A Lead Score Is a Routing Decision.

Here is what scoring usually looks like six months after it is switched on. Every lead has a number. The reps sort by it sometimes. Nobody can say what a forty means or what happens at fifty, and the lead that came in at nine on a Tuesday morning is still waiting at four. The number was built as information. It only pays when it is built as a decision, and the decision is one threshold with a workflow on it. Everything on screen is a mock demo of Zoho CRM with sample data.

The leads, the companies and the points are invented for this video. This is the Leads list in the mock demo, with the score column on the right. Fourteen leads, scores from minus four to sixty-one. Look at the top two: a sixty-one that has been in the system for three days, lead status new, no owner yet; and a fifty-eight whose owner is a rep who left last month. The scoring is working. The scoring is also changing nothing, because nothing reads it. That is the first mistake the article behind this video names, and the most common: a score with no automation on it.

The other three are scoring only positive signals so every lead climbs forever, confusing these rules with the assistant's predictive score, which answers a different question, and never checking the threshold against who actually bought. We will fix all four, starting with the one that matters. Scoring rules add points from two places. Field values: a lead in a target industry earns points, a company above a size earns points, a lead from the referral source earns more than one from a list. Touchpoints: an email opened, a webinar attended, a form filled, each earns a little.

Each item scores between minus ten and plus ten, and the minus matters: a competitor's email domain, an unsubscribe, a free mailbox on a business enquiry, each takes points away. The mock rules on this card use only standard fields: Industry, number of employees, Lead Source. Six positive items, three negative. The negatives are what keep the number honest; a scoring model with no way down is a scoring model where every old lead eventually looks hot. Now the decision. At what score does a person act? Not sort, not glance, act: pick up the phone within the hour.

In the mock demo the sales manager chose twenty, and the reasoning is the important part. She looked at the last quarter's closed deals, asked what they scored on the day they came in, and most were above twenty on arrival. Below twenty, almost none. Twenty is not a magic number; it is her number, read from her history. One threshold, one action. Companies that set three thresholds and five actions on day one end up with none of them trusted. Twenty means a human calls. Everything below twenty is nurtured by the marketing system and re-scored as it moves.

That is the whole design, and it fits on this card. If your history is thin, a new company or a new product line, start at the number the sales manager would act on by instinct and write down the date you chose it; the quarterly check on the last card will correct it. Here is the workflow that makes the threshold real. The trigger is the score itself: when a lead's score changes and crosses twenty, the rule fires. Two actions. First, assign the lead to a rep, round-robin across the team or by region, which is the same assignment logic we showed in the Friday-at-six video.

Second, create a task on that rep's list: call, due in sixty minutes, with the score and the reason in the subject. That is it. The score is no longer a column; it is the thing that decided who owns this lead and what they do before lunch. And the sixty-one that sat for three days in the first scene would have been on a rep's list within a minute of crossing twenty, because the rule reads the score, not the calendar. Scoring and assignment are two steps, not two choices.

The rules decide how good the lead is; the assignment decides whose it is; and a setup that only has the second one treats every lead as equally urgent, which is the Friday-at-six problem in a different coat. The last card is maintenance, and it is quarterly, not daily. Pull the leads that crossed twenty last quarter and see how many became deals. Pull the ones that never crossed and see how many became deals anyway. If the second number is large, the threshold is too high or a signal is missing; if the first is small, a positive item is too generous.

Move the number, not the whole model. Two closing notes. The assistant's predictive scoring, which learns from your history, answers a different question from these rules and is worth reading about separately; the article is linked. And the reason to do any of this is in one line from that article: a lead contacted within the hour is far more likely to qualify than one that waits. The threshold is how the hour happens. A lead score is a routing decision. Points from fields and touchpoints, up and down; one threshold read from your own closed deals; one workflow that assigns the lead and creates the call within the hour; one quarterly check.

Build it that way and the number on the record is the reason the right rep called first. We are CodeStringers, a Zoho authorised consulting partner. The two scoring articles and the assignment-rules video are linked below, and so is a no-risk discovery, paid only if you proceed.