Preview impact lets you model a change to your matching settings and see how it would play out, without touching your live account. Nothing you do here changes how your contacts are matched.
Go to Settings > Identity resolution > Preview impact.
The screen is marked Preview only and works in two steps. First you set up a draft, then you compare sample results against your live settings.
Step 1: Set up a draft
The Live card on the left shows what is running on your account right now: your merge threshold, how many contacts you have after merging, and your merge rate. Select View settings to see the full configuration.
Select Create draft settings to start a draft. It begins as a copy of your live settings, so the two cards match until you change something.
Select Edit draft settings to model a change. You can adjust the merge threshold, the similarity functions used to compare identifiers, and the source weights.
Reading the draft numbers
Once the simulation finishes, the draft card shows the same figures as your live card, with the difference against live next to each one.
| Figure | What it tells you |
| Merge threshold | The minimum confidence needed to merge automatically under the draft. |
| Contacts after merging | How many contacts you would have. Lowering the threshold merges more contacts, so this number falls. |
| Merge rate | The share of contacts that would be merged. Lowering the threshold pushes this up. |
The footer tells you how many settings differ from live. Select Compare with Live to see them side by side, or Reset to Live to discard the draft and start again.
While the simulation runs
Running a draft can take a while, so it runs in the background. You can leave the screen and carry on working.
Ometria notifies you when the simulation is ready, and you can check progress in the notifications dropdown in the top right. Select Cancel simulation if you want to stop it.
Step 2: Compare sample results
This table shows individual contact pairs and what would happen to each one under your live settings and under your draft. The pairs range from a certain match to a weak one, so you can see exactly where your change flips a decision.
| Column | What it shows |
| Candidate pairs | The contacts in the pair, with a label showing where they sit relative to your cutoffs. |
| Top matching signal | What linked the two contacts, for example, a loyalty card match or a name and address fuzzy match. |
| Live outcome | The confidence score and decision under your current settings. |
| Draft outcome | The confidence score and decision under your draft. |
The three bands
Every row is labelled and colour-coded by where it falls:
| Label | What it means |
| At / above your cutoff | The pair merges under both live and draft settings. Your change does not affect it. |
| Between your Live and Draft cutoffs | The pair merges under one and not the other. These are the rows your change actually decides. |
| Below cutoff | The pair merges under neither. Your change does not affect it. |
The middle band is the one to review. Those are the merges you would gain or lose, and reading a handful of them is the fastest way to tell whether your change is safe.
Scores can move as well as decisions
If you only change the threshold, the confidence scores stay the same, and the decision flips. If you change a similarity function or a source weight, the score itself moves, because you have changed how the evidence is judged. Both kinds of change show up in this table.
Select View reasoning on any row to see the field-by-field evidence behind a pair.
Applying a draft
A draft never goes live on its own. When you are happy with what the preview shows, contact your Technical Project Manager (TPM) to implement the settings on your account.
Comments
0 comments
Please sign in to leave a comment.