Praxus Research

Current note ·

Price a contractor’s backlog job by job

September construction releases show rising input costs and nonresidential spending growth concentrated in data centers. Acquisition terms should follow the unfinished jobs, not the sector headline.

Current note

ConstructionBacklogDeal structure

Praxus Research
3 min read

Research cutoff: September 14, 2026. Observation periods are identified in the text.

Prepared . Published .

A highway interchange from above at night.

For an owner buying or selling a contractor this autumn, a large backlog deserves a closer reading than a sector-growth headline. ABC's September 1 analysis found that July nonresidential construction spending rose 0.1%; chief economist Anirban Basu attributed the entire increase to data centers.[24] Its September 10 materials report showed nonresidential input prices rising 1.2% in August and 8.8% over the year.[28] Our transaction conclusion is specific: price the earnings still available in unfinished jobs, and decide explicitly who carries their remaining cost risk.

The sector average hides different businesses

ABC's August 11 release put July backlog at 8.0 months, based on a member survey conducted July 20 through August 4.[6] The split was substantial. The 88% of surveyed contractors without data-center work under contract averaged 7.5 months of backlog; the 12% with that work averaged 11.4 months.[6] ABC also said backlog among firms with $30 million to $50 million of annual revenue was at its lowest level since March 2020.[6]

These are observations from ABC members, not a census of contractors or a measure of acquisition prices. Backlog measures work under contract, while ABC's separate confidence indicator measures expectations for the next six months.[6] A positive confidence reading therefore should not be presented as realized growth. This note uses July backlog and spending because those are the latest corresponding releases we could verify; the newer materials figures cover August.[6][24][28]

The September spending analysis reinforces the distinction between a general contractor and a business with specific exposure to large projects. Basu said nonresidential spending excluding data centers had fallen for a second month.[24] That does not establish that every contractor outside data centers is shrinking. It does make a broad construction-growth assumption a poor substitute for the target's actual job schedule.

Re-estimate the unfinished work

The September 10 price report is relevant to that schedule. It describes an industry input-price index, not the price increase on a particular company's remaining purchases.[28] Do not mark down every project's margin by the index. Ask management to reconcile each material job's remaining contract revenue with current supplier commitments, labor requirements and estimated cost to complete.

Separate signed scope from work awaiting authorization. Identify materials already bought, prices locked by enforceable orders, and purchases still exposed to repricing. Review escalation clauses and unapproved change orders with counsel. If the seller's forecast assumes a customer will reimburse higher costs, show the contractual basis and the collection timing rather than burying the assumption in gross margin.

For a contractor concentrated in data centers, examine the schedule by customer and project phase. Test what happens if one large project pauses while the company is carrying dedicated crews or committed equipment. For a contractor without that exposure, build the base case from its own awards and bid conversion rather than borrowing the strongest segment's growth story. Neither approach assumes the current split will persist indefinitely.

Put the disputed risk into the terms

Where the parties disagree about identifiable jobs, consider a narrowly defined contingent payment tied to their collected margin rather than an earnout based on total company revenue. This is a structuring option, not a claim about prevailing market practice. Counsel and accountants would need to define job costs, shared overhead, buyer-controlled changes and the seller's information rights before either side could assess it.

Keep ordinary working capital and any specific job-loss provision distinct, with an explicit reconciliation to avoid counting the same exposure twice. The seller should understand which unfinished jobs support cash paid at closing and which remain contingent. The buyer should know how much funding those jobs require before collections arrive. That is a more useful basis for an autumn contractor transaction than assuming all backlog converts into the same earnings.

More notes

All notes

No. 07

When the strategic comes to the table

M&AStrategic buyers

Praxus Research
3 min read

Read the note

Corporate acquirers are the largest buyers in the market and the most selective. In 2025 they drove software M&A and paid the widest premium in a decade. By mid-2026 they had pulled back from sponsor-owned assets. A credible strategic bid changes the design of a process, and the process has to be designed for it.

No. 06

The cost of waiting for a better market

Capital raisingTiming

Praxus Research
3 min read

Read the note

Timing a raise or a sale to the rate cycle is a trade most private companies lose. The window that matters is company-specific, set by momentum, runway, and buyer attention. It rarely lines up with the macro one.

No. 05

Secondaries move to the center

Capital raisingSecondaries

Praxus Research
2 min read

Read the note

What began as a release valve for institutional portfolios has become ordinary market plumbing. Sponsors, LPs, and founders alike now treat the secondary market as a first-class source of liquidity, and structure for it from the start.

Contact

Contact us to begin the conversation.

Whether the transaction is now or three years out, it is the same firm, the same analysis and the same partners.

Contact us