QA Runs
See the Glossary for what a QA Run is. Each one launches one simulation for every selected Persona (or Generic) × Journey × viewport combination and records an outcome of Passed, Failed, or Indecisive for each one. Journeys in the same Persona/Generic × viewport lane run serially and carry the same browser forward, while distinct lanes can run independently.
Starting a run
Section titled “Starting a run”From a QA Flow, choose Run and configure the run:
- Journeys grid — Journeys against the Personas (or Generic) bound to them. A cell exists only where that Persona drives that Journey. Each ticked cell runs as one simulation; untick the cells you don’t need this time. A run can only pick from the cells the QA Flow already has — it never adds a Persona to a Journey.
- Viewports — Choose one or more of Desktop, Tablet, and Mobile. Runs are Chrome-only; each selected viewport multiplies the ticked cells so you can compare every Journey across screen sizes.
A run supports at most 50 simulations in total across every ticked cell and viewport.
A run starts immediately and you’re taken to its results page. Journeys update in real time as each persona works through them. To watch one browser live, open its Simulation from the Knowledge page’s App explorations list, filtered to QA.
Watching a run
Section titled “Watching a run”The run itself reads Running until it ends as Completed, Failed, or Stopped. It shows Paused instead of failing once your plan’s included credits and your Usage Credits are used up. A workspace admin can buy Usage Credits or turn on auto-reload, and the run waits until you resume it — see Why did my run pause?. Each outcome stays In progress until it settles, then reads as one of:
| Outcome | Meaning |
|---|---|
| Passed | The persona reached the expected outcome and the followed steps aligned with the plan. |
| Failed | The expected outcome was not reached. |
| Indecisive | Marketrix could not conclusively pass or fail the Journey; you resolve it, and the resolution is remembered per application. |
The run summary shows live counts of Passed, Failed, and Indecisive outcomes. You can Stop QA Run at any time — the run ends immediately and the outcomes collected so far are kept.
Reading results
Section titled “Reading results”Every Journey is evaluated automatically against its expected outcome. The outcome explains:
- Whether the outcome was reached.
- Whether the followed steps aligned with the planned steps.
- A short, human-readable reason for the outcome.
Resolve an Indecisive outcome via Resolve — choose a Ruling of Passed or Failed and explain why. See Common Issues if a run keeps landing on outcomes you didn’t expect.
Reports
Section titled “Reports”From a run you can generate its report on demand with Download report: one PDF for the whole run. Unlike the live run view, it also includes each Journey’s walkthrough — the steps Marketrix actually followed, each with its screenshot — alongside its outcome and planned steps.
Run history
Section titled “Run history”Every run is kept, so you can revisit the results and the report for any past run. The run history table lists each run with when it ran and the viewports it covered. It also shows the simulation count (selected Persona/Generic × Journey cells multiplied by selected viewports) and the status. Coverage is landed outcomes out of the expected Persona/Generic × Journey × viewport cells.
Triggering runs from workflows
Section titled “Triggering runs from workflows”QA Runs don’t have to be started by hand. A Trigger — a GitHub pull request or a scheduled timer — can start a workflow that runs a QA Flow, or you can run one on demand from Slack. See Connectors for what each provider contributes.