Measure

Measure can repeat configured questions across AudiAInce's AI audience archetypes and show how model-generated responses change between runs. It does not measure actual brand health, awareness, market share, or customer sentiment unless a view explicitly includes separate real-world data.

Model-Based Tracking

Configured questions may explore:

• modeled unaided or aided recall
• modeled attribute associations
• NPS-style recommendation responses
• sentiment themes in synthetic responses
• comparisons with named competitors

Keep these labels when exporting or presenting results. A model response is not a customer response.

Setting Up Tracking

1. Go to Measure in the app.
2. Select the products or concepts to compare.
3. Configure a run schedule where supported.
4. Define the questions and model diagnostics to retain.
5. Activate tracking.

Configuration

| Setting | Description | | --- | --- | | Frequency | How often to repeat the model run | | Archetypes or iterations | Model perspectives or mathematical repetitions, not people | | Competitors | Brands supplied for model comparison | | Audience context | Defined attributes supplied to the run |

Viewing Results

Dashboard

The dashboard may display:

• current model-generated scores
• change from an earlier model run
• model-simulation ranges
• archetype-level differences

Changes can result from context, prompts, model behavior, or configuration. They do not establish a market trend.

Alerts

Alerts can flag changes in stored model results or configured thresholds. Validate any consequential alert with first-party, human-research, or live-market evidence before acting.

Integrations

Supported workflows may provide API retrieval, notifications, exports, or dashboard embedding. If real-world outcome data is connected, keep it clearly separated from model-generated values and document the comparison method.

Best Practices

1. Keep questions and supplied context consistent across runs.
2. Record model and configuration changes.
3. Treat repeated runs as model monitoring, not population tracking.
4. Use customer research, analytics, or market data to measure actual outcomes.
5. Document where model and real-world evidence agree or diverge.

Related Articles

• Products - Setting up products
• Discovery - Initial audience exploration
• Calibration - Understanding reference comparisons