Jobber's Schedule With Eight Vans: Where It Struggles, and the Custom Calendar We Built Instead

Where does Jobber's schedule struggle for a window cleaning round?

Not in storing the work, but in planning it. Jobber holds clients, recurring jobs and visits well. The friction starts when the question is geographic and multi-van:

• The van is not the planning unit. Jobber schedules team members. Vehicles exist only on some plans, the API cannot create team members, and nothing ties a vehicle to a patch of postcodes. A round is planned by van, and crews swap vans on the morning.
• No sense of territory. Nothing tells you a job is sitting in the wrong van's patch, or that CF83 has 63 jobs today and CF24 has one.
• Busy days get crowded. A day with a few hundred visits across eight vans is hard to read in a general-purpose calendar.
• No capacity view. It shows jobs, not whether each van has a full day in all-around and front-only terms.

None of this is a criticism of Jobber. It is a general field service tool. A window cleaning round is a specific shape, and the planning view should match it.

What does a window cleaning schedule need?

Why not just read Jobber live?

Because a round at 7am is a lot of reads at once. The first version read each day straight from Jobber whenever someone opened it. On launch day, with every cleaner opening their phone at the same time, that overloaded Jobber's API.

The design that works keeps the planning layer's own copy of each day in a database and serves screens from that. The copy stays honest three ways:

1. Jobber webhooks report every visit change as it happens, including changes made inside Jobber.
2. Every change made in the app updates the copy immediately.
3. An hourly check catches anything the other two missed.

Days further out, weeks five to eight, are fetched slowly in the background, about a day a minute, so Jobber is never flooded. If Jobber is slow or down, the crew still sees their round, and any change that cannot be written fails with a clear message instead of pretending it worked.

Should you build a custom calendar?

Only if the planning view is genuinely costing you time. With one or two vans, Jobber's calendar plus good territories is usually enough. Past three or four vans, if someone is spending every evening building tomorrow by hand, a planning layer pays back quickly, and it is the foundation for the parts that save the most: territories, capacity targets and missed-visit handling.

Frequently asked questions