SEEScore

SEEScore™ is a composite performance score ranging from 25 to 200, calculated for each location in your network. It combines multiple configurable metrics — each with its own weight and performance thresholds — into a single comparable number that reflects the overall health of a location.

Available Metrics#

The following metrics can be included in your SEEScore™ configuration:

Sales#

Labor#

Food & Product Cost#

People#

Customer Experience#

Compliance#

Custom & Competitive#

These metrics are defined per account rather than being fixed, numbered slots. Each one is configured with its own name, and is identified by a key derived from that name (for example, a "Loyalty Signups" custom metric is keyed custom:loyalty_signups, and a "Rival Brand" competition metric is keyed competition:rival_brand).

Calculation#

SEEScore™ is calculated in three stages.

1. Metric Weights#

Each metric in the configuration is assigned a weight from 1 to 5, which controls how much that metric influences the final score.

2. Performance Component#

Every metric contributes a performance score based on two configured thresholds — a lower and an upper threshold — and a per-metric choice of scoring curve. The performance multiplier (0–10) is then multiplied by the metric's weight.

The scoring curve is chosen per metric on the score configuration page; new metrics use the smooth curve by default. The maximum performance points for a metric is weight × 10 with either curve.

Smooth curve (default)#

The performance multiplier ramps linearly from 0 at the lower threshold to 10 at the upper threshold — clamped to 0 below and 10 above. This avoids abrupt jumps at the thresholds and differentiates locations that fall within the same range. A location halfway between the thresholds scores 5, a quarter of the way scores 2.5, and so on.

Step curve (alternative)#

Instead of a continuous ramp, the thresholds define three fixed tiers:

Directionality: For metrics where lower is better (labor %, food cost %, turnover rates), the threshold comparison is inverted automatically — a sufficiently low value earns full points, and a value above the upper threshold earns zero. Both curves invert the same way.

Compliance: Treated as binary regardless of curve — compliant earns full points (10 × weight); non-compliant earns zero.

Setting thresholds#

Thresholds can be entered manually or generated with the Suggest range tool, which reads the actual distribution of location values for the metric over a chosen period and proposes a lower and upper threshold from percentiles of that data. You can then fine-tune before applying.

The tool offers three splits. They differ in how much of the network lands in the poor/excellent extremes versus inside the band:

Which to choose: quintiles is the sensible default — it keeps most locations inside the band and flags only the clear standouts and laggards. Move toward quartiles or tertiles when you want stricter grading that spreads more locations to the extremes.

This choice matters more with the smooth curve: only locations inside the band are graded continuously, while anything past a threshold clamps to 0 or 10. A wider split (quintiles) therefore keeps more of the network on the differentiating ramp, whereas a narrow split (tertiles) clamps the majority at the extremes — which makes the smooth curve behave more like the step curve. For smooth metrics, lean toward the wider splits.

3. Normalisation#

After summing earned points and total possible points across all metrics, the raw score is normalised to the 25–200 scale:

SEEScore = round(earnedPoints × (175 / totalPossiblePoints)) + 25

• The minimum achievable score is 25 (every metric at Poor — i.e. each value at or below its lower threshold).
• The maximum achievable score is 200 (every metric at Excellent — each value at or above its upper threshold).
• With the smooth curve, scores vary continuously between these bounds: a location's score reflects exactly where its values fall within each band, rather than snapping to a small set of tier-based totals as the step curve does.
• Metrics with no data for a location are excluded from both earned and possible points — a location is scored only on the metrics it has data for, so a missing value neither helps nor penalizes its score.
• A score of 0 indicates no metric data is available for the location at all.

Score Reference#