Field notes · Distributor management

The 90-day distributor check

By the Sightglass team5 min read10 July 2026

A distributor going quiet is not one lost customer — it is every account behind that distributor going dark at the same time. One wholesaler can sit between you and dozens of bars, a run of bottle shops and a chain reset. When it stops moving your cases, all of them stop together, and each one lands in your data as its own small, unremarkable gap. Nothing in the file says they share a cause.

That is why the distributor check belongs at the top of the list rather than somewhere in the middle. It is the cheapest report you will run all month — one derived column, one sort — and one of the few where the next step is a single phone call rather than a project. When we ran that column across one importer's own exports, 113 of 242 distributors had no sale in 90 or more days.

Build the column

The check is a single derived column: for every distributor in your book, the most recent date on which any case moved to any account. Not volume, not trend, not mix. Only when the file last showed a sale.

01Pull the widest depletion export you have — every distributor, every SKU, every account, the longest date range the report will give you.
02Group by distributor and throw the rest away. Account and product don't matter here; you want one row per distributor.
03Take the latest date on any row with cases above zero. That is the distributor's last-sale date.
04Measure the gap from the file's own end date, not from today. The export's window is the truth, and it always trails the calendar.
05Sort ascending. The oldest dates at the top are the check.
06Add one more column: cases that distributor moved in the twelve months before it went quiet. That, not the date, is your call order.

Group on the distributor number, not the name. The same wholesaler routinely appears under two or three spellings across a multi-state book, and a name-based grouping splits one quiet house into three noisy ones — hiding the row you were looking for.

Depletion data lags a day or two behind the file's end date, and a mid-cycle pull under-reports the most recent weeks. Confirm on a second file before you treat a distributor as quiet. One thin period is not a trend.

Why ninety days

Ninety days is not a magic number. It is the point at which the ordinary excuses have all run out.

Most wholesalers reorder on a cycle measured in weeks, not months. A missed cycle is a holiday, a truck, a warehouse move, a buyer on leave. Two missed cycles is a pattern. By ninety days you are past a full quarter of normal rhythm, past a seasonal lull and well past the reporting lag — so a distributor still showing nothing is quiet for a reason somebody at that house could give you in a minute.

Tighten the window where the rhythm is faster than the rule:

Fast-turn packages. If a distributor takes draft or cans every month, ninety days is two months late. Forty-five is closer.
Chain-driven volume. Where the business rides on resets, the window is the reset calendar, not the calendar month.
New placements. A distributor that has ordered three times ever has no rhythm to compare against; watch those weekly for the first two quarters.
Single-state, single-brand houses. Ninety days of silence there is most of a selling season.

Loosen it where the whole line is seasonal or allocated. The distributor-level column will not trip on a winter beer as long as something else is moving — but run the same check at distributor-and-SKU level and a beer that ships once a year fails it every spring, when it should not. Either exclude those SKUs or run them on their own calendar.

Stopped ordering, or stopped reporting

Before you pick up the phone, work out which of two very different things the blank means. A distributor that stopped ordering has a commercial problem: the brand slipped out of the book, a competitor took the slot, the buyer left. A distributor that stopped reporting has an administrative one: a feed changed, ownership changed, SKUs were recoded, your items landed under a supplier number nobody mapped.

Two kinds of silence
A distributor that stopped ordering is a sales problem. A distributor that stopped reporting is a data problem. They produce an identical blank in the file — and calling the second one as though it were the first is how you lose standing with a house you still need.

Three cheap tests separate them before anyone dials:

Look at the shape of the stop. Commercial decline tapers across periods. A reporting break tends to start clean on a period boundary and take every SKU with it at once.
Look at your own shipments. If you invoiced that distributor recently and nothing has depleted since, the cases are sitting in their warehouse — an inventory problem with a date on it, not a lost brand.
Look at whether anything is reporting from that house at all. If the whole distributor is blank across your entire portfolio on the same day, suspect a reporting change at that house before you suspect the market.

The week you find one

Assume you have confirmed it on a second file and it is a genuine stop. The first week is mostly about arriving with the right question.

01Check whether the ordering contact changed. Turnover is a common cause and the cheapest to rule out — the person who reordered you every few weeks left, and nobody inherited the line. Rule that out before you look for anything more dramatic.
02Check for an inventory build. Pull what you shipped that distributor over the same window. Stock on hand and no depletions is a work-with problem, not a listing problem, and it changes the whole conversation.
03Pull the accounts behind the distributor. Which of them were taking cases twelve months ago, and how many? That list is what the call is actually about.
04Call. A named person, a real phone, and one question: what changed?
05Email afterwards — a short recap with the account list attached, so there is a written record and something they can forward internally.

The order matters more than it looks. An email about a quiet distributor reads as a complaint, gets forwarded sideways and answered next week. A call gets you the reason in a couple of minutes, and the reason is usually mundane: a person, a pallet, or a code. Send the email second, as the record of what you agreed — never as the opening move.

If the answer turns out to be that the brand is genuinely out of the book, what you can do next depends on where you are. Distributor termination and transfer rules vary considerably from state to state. That is a question for your own counsel, not for a report.

The check nobody has to remember

The column is honestly a ten-minute job the first time. The difficulty is the second, fourth and eleventh time. It is the report that gets skipped in the busy quarter — reliably the quarter it would have caught something — and the history it needs reaches back further than any single export window holds.

Sightglass reads the VIP depletion feed nightly and keeps every snapshot, so last-sale dates survive past any one file. Quiet distributors surface in the leader view with the accounts behind them already attached, and the outreach drafted both ways — to the distributor and to the account. Nothing sends until a person approves it, and approved mail goes from the rep's own mailbox, so replies come back to them. Your reporting stays exactly where it is.

Want to see it on your book? Twenty minutes, screen shared. No file required. Book a 1-1 demo