Part B · The screens
B.17Measurement
What the outreach actually earned, measured against a held-out group that got nothing, with the range of the result and a verdict in words. A second tab backtests the scoring on past deals.
- No.
- B.17
- Dashboard address
- /measurement
What you can do here
Each core task, and how many clicks it takes from this screen.
| Task | Clicks | How |
|---|---|---|
| See the lift of every wave | 1 | Measurement: one row per wave, treated against held out, the lift with its range, and the verdict. |
| See how a number was counted | 2 | Measurement → Why? on the number: the arms, how the lift was counted, the interval and the verdict. |
| See a wave's accounts | 2 | Measurement → the wave: its accounts, filterable to the held-out group. |
| Backtest the scoring | 2 | Measurement → Backtest: had the AI scored past accounts, where would the winners have ranked? |
How it works behind the scenes
- Each outreach wave is an experiment: a fixed, stratified 20% of the chosen accounts is held out and gets no e-mail, so the difference in outcomes can be credited honestly.
- Outcomes (replies, meetings, deals and revenue) come back from Nutshell through the sync; the lift is shown with a 95% range, and the verdict says plainly when the range is too wide to call.
- Views by segment and by tier are one click away and live in the address, so a view can be shared.
- Held-out accounts are never shown to reps as targets and nothing about the holdout is written to Nutshell.
Rules and switches that apply here
- Any history made up for the prototype is labelled "simulated" on the screen and never mixed with real results.
The tables behind it
PK primary keyFK reference the database enforces (solid line)ID reference kept by id (dashed line)
| Table | What it holds | Columns |
|---|---|---|
| ctrl.experiments | Each experiment: its arms, the holdout share, how it is stratified and measured. | 17 |
| ctrl.outreach_waves | Each outreach wave, its segment, size and the experiment it belongs to. | 11 |
How the tables connect (2)
| Column | Points to | Kept by |
|---|---|---|
| ctrl.outreach_waves.experiment_id | ctrl.experiments | The database (foreign key) |
| ctrl.outreach_waves.approval_id | ctrl.approvals | The application (by id) |
The screen
localhost:13301/measurement
Measurement: each wave against its held-out group.
- 1Simulated history is labelled as such.
- 2All waves together: treated, held out, lift and the verdict.
Dashboard · Measurement. Captured from the running POC on September 25, 2026; synthetic data.
Not yet
- The prototype's wave history is simulated and labelled so; real lift comes from DMA's own sends.
See it live
Opens the screen in the running POC, in a new tab.
| Open | Where | What to try |
|---|---|---|
| Measurement | DMA AI dashboard | Why? on any number explains how it was counted. |