Kind: Measured report
By Asif Hussain, Founder and CEO, SpecIndex
A rep pulls the airport list, sees a runway job in the territory, and opens the project looking for the manual. There is no manual.
There is a name, a state, a status, and nothing to read. That is the shape of airport work in this index, and it is the message of this article: the projects are counted, but the paper behind them is close to absent, and the little paper that exists arrives through federal jobs.
Facts at a glance
- On October 16, 2026, we ran search_projects with project_type set to aviation, paging at a limit of 10, and read all 264 projects returned. (From: Method note.)
- Of those 264, one project holds documents and the total documents held across the population is one. (From: Method note.)
- The status mix across the 264 is 186 completed and 78 under construction. (From: Method note.)
- The comparison in the second section uses coverage_report run with no arguments on October 16, 2026, which reports 322,371 documents searched and 62,117 projects with documents across the corpus. (From: Method note.)
The project list is real, current, and broad
The index holds about two hundred and sixty aviation projects. About eighty of them were announced since late December 2024, so the list is not a museum. California carries the largest count of any named state, followed by Texas and Florida, and the projects run from runway lighting to hangars to terminal work.
Read the statuses and the list changes character. Most of the projects are marked completed, and the rest are marked under construction. This is a record of work awarded and built, drawn largely from public award and announcement records.
It tells a rep where airport money moved. It does not, by itself, tell a rep what got specified.
The documents behind the list are missing
Of the aviation projects in the index, exactly one holds a document, and it holds exactly one document. The wider corpus is not thin this way.
Across all sectors, the corpus holds documents for more than sixty thousand projects. Aviation stands apart: a whole sector with a project list and, behind it, one runway extension file from Alaska.
So a seller who treats the airport list like the hospital or school lists will waste a week. Those lists sit on top of manuals that name products. This list sits on top of award records.
The work happened, the money moved, and the specification that decided the brands was never part of the public record this index could capture for these projects. That is a statement about the public record, not about the market.
Airports are built and products are chosen. The choosing documents are just not public here.
The airport manuals the corpus does hold are federal
We went looking for airport specifications the other way, through brand positions, and the documents surfaced on three projects. All three are federal sourced. One is a control tower in Anchorage, one is a tower roof in Kenai, and one is a base building at Montgomery Regional Airport in Alabama.
For the project "HVAC Replacement Project located at Montgomery Base Building, Montgomery Regional Airport, Montgomery, AL", the specification states "Class L fast acting BUSS or approved equal". The citation is "MGM Basebuilding HVAC Replacement Project SOW 09.23.2024.pdf", pages 104 to 167. A fuse class, a maker family, and an open door for an equal, written into an airport heating and cooling replacement.
For the project "Anchorage Airport Traffic Control Tower (ANC ATCT) and Anchorage Terminal Radar Approach Control/Base Building (A11 TRAC", the specification states "Cirrus Second Look II by Armstrong World Industries, Inc.". The citation is "Attachment J-3a ANC ATCT and TRACON Final Specs Vol 2.pdf", pages 551 to 622. The project name is reproduced exactly as the index returns it. A ceiling tile, named as the basis of design, on page 551 and beyond of a tower manual that runs to hundreds of pages.
When an airport manual does reach the public record, it reads like any other federal manual. It names products, fixes classes and series, and states where an equal may enter. The scarcity is in the count of manuals, not in their quality.
Our judgment: for airports, plan territories from the list and sell from the federal manuals
The thinness is the finding. A rep cannot run a specification strategy against a list with one document behind it, and should not try.
The list earns its keep earlier in the sale, as a map of where airport work is awarded and which owners build again. The selling, where a document exists, runs through the federal channel, because that is where the manuals are public.
What to do with airport work in this index
Use the project list for coverage. Note the owner, the state, and the rhythm of awards, and treat a repeat owner as the account. Then watch the federal postings for the same airports, because a federal manual is the version of the job you can actually read and work.
You can [search projects by state, product and stage](/projects/) to build that list. When a manual appears, work it by division while the bid window is open, the same as any federal job.
Method note
On October 16, 2026, we ran search_projects with project_type set to aviation, paging at a limit of 10, and read all 264 projects returned. Of those 264, one project holds documents and the total documents held across the population is one. That project is "Jbe054 Runway Extension, Jber, Alaska", in Alaska, marked completed, announced July 22, 2022.
The status mix across the 264 is 186 completed and 78 under construction. The announced dates run from July 28, 2011 to September 17, 2026, and 78 projects carry an announced date on or after December 25, 2024, which is 660 days before the read. The state field reads California on 29 projects, Texas on 18, Florida on 16, Virginia on 13, and Hawaii on 11, with 56 records carrying the value US in place of a state and 46 distinct values in the field in total.
For the exhibits, we ran find_brand_positions on October 16, 2026, paging at a limit of 10, and read the first 250 rows returned for Eaton, the first 250 for Siemens, the first 380 for Johnson Controls, the first 250 for Armstrong, and, widening the brand list beyond those four, the first 450 for Zurn. That is 1,580 rows read. We filtered those rows on the client side to rows whose project name contains Airport or Airfield and that carry a citation, and excluded names for bus and rail terminals.
The filter returned 11 rows covering three distinct projects: the Anchorage tower, the Kenai tower roof, and the Montgomery base building. All three projects are federal sourced, with project identifiers that carry the SAM marker. Each specified product and substitution line in this article is quoted exactly as returned, and each citation gives the project name, the document title, and the page span exactly as returned.
The comparison in the second section uses coverage_report run with no arguments on October 16, 2026, which reports 322,371 documents searched and 62,117 projects with documents across the corpus. The limit of this article is the gap it reports. The aviation population counted here and the document population behind the exhibits are different populations: the exhibits come from federal projects whose manuals the corpus holds, and those projects are not the document holders of the aviation list.
Nothing here shows that airport construction itself is small. It shows that, in the public record this index holds as of October 16, 2026, airport projects arrive with records and federal airport jobs arrive with manuals, and the two sets barely meet.