Datasets

Datasets let you prepare realistic sample data without changing your live organisation. Each demo, sandbox, or training dataset is isolated from Production.

Non-production datasets and their client workspaces don’t send email. You can still see in-app notifications and test workflows, forms, invitations, and other features, but Runnit suppresses their email delivery. Password reset and email verification messages for your Runnit account still work.

Organisation owners and administrators can open Settings > Datasets.

Create and switch datasets

To create a dataset:

1. Enter a New dataset name.
2. Choose Demo, Sandbox, or Training.
3. Select Create and switch.

Use the buttons under Active dataset to move between Production and your other datasets. A dataset that’s still seeding or has failed can’t be selected. You can’t switch or create another dataset while a seed request is running.

The switcher lists only the datasets that belong to the workspace you’re signed in to. If you belong to several workspaces, each workspace’s site shows its own datasets. When the list covers more than one organisation, for example a client workspace with its own datasets, each button is prefixed with the organisation name.

Impersonating a teammate keeps your current dataset: if you’re in Production when you start acting as another user, you stay in Production. You return to your own active dataset when you end impersonation.

Seed sample data

1. Switch to the dataset you want to seed.
2. Under Staff to plan against, choose Existing agency staff or Fictional employees.
3. For fictional employees, choose a team size between 10 and 300.
4. If the dataset already contains work, tick Replace the projects and planning data after checking the dataset name.
5. Select Seed dataset or Re-seed dataset, then wait for completion.
6. Open Schedule to review the sample bookings and team capacity.

Existing agency staff is the default. It assigns sample projects and tasks to your active staff without changing their live projects, login passwords or profiles. It also creates separate copies of their working patterns, leave and office closures, so part-time hours and unavailable days appear in capacity planning. These copies are taken when you create the dataset and refreshed when you re-seed against existing staff. Changes made in the dataset stay there. Production changes made later need another seed to appear in the demo. Connected calendars and external busy events aren’t copied.

Sample bookings follow your dataset’s local working hours and staff availability. Capacity slots use the same bookings and assignees. Limited hours can push work past its sample delivery date, so you can review late plans as well as overload. If a staff member’s working pattern has no working days, seeding stops without replacing the dataset. Review that pattern in Production or choose fictional employees.

Shared staff profiles, pictures and work history are read-only from the demo. Switch to Production to edit them. Completing demo tasks doesn’t add fake experience to a staff member’s portfolio. Staff accounts created only for the dataset can still be edited there.

Fictional employees creates a sample roster with reserved email addresses. Runnit refuses to replace an existing login account if a sample address clashes.

Re-seeding replaces projects, tasks, schedules, time entries, rates and generated assets in the selected dataset. It refreshes staff availability copies in existing-staff mode and moves generated dates so they’re anchored to today. Production records and shared staff accounts are preserved.

Generated projects are spread across a useful demo horizon. Some begin up to two weeks before the seed date, while delivery dates extend up to two months after it. The portfolio includes draft, active, review, client-review, and completed work.

Each project has a client-aware brief with a realistic objective, audience, deliverables, channels, measures, and approvals. Its milestones fall inside the project dates, and task descriptions refer to deliverables in that brief. Staff availability determines when the task bookings finish. Draft-project schedules and resource bookings are marked tentative; bookings for other project states are confirmed.

Caution

Re-seeding is permanent for the selected dataset. Check the active dataset name before you continue.

If seeding is interrupted

If the connection drops or the server returns an error, Runnit might have finished seeding without being able to confirm it. It refreshes the dataset status and does not repeat the seed automatically. While seeding is still running, the page checks for completion for up to five minutes. You can also choose Refresh status, including if a status check fails.

Review Last seeded and the dataset’s projects before starting again. If status is unavailable, choose Refresh status. You cannot start another seed until status loads, and replacing existing work requires a new acknowledgement.

Production overrides

Production replacement is restricted to the specifically authorised production-data owner. Other owners and administrators must switch to a non-production dataset before they can seed data.

For the authorised account, the page shows Override production dataset and an acknowledgement checkbox. The button stays disabled until the warning is accepted. The override keeps that account’s existing login password, but replaces the Production data described above.

An impersonated session can’t run a Production override. Existing-staff seeding is available only in non-production datasets.

Client assets are isolated

Generated datasets don’t connect demo clients to Production asset collections. They create separate client records, separate brand collections, and fresh generated assets inside the selected dataset. Re-seeding a demo dataset rebuilds those demo records without changing or linking to Production records.

Next, see Client Organisations to learn how clients, teams, and shared assets work in a live organisation.