The Coverage Problem in Site Services

Coverage, not price, is the binding constraint

Construction and facilities work has moved to geography that no equipment fleet can economically serve. The site services industry has responded by selling harder on price, which addresses the wrong variable. Evidence from 106,671 delivered orders points to a different conclusion.

The argument in five points

1. Demand is geographically atomized. In 2025 we delivered orders in 5,908 cities. The top 25 accounted for 10.8% of volume. The remaining 5,883 markets averaged 4.1 orders each.
2. Fleet economics cannot follow that demand. Pre-positioning assets against four orders a year in a given market does not return its cost of capital. This is why coverage gaps exist and why they persist.
3. Single-provider fulfillment fails at scale. Across programs over $50,000, 62% required three or more distinct providers for sanitation alone. The median was four. The maximum observed on one site was eleven.
4. Two demand curves share one market. Most orders are small and transactional, and together they account for roughly half of revenue. A very small number of large programs account for most of the rest. They demand opposite operating models.
5. The solution is a records layer above the fleet. Coverage, rate, insurance, and service history held centrally lets work be placed with whichever provider can actually serve it, without making the buyer manage the complexity.

Demand has atomized faster than supply has followed

Utility-scale renewables, transmission, data centers, and national facilities programs share a trait that older construction did not: the work happens where the land, the interconnect, or the store is, not where the vendors are. A solar build sits in a county chosen for irradiance and transmission access. A retail portfolio has locations in towns nobody would build a service yard to reach.

Our own delivery record shows how thin this gets. In 2025 we delivered site orders in 5,908 distinct cities. Concentration is far lower than most people assume.

Derived from Detritus order records, calendar year 2025.

Why this is structural, not an execution problem

A regional provider sizes its fleet, yard, and routes against predictable local demand. Four orders a year does not justify a truck, a driver, or the inventory sitting idle between them. The provider is behaving rationally by declining to serve that market, or by pricing it as an exception when asked.

The buyer experiences this as a coverage gap and usually interprets it as a vendor quality problem. But the fleet is doing exactly what the math predicts when demand is spread this thin. Changing vendors does not change the arithmetic.

Single-provider fulfillment breaks at program scale

The coverage problem compounds when a single site is large. Peak crew counts on a utility-scale build can exceed what any one regional provider can service at the required frequency, even in a market where that provider is otherwise strong. Inventory and route capacity, not willingness, is the limit.

We examined every program in our record billing more than $50,000 and counted the distinct providers required for portable sanitation alone, excluding waste, water, and other lines.

Sample of 60 programs over $50,000. Median 4 providers, maximum 11 observed on a single site. Sanitation only.

62% of large programs required three or more providers. A buyer who contracts one provider for a program of this size is, in most cases, contracting for a shortfall that will surface at peak.

The three ways this fails in practice

• Silent under-service. The provider accepts the whole scope, then services on its own capacity rather than the agreed frequency. The gap appears as complaints from the field, not as a contract breach.
• Serial re-sourcing. The buyer discovers the shortfall and sources a second and third provider mid-program, each with its own rate negotiation, insurance review, and invoice.
• Concentration risk. Where one provider does carry the whole load, a single commercial dispute or equipment failure degrades the entire site. We have seen a provider raise its own receivables directly with a client site and threaten a service hold on active work.

Two demand curves share one market

Site services is usually discussed as one category. Our transaction record shows two distinct businesses with opposite operating requirements, and most vendors are built for only one.

Detritus order records, calendar year 2025. Bands are stated as proportions rather than dollar values.

Both curves have grown at close to the same rate over the past three years. Neither one is going away, so a vendor built for only one will always leave part of the market uncovered.

Why this matters to a buyer

A vendor optimized for thousands of small, fast orders is unlikely to hold an eighteen-month program in a thin-supply county. A vendor built for long placements rarely holds a service-level commitment across four hundred locations. Both are honest specializations.

If your organization has both profiles, and most enterprises do, then either you run two vendor programs or you engage a party whose coverage and records sit above the fleet and can serve both.

What a records layer is, and what it requires

A bigger fleet does not solve this. A maintained record does: who can actually serve a given location, at what price, with what insurance, and with what service history. The record supports four functions.

Why it compounds

Each order adds to the record. A provider paid once in a market becomes a known quantity for the next requirement there, with a rate, a radius, and a performance history attached. Coverage in a thin market is therefore a function of having transacted there before, which is why the model strengthens with volume and cannot be replicated quickly.

Our record currently holds 7,700+ provider accounts. In 2025 alone, work was placed with more than 2,400 distinct portable sanitation providers and more than 1,200 distinct waste providers.

Evidence from the delivery record

Four programs are documented in detail in accompanying past performance profiles. Each was selected because it isolates a different failure mode described above.

The pattern across all four

• Peak demand exceeded single-provider capacity in every case. Provider counts of 9, 11, 6, and 480 were required.
• Scope expanded mid-program in every case, and in each case was absorbed onto the existing agreement rather than returned to the client to source.
• Provider-side commercial problems occurred in every case and were resolved without interrupting site service.
• Deployment scaled by factors of four to twenty-two within a single program without a new sourcing cycle.

Client names, exact site locations, provider names, and contract values are withheld. Program values, exact figures, and reference contacts are available to a qualified party under mutual non-disclosure agreement and with client consent.

How to evaluate a site services partner

If the analysis above is right, the standard evaluation criteria are mis-weighted. Unit price is easy to compare and rarely the variable that determines whether a program succeeds. These nine questions test the things that do.

Coverage

1. In how many distinct cities did you deliver work last year? A count, not a claim of national coverage.
2. For a site in a county you have not served, what is your process and what is your typical time to a firm price?
3. How many distinct providers did your largest single site require, and how was service frequency maintained across them?

Accountability

4. When a scheduled service is missed, how do you find out? If the answer is that the customer calls, there is no exception detection.
5. Who owns an open exception through to resolution, and is that person named to us?
6. What happens to a provider’s record after a service failure? Absent a performance record, nothing changes.

Commercial control

7. When a provider raises a surcharge mid-program, does it appear on our invoice automatically or as a decision presented to us?
8. How is each provider charge tied to a specific delivery, service, or removal before it reaches our invoice?
9. If a provider has a receivables dispute with you, what protects our site from a service hold?

Methodology and sources

All figures are drawn from Detritus operational records rather than industry estimates or third-party research. Queries were run against the production system in August 2026.