Date Values Scoring
How to score date columns
Date Scoring lets you translate how recent (or how far in the future) a date/time value is into a numeric score. You define what counts as good, what counts as bad, and the system automatically scales everything in‑between.
How it works
- Choose a column of type
DateTime.
- Pick the metric direction:
- Lower is better – newer dates receive higher scores.
- Higher is better – older dates receive higher scores.
- Set two threshold values (see Thresholds panel in the screenshot):
- Good threshold – any value better than this gets the maximum score.
- Bad threshold – any value worse than this gets 0 points.
- Values that fall between the thresholds are linearly interpolated.
Example – Created Date (Lower is better)
To flip the behaviour simply switch the metric direction to Higher is better.
Empty values scoring
In addition to thresholds and metric direction, you can now configure how empty / missing values are scored.
Use Empty value score to choose what score should be assigned when a metric value is not available (e.g., NULL, empty, or not provided). Depending on the use case, you can treat empty values as:
- No score (0 points) — exclude the metric from scoring when value is missing
- Low (0 points) — consider missing value as bad
- Medium (half points) — neutral/partial score for missing value
- High (full points) — consider missing value as good
If the metric uses negative scoring (i.e., it assigns negative points), the Empty value score follows the same logic: selecting Low or Medium will apply a negative score, while High will result in 0 points (no penalty). For example, an empty value can be scored as -5 (Low) or -2.5 (Medium).
Threshold units
You are not limited to counting in days. Pick the unit that matches your use‑case from the Unit dropdown:
- Days ago
- Days in future
- Months ago
- Months in future
- Quarters ago
- Quarters in future
The platform automatically converts the chosen unit to an exact date relative to today, so your thresholds stay accurate over time.
Tips
- Use Days ago for short‑term freshness checks and Months/Quarters for longer lifecycles.
- Keep the “good” threshold inclusive (≤) and the “bad” threshold exclusive (>) for predictable scoring.
- Revisit thresholds periodically — business expectations of “fresh” vs “stale” can change.
Last updated on January 26, 2026