Signet A division of Circle Square Consulting

For specialty manufacturers

A machine makes chips. Software doesn't. Here is the exception.

Your instinct about software is correct and you should keep it. Put $340,000 into a machine and you know what you bought: a spindle, parts, second shift, and a resale value somebody will pay in cash. Put $340,000 into software and you get a login, an implementation partner, and dashboards three people open in the first month and nobody opens in the fourth. The exception is not about technology. It is about which lever you have left. You cannot control input costs and you cannot hire the estimator. What is left is how much of the office load runs without adding a person — margin defense, not innovation. Below is the arithmetic. Run it yourself. If it does not clear, do not do it.

The three levers, and two of them are gone

Input costs. In NAM's Manufacturers' Outlook Survey for Q2 2026 — 215 respondents, fielded May 12 to 28, 2026 — 83.1% named raw material costs their top challenge, up 25.6 points from 57.5% in Q1, and 94.4% expect further increases over the next twelve months, 59.6% above 5%. Trade uncertainty 71.8%, healthcare 76.8%, energy 72.0%. You can hedge a little and requote faster when a mill price moves. You cannot make it stop.

Labor. Same survey: attracting and retaining talent cited by 46.95%, average unfilled-job rate 4.25%. The CADDi/SME 2026 American Manufacturing Survey puts it harder — 79% call the skilled-labor shortage the single biggest challenge heading into 2026 (methodology not fully published; directional). The people you cannot hire are not only machinists. It is the estimator: the man who has quoted your kind of work for thirty years and knows which customer says 500 and orders 200. There is no queue of those people. You cannot fix a quoting bottleneck by posting the job.

Price. You know your customers. Some you can go back to. Some have a purchasing organization whose compensation depends on you not going back to them.

That leaves the office load — the only cost line still fully in your hands.

Why now and not five years ago

Your ERP is still your ERP and your customers still send POs as PDFs. What changed is that reading unstructured documents — a scanned print, a customer's spreadsheet format, a PO with terms on page two — went from a research problem to a working one. That was the hard part for twenty years.

Demand moved too. ISM's Manufacturing PMI hit 55.6% in July 2026, the highest since May 2022 and the seventh straight month of expansion, New Orders 56.7%, Employment 52.8% and the first expansion in 33 months (ISM Report on Business, July 2026). More RFQs, same office, same estimator.

The arithmetic — yours, not ours

We are not going to tell you how many hours this saves. We have not measured it in your shop, nobody has audited it anywhere, and a number we cannot show the working for in five minutes is not a number we will put in front of you.

There is no independently audited industry figure for RFQs per shop per month, for quote win rate, or for the dollar cost of slow quoting. It does not exist. Anyone handing you one is quoting a vendor's survey of that vendor's own customers. The two shop turnaround figures circulating here — Jacksonville Precision and Sweetwater — are vendor-supplied case studies for a competing product, Paperless Parts, in Modern Machine Shop, November 1, 2021. Read them as vendor claims. The durable line from that article is that manual quoting for a single part can take up to two hours — touch time, not sit time.

So fill these in yourself. You have them or can get them in a morning.

Then three questions.

What is line C worth? C × D × E × F. Gross margin on work you never quoted at all — not lost on price, lost by never entering the ring. It usually stops the conversation, because C is almost never zero and nobody has ever looked at it.

What would one more day of turnaround be worth? You cannot compute that cleanly, so do not pretend to. Do something better: go through last month's losses and count the ones where the buyer told you or implied they had already placed it. That is not a model. That is a list, with names on it.

What does the alternative cost? Price the next estimator — fully loaded salary, benefits, payroll burden, recruiter fee, and the twelve to twenty-four months before a new hire quotes your work unsupervised. Then ask whether you could find one. The honest comparison is not against zero. It is against hiring.

Set those against real prices: Agent Blueprint $4,900, Build $10,000, both $14,900, Claude Setup Connect $3,900 where the ERP is the risk, Agent Care $750 per agent per month.

What to do Monday

  1. Pull lines A and B. One morning.
  2. Read the customer names on line C out loud.
  3. Get your real F from the last four closed jobs, not from memory.
  4. Price the next estimator, fully loaded, and ask your recruiter how long it would take.
  5. If the arithmetic does not clear, do not do it. Say so and move on.

Close

The case is narrow: the office load is the only cost line you still control, the constraint there is one man's time, and most of what fills his day is reading and typing rather than judgment. The limit is equally narrow. This does not make you faster on the floor, does not reduce scrap, and does not improve on-time delivery directly. If your bottleneck is capacity at the mills, buy the mill — we will tell you so.

What we refuse. No shop floor automation, machine vision, or anything on the OT side. The agent does not price a job, commit a lead time, accept a change order, or touch a traveler in process or any quality disposition. No build without a Blueprint. No fixed fee against an ERP API we have not tested. No markup on model or API consumption — we will not make margin on your token spend while telling you we are defending yours. And no percentage saving, no ROI multiple, no payback period, not one.

Signet — AI that holds up. A division of Circle Square Consulting, Radnor, PA.

The Quote Clock — free, 30 minutes Published August 29, 2026