How Sprintr calculates the Pricing Score (0–100)
The Pricing Score is a deterministic “distance from Sprintr target” score:
Step 1) Sprintr creates an internal regular-price target from the AI guidance
Sprintr converts the AI’s direction + percent-range bucket into a single internal target (using a consistent midpoint per range).
This ensures:
The guidance stays “bucketed” in the UI (safe and non-fussy)
The score can still be calculated consistently
Step 2) Sprintr measures how far your current regular price is from that target
The score is based on how close your current regular price is to Sprintr’s target regular price.
Step 3) Confidence softens the penalty (so low-confidence calls don’t over-punish you)
When confidence is lower, Sprintr reduces how harshly the score drops for being off-target. When confidence is higher, the score is stricter.
Multiple variants → one product score
If a product has multiple variants, Sprintr calculates per-variant alignment and then combines them into one product score, weighted toward the variants with more commercial impact.
What happens to the score after you run a Price Update
If you apply Sprintr’s suggested regular price
Pricing Score becomes 100 (fully aligned)
If you tweak the regular price during the update
The score updates based on how close your chosen regular price is to Sprintr’s suggested target
When Sprintr re-runs pricing AI after an update
Sprintr only re-runs pricing analysis if the applied result meaningfully diverges from the recommendation (to avoid stale guidance):
If your applied regular price is more than 5% away from Sprintr’s suggested regular target → Sprintr queues a re-run.
If your applied regular price is within 5% → no re-run; the score updates locally.
If you’re on sale:
If Regular matches, but your applied sale price is more than 5% away from Sprintr’s suggested sale target → Sprintr queues a re-run (because the sale state has meaningfully changed).
If the sale difference is within 5% → no re-run; Sprintr updates the sale indicator only.
What the score ranges mean (updated)
90–100 — On target Your regular price is aligned (or extremely close) to Sprintr’s current recommendation.
70–89 — Close You’re near the target. A small tweak would fully align you, but nothing is urgent.
50–69 — Needs a tweak Your regular price is clearly off target. A considered adjustment is recommended.
0–49 — Significantly off Your regular price is far from where Sprintr believes it should be right now. Action is strongly recommended.
What the score is not
To keep the score stable and trustworthy, it is not designed to:
Reward you for being the cheapest
Optimise around promotions by default
Push constant repricing
Treat a sale as the main lever
The score is about regular price alignment with your situation and Sprintr’s current view.
Examples: Pricing Score + Regular/Sale outcomes (updated)
Example 1 — Accepting the recommendation (Regular aligned)
Before: Pricing Score: 72
Action: You apply Sprintr’s suggested Regular price
After: Pricing Score: 100 Why it happens: the dial represents regular alignment. Applying the suggested regular price puts you exactly on target.
Example 2 — You apply a slightly different regular price
Before: Pricing Score: 72
Action: You apply a regular price close to the suggestion (a small tweak)
After: Pricing Score: high but not 100 Why it happens: the score reflects how close your chosen regular price is to Sprintr’s target.
Example 3 — On sale: Regular matched, Sale modified
Regular: matched Sprintr’s suggestion
Sale: changed away from Sprintr’s suggestion What you’ll see:
Pricing Score dial: 100 (regular is on target)
Sale indicator: Modified What happens next:
If the sale change is small, Sprintr won’t re-run analysis.
If the sale change is more than 5% away from the suggested sale target, Sprintr queues a re-run to keep guidance current.
Example 4 — On sale: both Regular and Sale modified
What you’ll see:
Pricing Score dial updates based on regular distance to target
Sale indicator shows Modified What happens next:
If either track is moved meaningfully (more than 5% away from the suggestion), Sprintr may queue a re-run.
