Lead Scoring Guide
4 min read
How It Works
Every customer gets a score from 0 to 100, built from their order history and measured against your own store rather than an industry average. Buyers are ranked on what they are worth to you; visitors who have never ordered are ranked on how recently they were active, and capped so they cannot outrank a real customer. The score is calculated from published weights - no black box - and it is the same number everywhere it appears: the leads table, the live feed, the lead profile and the embed.
Scoring Factors
Customers with at least one order:
Lifetime spend, measured against your store’s other buyers
How often they order, against your store’s buying cadence
Days since the last order, decaying on a 180-day half-life
Leads who have not ordered yet:
Days since the last tracked interaction, decaying on a 180-day half-life
Estimated upside from profile signals, capped so a lead with no order can’t outscore a real buyer
Percentiles come from your store’s own buyer history, so a $500 average-order-value shop does not score the same as a $20 one. Custom rules you configure in Scoring Settings are added on top of this base score, between 0 and 100.
Scoring Models
The weights above belong to the Balanced model, which every account starts on. Five presets ship, all running the same engine with different numbers - pick the one that matches how your store sells, from AI Studio in the dashboard. Switching recalculates your scores; custom rules still apply on top either way.
Value 45% · Frequency 20% · Recency 35% · 180-day half-life
The default. No single signal dominates.
Value 70% · Frequency 10% · Recency 20% · 365-day half-life
For high-ticket, low-frequency sales. A large one-off order keeps its worth for a full year.
Value 20% · Frequency 20% · Recency 60% · 90-day half-life
For replenishment and subscriptions. Yesterday’s buyer outranks last year’s whale.
Value 20% · Frequency 55% · Recency 25% · 240-day half-life
For repeat-purchase brands. The regular outranks the occasional big spender.
Value 15% · Frequency 15% · Recency 70% · 60-day half-life
For win-back campaigns. A drifting customer falls while their spend still looks healthy.
Lead Segments
High purchase intent. Target immediately with personalized offers.
Interested but not ready. Nurture with content and follow-ups.
Low engagement. Re-engage with win-back campaigns.
A past customer whose orders have stopped - at least 90 days of silence, or quieter than your own buyers’ 75th-percentile gap. The label follows the order history, not the number.
Re-scoring
Scores recalculate themselves as orders arrive through Shopify webhooks, and each lead profile shows why it scored the way it did. To rewrite every score after changing a rule or editing orders, use the Rescore All button in the dashboard action bar.
Switching scoring model works the same way: the change is recorded immediately, and every score still carrying the old model is rewritten by the next scheduled rescore, so your numbers stay consistent once it has run.
Want custom scoring rules?
Add your own conditions and points in Scoring Settings (Growth and above), or contact us at support@leadup.business.