My ACD only reports in 30-minute intervals. Can I still plan in 15?

Only by splitting each half hour in two, which invents a shape the data does not contain. Plan at the resolution you can measure. If the peaks are sharp enough to matter, the fix is a finer report from the ACD, not a finer spreadsheet.

Does the interval change the total staffing?

Yes, a little. Finer intervals see sharper peaks and troughs, so the sum of the per-interval requirement usually rises, and the roster fits it less well. Coarser intervals average the peaks away and understaff them. The sum is not the point; the peak is.

Which reporting interval to plan in: 15, 30 or 60 minutes

Every staffing calculation starts with a choice most planners inherit rather than make: the length of the interval. Fifteen minutes, thirty, or an hour. It sets what the forecast can see, what the Erlang calculation assumes, and how well a roster can be made to fit. This guide gives the rule, explains the one effect that catches people out, and shows what moves when you change it.

What the interval decides

The Erlang calculation takes the contacts in an interval and treats them as arriving at a steady average rate across it. Inside the interval there is no shape; the shape lives in the differences between intervals. A day planned in hourly intervals sees twelve numbers; the same day in fifteen-minute intervals sees forty-eight, and the peak in the finer view is always higher than in the coarser one, because averaging over an hour flattens the busiest quarter.

That is the argument for short intervals: they see the peak. Staff to an hourly average and the busiest fifteen minutes of that hour is understaffed by whatever the average hid.

The argument against: overhang

A contact that starts at 09:14 and lasts seven minutes is counted as one contact in the 09:00 interval, and it occupies an agent until 09:21, which is in the next one. The Erlang calculation for 09:00 to 09:15 counted its whole handle time; the calculation for 09:15 to 09:30 counted none of it. When handle time is short relative to the interval this washes out. When it is long, it does not: with a seven-minute handle time and fifteen-minute intervals, roughly half the work counted in an interval is actually done in the next, and both calculations are wrong in opposite directions.

The working rule is that handle time should not exceed half the interval. Seven and a half minutes for a fifteen-minute interval, fifteen for a thirty, thirty for an hour. Every calculator on this site checks this and warns when it is breached; it is the most common reason two calculators disagree about the same day.

Choosing

Voice with a handle time under five minutes: fifteen-minute intervals if the ACD reports them, thirty otherwise. The peaks are sharp enough to matter and overhang is small.
Voice with a handle time of five to fifteen minutes: thirty-minute intervals. Most inbound centres sit here, which is why thirty is the default everywhere.
Technical support, complex sales, anything over fifteen minutes: hourly. Overhang at thirty minutes is too large, and long-handle queues have gentler peaks anyway.
Chat: by elapsed handle time, which at concurrency two or three is often ten minutes or more. Thirty minutes, usually.
Email and back office: intervals barely matter; the turnaround time is the unit.

If the ACD reports at a coarser resolution than the rule suggests, plan at the ACD’s resolution. Splitting a half hour into two identical quarters invents nothing and reassures nobody.

What changes when you switch

Take the reference interval used across these guides: 200 contacts in thirty minutes at a five-minute handle time, 80/20. That needs 39 agents. Reported hourly, the same demand is 400 contacts in sixty minutes: still 33.3 erlangs, still 39 agents if the hour is flat. If the hour was in fact 230 contacts in the first half and 170 in the second, the hourly view staffs both halves at 39 and the first half fails; the half-hourly view staffs them at 44 and 34.

The finer view costs more in total (78 against 78 in this case, but usually more) and fits a roster less well, because shifts cannot change every fifteen minutes. That is the trade: fidelity against roster efficiency. Most centres settle on thirty minutes because it loses least on both.

Try it

Enter your own handle time and change the interval. The overhang warning appears when the rule is breached.

Inputs

Results

Agents required

On the phones for the whole interval, before shrinkage.

Service level achieved: 82.6%
Occupancy: 85.5%
Average speed of answer: 13.4 s
Contacts that will wait: 25.4%
Workload: 33.3 erlangs

Show the working

Workload = 200 contacts ÷ 30 min × 300 s ÷ 60 = 33.3 erlangs.
The queue is only stable with more agents than erlangs, so the search starts at 34.
With 39 agents, Erlang C gives a 25.4% chance that a new contact finds every agent busy.
Service level = 1 − 0.254 × e^(−(39 − 33.3) × 20 ÷ 300) = 82.6%, which clears the 80% target.
With 38 agents it would be 75.6%, which does not.
Occupancy = 33.3 ÷ 39 = 85.5%.

Doing this for every interval of the week? Pebble WFM computes the requirement from your forecast and builds the roster. Free month, no card needed.

Where next

How to calculate Erlang C staffing: what the calculation assumes inside the interval.
How to calculate FTE from contact volume: the intervals summed into a week.