Why Joe sometimes tells you not to buy a robot

October 1, 2026 · 4 min read

The concierge on this site turns down projects. Not because the robots cannot do the work, but because the station, the rate or the arithmetic says the project would disappoint. Here is what it is checking, and why a dealer that says no stays in business.

The short answer

A good share of conversations on this site end with the concierge suggesting something other than a robot: a fixture, a bar feeder, a process change, a different station, or waiting a year until a part family settles down. That happens more often than our margin would prefer, and it is on purpose. The company's margin comes from the projects that go in and stay in, and the fastest way to lose a shop for good is to sell it a cell that gets parked.

What is Joe actually checking?

Five things, in roughly this order. Is this station the constraint, or is the real bottleneck two stations upstream? Does the cell keep up at your peak rate on your busiest afternoon, not at the average on the monthly report? Do parts arrive in a way a machine can pick, or do they arrive in a bin, upside down, three ways? Is the payback honest once you include the integrator, the gripper, the guarding and the risk assessment, and once you admit whether the cell runs one shift or two? And is there a simpler tool further down the ladder that solves eighty percent of the problem for a tenth of the money?

Any one of those can turn a yes into a not yet. The concierge says so in plain words and tells you which one it was.

Why would a dealer turn down a sale?

Because the alternative is worse for everyone. A cell that falls behind at peak does not fail quietly. Cases pile up, someone steps in to hand stack the overflow, and within a month you have a robot and a person doing the job the robot was bought to do. That shop never buys a second robot, and it tells its neighbors. A dealer that depends on referrals from small manufacturers cannot afford that story, so the cheapest sale we ever make is the one we talk you out of.

There is also a plainer reason. Every machine the concierge names is a candidate, not a specification, and a certified integrator still has to walk your floor and sign the risk assessment. If the honest preliminary read is that the integrator will say no, telling you that now saves you a site visit and a quote you would have had to turn down yourself.

What does a no look like in practice?

Picture a deburring station where parts come off three different machines in random order. The honest read is that the robot is fine and the part presentation is not, and that a divider and a staging fixture at the machines would make the job automatable for a few thousand dollars, with or without a robot later. Or picture a lathe that runs fourteen hours a week. The arithmetic on spindle hours says a tending cell would sit idle most of the week with the machine, and the better first robot is on the mill that is already the constraint. Both are illustrations, not client stories, and both are the kind of answer the concierge gives every week.

Neither of those is a lost sale. Both are a shop that now trusts the next answer.

When does Joe say yes?

When the station is the constraint, the peak rate is inside the class of machine with margin to spare, the parts arrive in a consistent way, the payback survives honest inputs, and nothing cheaper on the ladder solves it. Then the concierge names two or three candidates from the matrix, says plainly which looks strongest and why, and offers to have the team confirm specs and availability with the maker and check whether a certified integrator who has built that application has the interest and capacity to take it on.

How to get the most out of a no

Ask what would have to be true for it to become a yes. Usually it is one thing: a fixture, a minimum batch size, a second shift, a different station. That single sentence is often the most useful output of the whole conversation, because it tells you what to change in the plant this month so that the robot makes sense next year.

Describe the station to Joe if you want that read on your own job. It is free, it does not ask for your email before it answers, and when the answer is no, it will tell you what the yes would take.

Related on this site

A no from a seller is information. It usually points at the fixture, the upstream station or the peak rate that would have sunk the project after the check cleared.

Put it to the test on your own job

Tell Joe what you are running and he will apply this to your numbers, including telling you when a robot is the wrong answer.

More notes