
Part of Support workforce tools
Estimating queue demand from historical contacts
Define the arrival unit, build comparable historical intervals, adjust for known events and check a support queue forecast against actual demand.
Estimate queue demand by counting comparable past arrivals in consistent intervals, then record adjustments for known changes. Define an arrival before calculating a baseline; otherwise, transfers, reopened work and unique customer issues may be mixed into one precise-looking total.
Choose the unit and interval
Start with the queue, channel and service window. A live phone or chat queue may need short intervals because people must be available near arrival.
Email can also be viewed across its response window alongside the backlog. Record the reporting time zone and how daylight-saving changes are handled when comparing weeks.
Decide whether the count represents unique issues, contacts entering a queue or work items needing agent time. Keep transfers, callbacks, reopened cases, spam and bot-resolved contacts visible until you know how the source system records them.
Zendesk WFM counts each entry into a workstream, including moves and reopenings; its history may include spam. This should not silently become a count of unique customers.
Build comparable history
Use the same unit and interval across several comparable weeks. Keep the original arrival timestamp, channel, queue, event type and any exclusion marker.
Group like days and intervals together so a busy Monday morning is compared with similar Monday mornings. Also inspect whole-week totals for a broader trend.
Annotate incidents, campaigns, public holidays, channel closures and routing changes. Do not delete a spike merely because it is inconvenient.
Decide whether its cause could recur. Keep a separate event scenario when it could; when it could not, record the exclusion and retain the original count.
Zendesk WFM offers historical-period selection, volume adjustments and outlier exclusions. Those controls do not decide whether an exclusion is justified.
Key Events Impacting Contact Volume Over Past 3 Weeks
- 10September 2026 – National Public Holiday (Labour Day) — Reduced contact volume; excluded from baseline
- 15September 2026 – Channel Routing Change (Chat to Email) — Increased email arrivals; documented as adjustment
- 22September 2026 – Promotional Campaign Launch — Spiked arrivals; modelled as recurring event
Make the estimate traceable
A simple baseline can use the average or median of comparable intervals. Record the number of observations and inspect their spread.
Adjust for a named reason, such as a planned launch or a documented channel shift. Keep the unadjusted baseline beside the revised estimate so another planner can see what changed.
Suppose four hypothetical Monday mornings have 32, 35, 34 and 47 email arrivals. Investigate the higher observation.
If it came from a recurring promotion, model that event; if it came from a one-off fault, document the reason before choosing a baseline. These numbers illustrate the decision and are not measured demand.
Compare forecast with arrivals
Compare the forecast and actual arrivals using the same unit and interval. Record the size and direction of misses.
A weekly total can appear accurate while repeatedly missing a morning peak. Investigate changed volume, routing, duplicate counting and new patterns.
For asynchronous work, review opening backlog separately. Today's arrivals do not describe everything awaiting a response.
Pass the dated demand estimate, uncertainty, queue definition and event assumptions to the capacity planner. Handling effort, response targets, shrinkage and skills determine coverage after the demand estimate is made.
