Enquirer Consulting Group

Reachable Buyer Map

Prepared for Paul Lewis, Pythian · United States · August 2026
Here is the map. Data and AI work is bought by a small number of large employers, so the useful question is not how big the market is but which slice of it can fund a program without a board conversation. This page counts the United States side. The headline band is 1,000 or more people on the payroll plan, with the 250 to 999 band shown next to it, because that is where a first project is signed by one executive.
Manufacturing and industrial
The largest block in the enterprise band and the one carrying the oldest estates. Plant systems, quality data and supply data usually sit in different places with different owners, and the work arrives as a migration or a consolidation rather than as an AI project.
Who signs: CIO, VP of information technology, head of data and analytics, director of plant systems.
2,950 to 3,050
US employers at 1,000 people or more; about 6,400 more between 250 and 999
Health systems, provider groups and care operators
The densest concentration of large employers in the country and the sector where reporting obligations force the data question whether or not anyone wanted it. Long buying cycles, and the work recurs once it lands.
Who signs: CIO, chief data officer, VP of analytics, head of clinical informatics.
2,400 to 2,500
US employers at 1,000 people or more; about 4,700 more between 250 and 999
Retail and consumer
Where the payback window is short enough to defend internally, which shortens the sale. Demand data, store or channel data and customer data rarely share a home, and a new commercial leader tends to reopen that question within the first year.
Who signs: chief digital officer, VP of ecommerce, head of data, CIO.
1,200 to 1,250
US employers at 1,000 people or more; about 2,500 more between 250 and 999
Banking, insurance and capital markets
The most regulated buyer on the page and the most likely to run a formal selection. Slower to enter, harder to displace once you are in, and the only segment here where a named reference in the same sub-sector reliably shortens the next one.
Who signs: chief data officer, head of enterprise architecture, chief risk officer, CIO.
1,050 to 1,150
US employers at 1,000 people or more; about 2,200 more between 250 and 999
Information, media and technology services
The segment most likely to have strong internal engineering and still buy outside help, because the constraint is capacity and specialist depth rather than capability. Shortest evaluation of any segment on this page.
Who signs: CTO, VP of engineering, head of platform, head of data.
900 to 950
US employers at 1,000 people or more; about 2,300 more between 250 and 999
Transportation and logistics
Operationally data-rich and historically under-invested in the reporting layer, so the first project is often a consolidation rather than a model. Decisions concentrate in a small executive group, which makes the account short to map.
Who signs: CIO, VP of operations, head of business intelligence, chief operating officer.
740 to 780
US employers at 1,000 people or more; about 1,300 more between 250 and 999

Where the openings are

1
The enterprise band is finite, and a large part of it sits outside these six sectors. The segments above come to between 9,240 and 9,680 US employers at 1,000 people or more, out of about 16,800 in that band across every sector. That leaves roughly 7,100 to 7,600 outside the six, mostly in education, construction, hospitality and professional services, which is a deliberate choice about where to spend attention rather than an oversight.
2
The buyer is a seat, not a company. CIO, chief data officer, VP of analytics. Those seats turn over, and a new one almost always reopens the roster of who does this work. A channel built on named roles catches that moment. A partner-sourced channel hears about it after the decision.
3
Partner-sourced pipeline selects for wherever a field team already sits. In this category most work arrives through cloud and platform referrals, which is a real channel and a narrow one: it reaches accounts a partner already holds and nothing outside them. The accounts outside are not unqualified, they are unaware.
4
This work is bought at a moment, not on a cycle. An end-of-support date, a merger, a new CIO, a platform program that stalled. Those moments are visible from outside if someone is watching several thousand named companies, and invisible if the channel is waiting to be told.
Built from public federal registry data covering US employers that file a benefit plan, current to the 2024 filing year. Counts are banded deliberately. Headcount bands use plan participants as a proxy, so they indicate scale rather than an exact staff count. Sector codes are self-reported by the companies themselves, and a company filing under two codes is counted once. Owner-only and very small employers are not published in this data, which does not affect a map drawn at 250 people and above. It describes the market rather than your business, and there is nothing to buy at the end of it.
ENQUIRER CONSULTING GROUP