How many tickets come in per week

The most basic report there is, and the one most often built wrong. Which date, which grouping, and which week you leave out.

The most basic report there is, and the one most often built wrong. What decides whether it is useful is which date you count on, how you group it, and which week you leave out.

Build it

1. In Explore, click New report and choose the dataset Support > Support - Tickets.
2. In Metrics, add Tickets > Tickets.
3. In Columns, add Time - Ticket created > Ticket created - Week.
4. Set the date range to the last twelve weeks and exclude the current one.
5. Show it as a line or column chart.

Two things called almost the same

The dataset also carries metrics called Tickets created - This week, Tickets created - Last week and so on. Each of those returns one number for a fixed period. The attribute Ticket created - Week is the one that splits a chart into weeks. Note the plural: the metrics are Tickets created, the attribute is Ticket created. Reach for the metric by mistake and you get a single figure where you expected a line.

Count on created, not solved

Volume means what arrived. The Tickets metric on Ticket created - Week answers that. Grouping on a solved date answers a different question and produces a number that moves for reasons that have nothing to do with demand.

Weeks beat months

Months contain different numbers of working days, so a monthly chart shows a February dip that is a calendar artefact rather than a change in demand. Weeks are comparable with each other. Leave the current week out as well: a partial week always reads as a collapse in volume to whoever opens the report on a Wednesday.

There is no spam filter to add

The Tickets dataset has no attribute for spam. The one that exists is Requester status, with the values Active, Suspended and Deleted, and that describes the requester rather than the ticket. If you want to exclude something, that is the attribute you have, and it is worth checking what it removes before you build a filter on it.

Split by channel as a second report

The total tells you the size of the work and the channel split tells you what kind of work it is. Build the total first, then add Ticket channel to a separate report rather than crowding both into one chart.

Read it against something

Volume on its own means little. Against the same weeks last year, against your customer numbers, or against a release date, it starts to explain itself. A sustained rise is either growth or something broken, and comparing volume by subject tells you which within minutes.

See also

• Which days and hours are busiest
• Which subjects come up most often
• Setting dates and date ranges