Naics 5415 Industry Data: Dataset Structure and Field Coverage

NAICS 5415 is the six-digit North American Industry Classification System code for computer systems design and related services - the bucket that holds custom software development, IT consulting, and the systems integration work sold under IT services contracts. It is the standard key analysts use to isolate the IT consulting economy inside government business statistics, because every Census business survey tabulates results against it directly.

One source anchors the series: the U.S. Census Bureau's Annual Integrated Economic Survey (AIES), paired with the archived Service Annual Survey tables it replaced. AIES took over from seven separate annual business surveys in 2024 and reports annual revenue, payroll and employment by six-digit NAICS industry - NAICS 5415 included, from reference year 2023 onward.

The companion archive fills the gap AIES cannot: the Service Annual Survey tables cover roughly 2012 through 2022, keeping the pre-consolidation years comparable with the new series.

The stitching problem defines this dataset more than any other. AIES begins with reference year 2023, so a long-run NAICS 5415 chart needs two older sources joined underneath it:

• Census API time series: basic summary statistics from 1992 through 2023 - a 32-year spine covering the dot-com build-out, the offshoring wave and the cloud era.
• Archived Service Annual Survey tables: approximately 2012-2022, the program AIES replaced when it consolidated seven annual business surveys in 2024.

A practical assembly reads top-down: AIES rows for 2023 onward, SAS tables for the 2012-2022 bridge, API history for anything before 2012. Because the three come from the same publisher against the same classification, the joins hold at the NAICS-code level - but the survey redesign means definitions and imputation methods shift at the 2022/2023 boundary, which is exactly where the CV columns earn their place in any published figure.

The Census series sizes the industry; two private sources price and populate it, and both index public or disclosed records rather than proprietary surveys.

H1BData.info mirrors more than 4.8 million Department of Labor Labor Condition Application disclosures spanning October 2013 through September 2025, with employer, job title, salary and worksite on each row. That makes it the deepest wage source in the vertical regardless of how you slice it: filter by employer to see what individual IT staffing and consulting sponsors pay specific titles, or aggregate across employers to benchmark a role against the market. Each LCA is a filing employers must make before hiring H-1B workers, so the underlying facts are US Government public disclosure data even though the site itself is a private, ad-supported mirror. See the labor condition application (LCA) entry for the filing mechanics and the H-1B disclosure records page for the record type.

Yes - the demand-side counterpart lives in the Data.gov Catalog's 'information technology services' search. The federated catalog holds 552,271 datasets harvested daily from federal agencies plus state, city and county publishers, and the IT services query surfaces federal IT spending, procurement and workforce records behind a cursor-paginated DCAT-US JSON search API. Records ship with file-level distributions, and robots.txt sets a 10-second crawl delay for polite harvesting.

The four datasets assemble in a fixed order, cheapest dependency first: