Now onboarding a limited number of teams Request access
This field note is not published yet. See what is live on the blog.
FIELD NOTES

An inbox is not a pipeline

Most bid teams run on a shared mailbox and a spreadsheet that is updated when someone has time. It works until it does not, and the failure is never visible while it is happening.

The mailbox is the real system of record in more industrial businesses than anyone would admit. The CRM has the opportunity. The spreadsheet has the deadline. But the actual state of the pursuit, what was asked, what was answered, what is outstanding, lives in a thread that only one person has read.

This arrangement is not stupid. It is what happens when work arrives as email and the tools available were designed for something else. It is also the thing that quietly caps how many pursuits a team can run.

An inbox tells you what arrived. It cannot tell you what is missing, and missing is what loses bids.

The four things a mailbox structurally cannot do

  1. Show what has not happened. A thread with no reply looks exactly like a thread you have not opened. Absence is the most important state in a pursuit, and it is the one state email has no way to display.
  2. Hold one truth across people. Two colleagues working the same tender have two different mailboxes and therefore two different versions of the deal. Whoever answers next answers from their own partial copy.
  3. Survive a handover. When someone is away, the pursuit is functionally paused, because the context is in their reading history rather than anywhere a colleague can reach.
  4. Produce honest counters. How many suppliers were contacted, how many replied, how many quoted. These are countable facts that nobody can count without reconstructing them by hand.

What a board changes

The change is not that email goes away. Email is how buyers and suppliers want to talk, and any system that fights that is going to lose. The change is that the thread stops being the record and becomes an input to one.

Once every request is a deal with a state, the questions that mattered all along become answerable at a glance. Which pursuits have nobody assigned. Which have had no inbound message in nine days. Which are three days from close with no usable quote. Where the work is stuck, rather than where it is loud.

RFQ-2026-0802closes in 3 days
Gearbox overhaul, two sites
8 contacted · 5 replied · 2 quoted · 1 needs you
taken by A. Bergquote 7/10 potential, one gap

Sample data. Every counter is a fact, not an impression.

Honest counters, or none

A board is only worth having if its numbers are true, which means it has to be willing to say uncomfortable things. If nobody has been contacted, the card says so. If no supplier has replied, the card says that, and it says what happens next.

The alternative, a widget that renders a zero as a blank or hides an empty section, is worse than no board at all. It converts a known problem into an invisible one, which is precisely the failure mode you adopted the board to escape.

The threshold where this stops being optional

Roughly, it is the point where the number of live pursuits exceeds what one person can recite from memory. Below that, a good bid manager genuinely is the system, and any tool is overhead.

Above it, the losses start, and they are invisible for a quarter or two because nothing announces them. The first sign is usually not a lost bid. It is a supplier mentioning, mildly, that they already sent that.

One board, one truth, honest counters

Celestix AI turns every tender and inbound RFQ into a live deal with a state, shared across the team and owned by name. Every card says what is true, including when the answer is that nothing has happened yet.

Field notes are written from work we do on live deals. All figures and documents shown are sample data. No customer, supplier, or buyer is identified.