What the output looks like

One sample row, walked end to end

A ranked list only helps if a rep can act on a row without asking what it means. Here is one row from a sample run, broken into the parts a rep actually reads.

The sample row

Names and figures illustrative. Ranked by possible loss.

Reading it in the field

1. Step 1

   Signal

   Orders fell from 6 to 3. The last order was 21 days ago against a usual gap of 9 days. Both facts come from the order file alone; no extra integration is needed to see them.

2. Step 2

   Reason

   The row states the change in plain language rather than a score, so the rep can see why this dealer is above the others before picking up the phone.

3. Step 3

   What to ask

   The action column tells the rep what to check first. For a fall in order frequency against the account’s own rhythm, that is a call to confirm whether demand, stock, or a service issue is behind it.

4. Step 4

   Outcome field

   What the rep finds gets recorded against the account. That feedback is what improves the next run, and it is the difference between a list that is used once and one that stays credible.

How the ranking is tiered

Every dealer lands in one of three action tiers, so the list reads as a work queue rather than a wall of names:

• Call now — a clear break from the account’s normal pattern, worth contacting this week.
• Win-back — the account has already slowed sharply or stopped, so the conversation is a recovery attempt.
• Watch — the pattern looks like accounts that churn, but there is no confirmed change in orders or spend yet.