What a wholesaler coordinator actually does all month

There is one person at most suppliers who knows exactly which distributor files are late, and it is not the VP of Sales. Their title is something like sales operations analyst, or wholesaler coordinator, or just "the person who does the reports." They pull the exports, fix the account names, rebuild the same pivots, and hand somebody a deck. When the number in the Monday meeting is right, it is because they made it right. When it is late, it is usually because a file is.

What follows is the month as they actually experience it — the close, the chase, the pull-and-pivot cycle, the interrupts, the decks. Then the question worth asking out loud: which parts of that month are identical every single time and could be handed to a machine, and which parts should stay exactly where they are, with the person who knows the accounts.

The month has a shape

Nobody publishes the reporting calendar, but it is as fixed as a delivery schedule. Allowing for a market or two that runs on its own clock, it goes like this.

Then it starts again. Overlaid on all of it, every week, are the interrupts.

The files that arrive late, and the ones that change

Late files are rarely carelessness. Distributors report on different cadences out of different systems, and a supplier with dozens of them is holding dozens of small dependencies. A market that just changed hands, a new distributor whose item mapping is half done, a data person who moved jobs — each one costs time somewhere. In some states there are separate filings running on their own calendar as well, and those rules vary state to state.

The harder problem is quieter: a corrected file lands and last month moves. The number already sitting on a slide is now wrong, and nobody outside the reporting seat will notice unless the coordinator says so. That is not a failing of anyone's reporting. A transaction file is a record of transactions as they were understood on the day it was produced, and restatements are a normal part of the trade.

The pull, the clean-up, and the same three pivots

This is the bulk of the hours: the least interesting part of the job to describe and the most consequential to get wrong. Scale is the reason. One importer's monthly depletion export, analyzed by Sightglass, looked like this.

Nobody reads thirty thousand rows. They summarize them, and the summarizing is where the judgement calls hide. The same bar can appear three different ways across three distributors. Chain locations arrive with store numbers that do not match the ones in the CRM. A SKU that changed package configuration mid-year will roll up wrong unless someone remembers it did. All of that lives in one person's head and, if you are lucky, in a crosswalk tab.

Six steps, performed identically, twelve times a year.

Can you get me the numbers for

The interrupts are the part nobody schedules. A rep has a chain meeting Thursday and needs twelve months by account. A brand manager wants a SKU trend before a supplier call. The GM saw something in a market and wants to know whether it is real. Each is a few minutes of pivoting wrapped in a much longer job.

None of these needs data the company does not already have. The cost is reconstruction: which file, which mapping, which definition of a case, and whether the answer has to agree with a number already on somebody's slide. Consistency is the real work, and it is invisible when it goes well.

Which four blocks to automate first

The instinct is to automate whatever annoys you most. The better test is repeatability.

By that test, four blocks come off the calendar first, in this order.

What should stay human is most of what makes the role worth having. Deciding whether a decline is real, or a re-file, or seasonality, or one large order sitting in the base period. Knowing that a chain re-routed through a different distributor, so the volume did not disappear — it moved. Calling the distributor rep to ask what actually happened at an account, which is a relationship question and always will be. And standing in front of leadership to say what the numbers mean and what the team intends to do about it. A report cannot do that and should not try.

What it looks like when it runs itself

Sightglass was built for the first three of those blocks — the pull, the normalization and the standing views. It reads the VIP depletion feed nightly — the feed you already subscribe to — and keeps every snapshot, so history outlives any single export window. Normalization and the standing views are done before anyone opens a file. Accounts that are slipping, losing SKUs or have gone quiet get surfaced with the reason attached, alongside the ones that are winning, and the outreach is drafted in the house voice, to the account and to the distributor.

Nothing sends without a person approving it, and approved mail goes from the rep's own mailbox so replies come back to them. Briefings land in Slack, an email digest, or the dashboard; leaders get portfolio, distributor, territory, lapsed and concentration views in one place. When the data lags, it stays quiet rather than flagging a gap the data cannot yet support. Your reporting stays exactly where it is.

The coordinator keeps the judgement, the relationships and the exceptions. They just stop spending the first week of every month assembling the thing they were hired to interpret.