Experiment Management

The whole process in one place:
hypothesis → test → card → archive

No more Slack for requests, Jira for kanban, Excel for the gantt, Confluence for the archive. One workspace for the entire experimentation workflow.

Kanban Board

Hypothesis pipeline visible to everyone

Stakeholders — product managers, designers, marketers — create hypothesis cards themselves. No bottleneck through the analyst to get an idea into the queue.

  • Backlog → Planned → Active → Done — four stages, one board
  • Each card tracks hypothesis text, metric, expected effect, tags, and owner
  • Results are attached to the card when the experiment completes
workbench.ab-labz.com/hypotheses
Hypotheses Kanban Board
Hypothesis Card

Everything the experiment needs — in one form

When creating a hypothesis, the team fills in everything upfront: the hypothesis text, key metric, expected effect, treatment location, tags, split system ID, sample size, audience slice, traffic allocation, and variants. Nothing gets lost in Slack or a separate doc.

  • Hypothesis text, key metric, expected change, treatment location, and tags
  • Split system ID, required sample size, audience slice, and traffic allocation
  • Up to 10 variants — control and any number of test groups
workbench.ab-labz.com/hypotheses
Add Hypothesis Modal
Gantt Calendar

See what's running and how much is left

Each experiment is shown as a bar on a timeline — with its audience share displayed directly on the bar. Instantly see which tests overlap and how much of the audience is still free.

  • Audience allocation shown on every bar — easy to estimate what's still available
  • Visual overlap detection — plan new tests without invalidating running ones
  • Feeds into Portfolio Analytics to track capacity utilization over time
workbench.ab-labz.com/calendar
Experiment Calendar
Experiment Card

One record from hypothesis to archive

The card you create before the test becomes the analysis card when it's done — and then the archive entry. No information is ever lost or re-entered.

Hypothesis

Metric, expected effect, MDE, variants, allocation

Active Test

Live monitoring, SRM checks, sample progress

Analysis

Engine results, AI conclusion, ship/reject decision

Archive

Searchable, feeds Portfolio Analytics

workbench.ab-labz.com/archive
Experiments Archive
Shared Cards

Share any experiment card as a public link

Every archived card can be shared as a read-only public page — hypothesis, analysis results, AI conclusion, and all metadata in one view. No login needed for the reader.

  • Send to a stakeholder or PM — without granting workspace access
  • Full context: hypothesis, variants, results, AI conclusion — in one page
See a live example →
workbench.ab-labz.com/share/…
Shared experiment card

Replace a stack of tools with one

Available free for solo users. For teams — register and invite your colleagues.