# SDR Call Intelligence Settings

> Define the playbook, personas, discovery behavior, next steps, and scoring weights used in call evaluation.

Canonical URL: https://marcenta.ai/docs/administration/sdr-call-intelligence-settings
Markdown URL: https://marcenta.ai/site-md/docs/administration/sdr-call-intelligence-settings

Owners and admins can open **Settings → Call Analysis** to configure the shared evaluation rubric. Complete this setup before relying on automated coaching or quality results.

## Evaluation inputs [#evaluation-inputs]

Configure the criteria that reflect the current sales motion:

| Input               | What to provide                                                 |
| ------------------- | --------------------------------------------------------------- |
| Pitch script        | The approved talk track or playbook used to evaluate the call   |
| Pitch points        | Individual ideas or value points representatives should cover   |
| Discovery questions | Questions or topics expected during discovery                   |
| Accepted next steps | Valid outcomes that represent forward progress                  |
| Ideal personas      | Titles or roles the team is expected to engage                  |
| Personas to avoid   | Titles or roles that generally fall outside the target audience |
| Seniority hints     | Terms that help interpret role seniority                        |

List-based fields accept one item per line. Use specific operational language rather than broad aspirations. Similar or contradictory entries make evaluation harder to interpret.

## Scoring weights [#scoring-weights]

Weights determine how the configured evaluation areas contribute to the overall quality assessment. The saved weights must total exactly 100%.

Start with the relative coaching importance of each area, then validate the rubric against a representative sample that includes strong, average, and weak calls. A mathematically valid total can still produce an operationally poor rubric.

## Defaults and blank fields [#defaults-and-blank-fields]

When a configurable rubric field is blank, the product uses its default evaluation guidance. Leaving a field blank does not necessarily disable that evaluation area.

If your organization intentionally does not evaluate an area, confirm the product behavior with your administrator or support before assuming an empty field removes it.

## Apply a change [#apply-a-change]

1. Record the reason for the rubric update internally.
2. Edit the relevant criteria.
3. Confirm all weights total 100%.
4. Save the settings.
5. Reprocess a small representative sample where appropriate.
6. Compare the new results with manager judgment before broader use.

New settings affect calls analyzed or reprocessed after the change. Existing completed analyses can continue to show results produced under the earlier configuration.

## Governance [#governance]

* Review the rubric when messaging, qualification, personas, or the sales process changes.
* Avoid frequent edits that make scores incomparable over time.
* Tell managers when the rubric changes.
* Use transcript evidence and recording context during reviews.
* Do not use an automated score as the sole basis for employment decisions.

<Callout type="info">
  Automated evaluation is decision support. Recording quality, transcript accuracy, call context, and manager review all affect how a result should be interpreted.
</Callout>