How to Read the Inbound Forecasting Report in VICIdial

The Inbound Forecasting Report is a planning tool. Instead of telling you what already happened call by call, it looks at past inbound activity for the campaigns and ingroups you pick, runs the numbers through some queueing math, and then estimates how many agents you should put on the phones to keep your dropped calls under a target you set.

If you are still getting comfortable with the reporting menu, start with the VICIdial reports guide for the full map, then come back here for the forecasting piece.

What you feed it

You give the report three things: a date range, the campaigns or ingroups to include, and a desired drop call rate. The drop rate is the share of inbound callers you are willing to lose before an agent picks up. A common starting target is 3%. Pick the Ingroup or Campaign that actually receives the traffic you want to staff for, not your whole system, or the recommendation will be off.

The report buckets everything by hour, so it works best over a span of similar days. Mixing a quiet Sunday with a busy Monday in one range blurs the per-hour picture.

Reading the columns

• Calling Hour is the hour interval the row covers.
• Calls counts every call with activity in that hour. A call that starts at 9:59 and ends at 10:02 is counted in both the 9 and 10 hours, so the totals can look higher than a raw call log.
• Total Time and Average Time give the combined and per-call duration, including time spent waiting in Call queue for an agent.
• Blocking is the drop rate for the hour: dropped calls divided by active calls.
• Erlangs is the traffic load for the hour, the engine behind the staffing math.
• GoS is the grade of service, the probability a call is dropped at the agent count being tested.
• Rec Agents is the headline number: how many agents you need that hour to hit your target drop rate.
• Est Agents is the report's guess at how many agents you actually had, based on the observed drop rate.
• Calls/Agent divides calls by agents so you can sanity-check load per person.

How the recommendation is built

Under the hood the report uses Erlang B math. It converts your call volume and average handle time into Erlangs, then asks: at this traffic load, how many agents keep the dropped-call probability at or below the target you typed in? That count lands in Rec Agents.

flowchart TD
  A[Inbound calls in hour] --> B[Average call length]
  B --> C[Erlangs = calls times length]
  D[Desired drop rate] --> E[Erlang B math]
  C --> E
  E --> F[Grade of Service per agent count]
  F --> G[Rec Agents for that hour]

Turning it into a schedule

Read down the Rec Agents column and you have an hour-by-hour staffing plan. Pad it for breaks and absences, because the math assumes everyone in the count is available to answer. Pair this with your live numbers, like the ones in the All Campaigns Summary report, to confirm the forecast matches what your floor is really doing. Watching your Service level alongside Rec Agents tells you whether the plan is holding up.

The forecast is only as good as the data it runs on, so feed it clean, representative days. If you want a hosted VICIdial box with this report already wired up and your historical data intact, see VICIfast pricing.