For Plumbing and PVF distributors
Plumbing and PVF: the bid is a document and the cost is a stream.
A project price given three weeks ago against a cost that has since moved twice is still the number the customer is holding you to, and the first place it shows up is a margin report after the job is invoiced.
A working session on your numbers, not a slide deck. No account, no card, and nothing connected until you say so.
Four things that go wrong on a plumbing and PVF book
Named rather than categorized. “Margin erosion” is a heading a reader cannot check against their own week; each of these is something that happened on a Tuesday, and if none of them is familiar then this is the wrong page and the rest of it will not improve.
Problem one
A bid is a price with a shelf life nobody set
Project work is quoted once and drawn down for months. Between the bid and the release your purchase cost moves — sometimes more than once — and nothing in the process compares the two. There is usually a validity clause, and it is usually not enforced, because enforcing it means reopening a price with a customer who is mid-job. The clause is also not really the point. Even a bid that could be reopened needs somebody to notice that it should be, and noticing means comparing a document written in April against a cost that moved in June.
- The bid is a single document with a single set of numbers
- The releases are spread across months you cannot choose
- The validity clause exists and is rarely invoked
- Noticing means comparing April's document against June's cost
Problem two
Commodity pipe reprices faster than any synced cost
Even a nightly sync is reporting yesterday's cost against a number that was given in April. On the commodity half of this book the movement is not noise around a mean — it trends, for months, and a bid priced at the start of a run is wrong by the end of it in a direction that was predictable to everybody except the quoting process. The direction is often known inside the business, too, which is the frustrating part: the purchasing side can frequently see a run coming, and there is no path by which that reaches a bid already in the market.
- The quote is a document; the cost is a stream
- A cost snapshot ages from the moment it is taken
- Nothing re-reads the pair after the bid is submitted
- Purchasing can often see a run coming; the bid cannot
Problem three
The engineered half carries the spread and hides in the total
Valves and specialties earn what commodity pipe does not, so an account quietly moving that half elsewhere matters far more than its share of revenue suggests. A customer total that looks flat can be a mix that has gone badly wrong — same money, different money. So the account that looks stable is sometimes the one to worry about, and the one whose revenue dipped is sometimes perfectly healthy. A total cannot tell those two apart, and a total is what gets reviewed.
- Revenue can hold steady while margin mix collapses
- A per-customer total cannot show a mix shift
- Decline needs to be read per item, not per account
- The account that looks stable is sometimes the one to worry about
Problem four
Within one book, two different honest answers
The commodity half is bought direct, and its invoiced cost is the real cost. The branded and engineered half is claimed back afterwards, and its invoiced cost is not. Which half a line belongs to decides whether a floor computed from it means anything — and no field in your ERP says which half a line is. Which makes any single book-wide statement about margin discipline wrong in both directions at once — too strict on the branded lines, and credible enough on the commodity ones that nobody goes back and questions the branded ones.
- Both halves are on the same invoices, in the same book
- The floor is trustworthy on one of them and not the other
- Nothing labels which, so a book-wide claim is wrong either way
- A book-wide margin claim is wrong in both directions at once
What it costs is the way it accumulates
A bid does not go wrong all at once. It goes wrong in the middle, and it is drawn down anyway.
How it compounds
The release schedule is what makes this expensive. A price agreed in April is drawn against in May, July and October, each release invoiced at the April number against whatever the material cost that month. One bid is not one mistake; it is a mistake with a quantity attached, and the quantity is the customer's schedule rather than yours. Meanwhile the next bid is priced from the last one's numbers, so the drift is carried forward as an input rather than corrected as an error.
Why nothing catches it
Nothing objects at any point along that sequence. The bid was competitive, the releases shipped on time, the invoices went out at the agreed price, and the whole thing reads as a job that went to plan. It closes as a margin figure at month end, aggregated with counter business that had nothing to do with it.
No figure on this page is a claim about your book. There are no industry averages here and no recovered-margin percentages: the only numbers on the page are on the card further down, and they are sample figures re-derived by a test from the operands the card itself shows.
What a plumbing and PVF desk does about it today
Not a list of mistakes. Every one of these is a sensible response to a real constraint, which is exactly why it has survived — and the last one is usually the most rational of them. A product that does not understand why the current answer is the current answer has nothing useful to replace it with.
A cost snapshot taken at bid time
Correct on the day, saved in the bid file, and never compared to anything again. It is the right input to the original decision and no input at all to the four releases that follow.
A validity clause nobody enforces
Thirty days, printed on the quotation. Invoking it means reopening a price with a customer who has already won the job on the back of it, so in practice it is a negotiating position rather than a control.
A month-end margin report
Real numbers over a closed period, mixing project releases with counter business. It can tell you the month was worse than the last one and not which bid did it.
Nothing, between bid and release
The honest baseline. For the months in which the exposure actually accumulates there is usually no process at all, because there is no moment in the workflow that would naturally hold one.
Where PIE intervenes, and at which moment
Every figure below is deterministic arithmetic over rows your ERP already wrote, stamped with the version of the margin policy that judged it. The AI on this platform reads those numbers and phrases them; it never produces one. What is not listed here does not apply to a plumbing and PVF book, and the limits further down say which of those are permanent.
The pairs where price stopped following cost
The pass-through detector reads your own purchase and invoice history and names the customer-and-item pairs where your cost rose and your price did not. On a commodity book that is the whole game, and it is computed from rows your ERP already wrote — which means it works on bids raised outside PIE, in the months while they are still being drawn against.
Which accounts stopped buying the engineered half
Decline and dormancy are computed per customer and per item, and need no cost at all to be right. A mix shift that a customer total cannot show — the valves going elsewhere while the pipe stays — is visible as the items that stopped appearing.
A floor on the line, where the cost is trustworthy
Where you buy direct and the invoiced cost is the real cost, the floor check works exactly as it does anywhere: applied at the moment the line is priced, with a breach held for a named approver rather than sent. On the branded half, where purchases are claimed back, it is not trustworthy and the page says so rather than averaging the two.
The enquiry, resolved against your own book
Paste a takeoff, a schedule or an emailed list and PIE reads it into lines against the catalog you uploaded — exact match first, nearest neighbour over the description otherwise, deterministically, with no model asked. PVF sizing and schedule notation is structured enough that the exact match usually carries it. It finds the line in your own book and proposes no substitute from another manufacturer's.
One quote line, start to finish
One item on one bid, from April to October. The card carries sample figures for the floor half; steps two and three are the half that works on a bid PIE never saw.
2 in SCH40 A53B ERW x 21 ft
200 units for A repeat account in Ohio, quoted at $41.20 — routed to a sales manager, no individual owner. The platform holds it, not the salesperson.
| Quoted unit pricethis quote | $41.20 |
| Effective unit costbill | $38.08 |
| Margin flooryour margin policy | $44.80 |
| Recommendedthis customer’s own history | $48.70 |
| Margin at this pricecomputed | 7.6% |
The manager’s decision card, as the product draws it — with sample figures, because no customer’s numbers belong on a public page.
April: the bid is priced, outside PIE
Whether PIE sees the bid itself depends on the book it sits in: quotes are imported from Zoho Books, Business Central, Acumatica and NetSuite, and not yet from Prophet 21 or Sage. What is visible on every book is the purchase history behind the item and what this customer has paid for it before — which is what the bid should have been priced against either way.
July: your cost moves and the detector notices
The pass-through detector reads your own purchase and invoice rows and names this customer-and-item pair: cost up, price flat. It is deterministic arithmetic against a threshold you set, and it opens into the two series it compared.
July, same afternoon: somebody knows which job
The pair names the customer and the item, so the job is identifiable from your own records. Whether the contract can be reopened is a commercial question; knowing before October rather than after is the part that was missing.
October: the next release is quoted inside PIE
Now the floor applies. Cost, divided by one minus your margin floor, on the AP-invoice cost as it stands in October rather than April. On the card the asked-for price is below it, and the line is held for a named approver with the rule named.
And the next bid is priced from a record, not a memory
The held line, the operands and the policy version are one row. The commodity half of that record is trustworthy; the branded half is not, for the reason the limits below give, and PIE does not average the two into one comfortable number.
Sample figures throughout. No customer’s numbers appear on a public page, every figure on the card is re-derived by a test from the same operands it shows, and the floor is cost divided by one minus the margin floor your policy sets — nothing rounded in your favor, because a floor rounded down is a floor that has moved.
The limits, before you find them in week three
Every item below is a thing this platform does not do on a plumbing and PVF book today. They are on the page because a capability list is not what a distributor who has survived one implementation believes — and because the alternative is you discovering them after a pilot.
- Special pricing agreements and manufacturer rebates are not modelled. On the branded and engineered half of this book, where purchases are claimed back afterwards, the floor PIE computes sits above your real one and will hold profitable lines. The commodity half, bought direct, is unaffected — but which half you mostly are decides how much of this product works for you.
- PIE holds no commodity index and no cost forecast. It reports what your purchase history says happened, not what pipe is about to do.
- Quotes are not imported from any ERP, so a bid's own history lives outside PIE until you start quoting here — and on a book of long-lived bids that gap matters more than most.
- Nothing resolves until a price list has been uploaded and its decoding confirmed by a person.
- Project and job costing are not modelled. PIE prices and checks lines; it does not track a job.
Why it can be trusted
The AI never computes a single number.
Your business data and your rules determine the number. AI explains it.
Same inputs, same answer, every time — with a paper trail.
- Deterministic calculations — every figure is arithmetic on your own records; turn AI off and every number still works
- Auditable decisions — each number names the policy that produced it, so a price you quoted last quarter still explains itself
- Your data — read from your own books, used for you alone, exportable and erasable on request; AI runs on your own account
- Your rules — you set the floors and thresholds, and PIE holds every quote to them
- Provenance — every row carries the connector, the connection and the id it came from, so nothing in PIE claims a source it did not have
What plumbing and PVF distributors ask first
Our bids stay open for months. Does PIE track a quote over time?
Partly, and which part depends on your ERP. PIE imports the quotes themselves from Zoho Books, Business Central, Acumatica and NetSuite — so a bid raised in one of those books is visible, with its lines, and counts in what was offered against what was won. It does not yet read them from Prophet 21 or Sage, where a bid stays invisible until you quote inside PIE. What PIE does not do on any book is judge how a bid ended from a status word its ERP wrote: only Zoho's own vocabulary is read that way, so elsewhere a quote reads as undecided until somebody records the outcome. On every book it reads the invoices that came out of the bid and tells you where the realized margin went.
Does PIE know what pipe is going to cost next month?
No, and it will not guess. It reports what your own purchase history says your cost did, and names the customer-and-item pairs where your price did not follow. A forecast presented with the authority of a measurement is the thing this platform is built not to produce.
How much of this works if we are mostly SPA-priced?
The detection half works fully; the floor does not. Where cost is claimed back afterwards the computed floor is above your real one. If your book is mostly branded and SPA-driven, read /industries/electrical — the trade differs and the caveat is identical.
Can a salesperson see cost or margin?
No. Those fields are absent from the response rather than hidden by a screen, and a rule whose boundary is cost is withheld with them.
Which of the seven this trade runs
Prophet 21 is a core Epicor vertical here, as it is in electrical, and PVF is named separately in the same list. The other systems this trade commonly runs — Eclipse, DDI — are not among the seven PIE reads: where the book cannot be read, none of this applies, and that is a gate rather than something a page can disclose its way past.
Your book already knows where the margin went.
30–45 minutes, your own numbers on the screen, and an honest answer about what PIE can and cannot see on a plumbing and PVF book.