CostCandor

= published prices × your house

Four contractors, one roof, one day

Published 2026-08-05 · records retrieved 2026-08-05 · our band fixed as of 2026-07 · how we build these numbers

Run your own numbers in the roof replacement calculator →

Our quote checklist tells you to collect at least three bids in writing on the same specification. That advice has sat here with no documentary evidence under it. Here is the first piece: a one-page tabulation from a Texas school district recording what four contractors put in sealed envelopes for one school roof in 2019. The highest is 2.39 times the lowest — and the rest of this page is what the document cannot tell you, which is most of what you would want to know.

1. What this record is

A bid tabulation published by Austin Independent School District for solicitation 19CSP098, "Roof Replacement and Improvements at Kealing Middle School." One page, four data rows, five columns: Offeror's Company Name, Base Proposal Amount, Proposal Bond, Addenda Received, Total Contract. The header reads COMPETITIVE SEALED PROPOSAL AND EVALUATION TABULATION — the Texas competitive sealed proposal route rather than a low-bid letting, a distinction that does real work below. The opening is dated February 5, 2019. The form's own label for that date reads "Big Opening Date" [sic] — the printed text, not a slip in transcription.

The record

Austin Independent School District, Contracts & Procurement — "Bid Tabulation for 19CSP098 Roof Replacement and Improvements at Kealing Middle School," 19CSP098-Bid-Tab.pdf. One page, SHA-256 4314df181c22105e269d7ed1dbf7ec4ada892a371da114f054d0047a0fb2c48a, retrieved 2026-08-05 (HTTP 200). License unconfirmed — the district publishes no terms-of-use or copyright page we could find and the PDF carries no reuse notice, so we quote the figures and link to the original rather than reproduce the page. If the link dies, the digest tells you whether the file you found is the file we read.

There is no database query here — the record is a PDF, so the reproduction is a shell session. Two of these commands are worth running for what they fail to find.

# 1. fetch the tabulation, keeping the headers so you can see the 200
curl -sS -D headers.txt -o austin19CSP098.pdf -L -A "Mozilla/5.0" \
  "https://www.austinisd.org/sites/default/files/dept/contract-procurement/docs/bid-tabs/19CSP098-Bid-Tab.pdf"

# 2. confirm you are reading the same bytes we read
shasum -a 256 austin19CSP098.pdf
# 4314df181c22105e269d7ed1dbf7ec4ada892a371da114f054d0047a0fb2c48a

# 3. the table, as printed
pdftotext -layout austin19CSP098.pdf -

# 4. look for an area or quantity term. there is none to find.
pdftotext -layout austin19CSP098.pdf - | grep -inE "sq|squar|foot|feet|area|unit|allowance|SF"

# 5. and the whole record is the one page you just read
pdfinfo austin19CSP098.pdf          # Pages: 1

2. What it measures — and what it quietly does not

This decides whether the numbers below are worth anything, so it comes first.

  • There is no area in it. No square count, no dimensions, no quantity takeoff — command 4 turns up nothing of the kind anywhere in the one page. A cost per square foot cannot be derived here. Looking up the school's roof area elsewhere and dividing would describe that other source, not this document. Even a public bid tabulation does not hand you a unit price.
  • It does not record who won. None of the five columns names a winner: no award column, no evaluation score, no recommendation, no final contract amount — despite "evaluation" sitting in the header. Under the competitive sealed proposal route the district may weigh factors besides price and may negotiate after selecting, so the lowest figure here is not "the winning bid."
  • Opening amounts are not contract amounts. What is printed is what was in the envelope on February 5. Negotiation after selection is permitted and change orders are invisible here, so treating the low figure as money actually spent reads a number the document does not contain.
  • The scope is one line long. "Roof Replacement and Improvements" is the entire description — which improvements, which roofing system, tear-off or recover, how many sections, none of it stated. These are prices for something we cannot fully describe.
  • The conditions are not certified identical. Three rows record addenda "1,2" as received; one records "NA". The form does not explain the difference and we will not guess at it. Without this line, "2.39 times apart under identical conditions" is a stronger claim than the paper supports.
  • It is 2019, and it is commercial. These prices are seven and a half years old, from public school work carrying proposal bonds — all four rows read Yes — and school-calendar scheduling. Residential re-roofing is a different market.

3. The cohort: four rows, no filters

n = 4, the entire table. Every printed row is in the numbers below; nothing was excluded — in a record this small, an exclusion rule would be indistinguishable from choosing a result. One opening, one building, one solicitation.

We read the Base Proposal Amount column. In all four rows Total Contract carries the identical figure to the cent, so nothing depends on which money column is used. The document lists the offerors in the order Longhorn, Tecta, Port, WCR; the ascending sort below is ours. And the limit no filtering can repair: four proposals for one building on one morning is an anecdote with a paper trail — it shows a spread this wide occurs, not how often.

4. The distribution, and why our band cannot go on the same axis

Offeror, as printedBase proposalBondAddenda received
Longhorn Commercial Roofing$225,420.00Yes1,2
WCR Construction$345,291.00Yes1,2
Tecta America$360,114.00Yes1,2
Port Enterprises$538,585.00YesNA

Those four figures are printed in the document. Everything in the next table is ours, derived from them — the tabulation carries no summary statistics of any kind.

Derived from the four amountsValue
Lowest$225,420.00
Highest$538,585.00
Spread, highest minus lowest$313,165.00
Highest divided by lowest2.3893× (138.93% above)
Median of the four$352,702.50
Mean of the four$367,352.50
Lowest as a share of the median63.9%
Highest as a share of the median152.7%

Our roof replacement calculator prices in dollars per square — 100 square feet of roof surface — publishing, for instance, $450 to $750 per square installed for architectural asphalt on data fixed as of 2026-07. Setting these bids beside that band requires this roof's square count, and the record has none. No shared unit, no comparison in either direction: this document cannot confirm our band and cannot contradict it. Nor should the two be blended informally — a reader who left thinking these totals describe a house re-roofing would have learned something false from a true document.

5. Verdict

Verdict

N/A — not a band check

The record has no area, so it yields no unit price; without one there is nothing to set against a per-square band. It is also a different market and a different decade. We publish it because it does something our own bands cannot: it is documentary evidence, from an outside party with no interest in this site, that four contractors answering one written solicitation returned figures 2.39 times apart. The "get at least three bids" line in our checklist was an assertion. It now has one piece of paper under it — one, from a market that is not yours, which is exactly as much as we claim.

6. If you are collecting roof quotes

Four questions come straight out of the gaps in this document, each asking a bidder to supply what the public record did not.

  • "How many squares did you measure?" Area is this tabulation's central absence; the same gap in your quotes makes them incomparable too. Get it in writing from every bidder, and make two who disagree reconcile it.
  • "Which revisions is this price based on?" Three of these four rows record addenda 1 and 2 as received; the fourth reads NA, and the form never says what that stands for. Your equivalent is the drawing, walkthrough note, or phone call that changed the job after the first bidder quoted.
  • "Is this a firm price or an opening position?" The figures above are proposals, and the procedure that produced them allows negotiation afterwards. Your quote should say plainly which it is.
  • "What exactly does 'and improvements' cover?" The scope line here is four words long and nobody can price it. A quote whose scope fits on one line has the same problem.

The temptation with a spread like this is to read the lowest figure as the market price and the highest as padding. This record supports neither — it never says who was right, or who won. What it supports is narrower: where one written solicitation fixed the job and four sealed envelopes were opened on one morning, the figures still came back multiples apart. One quote tells you nothing about that range; three tell you where in it you stand.

Frequently asked questions

Do roofing bids really vary that much?

In this one public opening, yes: four sealed proposals for the same school roof, opened the same morning, ran from $225,420.00 to $538,585.00 — the highest 2.39 times the lowest. But that is four proposals for one building on one day in 2019. It shows a spread this wide can happen on a specification fixed in writing, and says nothing about how often.

Can I work out a cost per square foot from this record?

No, and that is the most useful thing about it. The tabulation contains no roof area, no square count, and no quantity takeoff — a search of the one page for area terms returns nothing. Importing the area from another source and dividing would produce a figure describing that other source, not this record.

Did the lowest bidder win the job?

The document does not say. It records four proposals at the moment the envelopes were opened: no award column, no evaluation scores, no recommendation, no final contract amount. The header describes a competitive sealed proposal, a Texas procedure that allows factors other than price to be weighed and negotiation after selection, so the lowest figure here is not automatically the contract price.

Does this apply to my house?

Not directly. This is public school construction from February 2019, carrying proposal bonds, procurement rules, and school-calendar scheduling that residential re-roofing does not. Read it as a contrast case: even where one written solicitation fixed the job and the envelopes were opened on one morning, the numbers came back multiples apart.

Cost math by CostCandor