Are we going to hit our number? Will we release $10M in deferred revenue by end of Q3 — give me a risk-adjusted read and the key dependencies.
Bottom line
Yes — risk-adjusted, you're on track to clear $10M. The model projects $12.6M in net risk-adjusted release by September 30, putting you $2.6M above target. Your 42.3% on-time execution rate drives most of the adjustment: the nominal recognition pool is $46M, but it reduces to $12.6M net after risk adjustment.
The two pools
Nominal
Risk-adjusted
Revenue recognition pool
$46.1M
$19.5M
Exposure / liability pool
$111.0M
$6.9M
Net release
−$65.0M
+$12.6M
Gap to $10M target
—
+$2.6M surplus
The nominal figures are misleading in opposite directions — don't anchor on either raw number. The risk-adjusted net is the number to use.
Key revenue dependencies — top 5 projects
Rank
Project
Entitlement
1
Project 0305a0db
$11.9M
2
Project 795f96e3
$11.4M
3
Project 205f2e03
$9.6M
4
Project fedd1f63
$7.5M
5
Project 080aeb05
$5.7M
These five projects carry the entire recognition pool. Projects 1–3 alone account for $32.9M of the $46.1M nominal pool — their milestone timelines are the single biggest variable.
Evidence
Every figure opens to its source: 5 projects, 306 exposure items, and the contracts, milestones, and pay applications behind them.
Recommendations
- Confirm milestone timelines on projects 1–3 — they carry $32.9M of the pool.
- Review the 5 material exposure items before month-end.
Keep digging
Which customers should we stop pursuing?
Three names stand out — one is a clear outlier.
Bottom line
Three customers are reliably unprofitable to pursue. Yellowstone Realty is the clearest outlier in the customer base.
The damage profile
Customer
CO / project
Pay-app accept
Overdue
Yellowstone Realty
10.5
43%
100d
+2 more flagged (Whitecap, Steel LLC) — same pattern, lower volume.
Evidence
Scored from change-order friction, pay-app acceptance, and receivables across every engagement — each figure opens to the underlying records.
Recommendations
- Decline the pending Yellowstone invitation to bid.
- Set a review gate before accepting new Whitecap or Steel LLC work.
Keep digging
Which projects are most likely to miss margin?
Three projects show the risk pattern — one is material.
Bottom line
Margin risk concentrates in three active projects. Project 3fa2c917 carries the widest gap between bid margin and the current read.
The watch list
Project
Bid margin
Current read
Primary driver
Project 3fa2c917
6.5%
3.1%
Schedule slip into premium labor
Project b81d4e02
5.8%
4.2%
Open steel buyout gap
Project 51c9aa38
7.0%
5.9%
Closeout rework
Evidence
Cost events, schedule updates, and buyout status post against each project's margin read — the list reorders itself as they land.
Recommendations
- Close the steel buyout on b81d4e02 before the price window moves.
- Review premium-labor exposure on 3fa2c917 with the PM this week.
Keep digging
Where are liabilities quietly accumulating?
Two places: LD exposure at the high end of the contract spread, and exposure items clustering against the Q3 window.
Bottom line
LD caps run 4.5–14.5% across the GMP corpus (median 8.96%), and two outlier contracts carry a disproportionate share. Separately, 306 exposure items sit against the Q3 window — 5 are material.
LD cap distribution
LD cap
Low
4.5%
Median
8.96%
High
14.5%
2 outliers flagged — both signed in the last 18 months.
Evidence
Every number traces to the clause it came from — clause-level extraction across the contract corpus, plus the exposure ledger against the Q3 window.
Recommendations
- Review the two outlier LD contracts before renewal conversations.
- Assign owners to the 5 material Q3 exposure items.
Keep digging
Why are RFP responses slowing down?
The constraint is response capacity, not inbound volume.
Bottom line
Author capacity is the bottleneck. Two authors carry 71% of recent responses, and four of six frequent authors are at or above allocation.
Queue profile
Current
RFPs in queue
14
Median age
11 days
Due within 10 days
6
Evidence
Authorship, queue age, and due dates tracked per response across the RFP corpus — the load imbalance is measured, not anecdotal.
Recommendations
- Rebalance authorship — two authors carry 71% of recent responses.
- Triage the 6 responses due within 10 days.
Keep digging
Where are we capacity constrained?
Two constraints: precon estimating firm-wide, and field supervision in the Southeast.
Bottom line
Estimating is the binding constraint — effective capacity runs 88% committed through August. Field supervision tightens in the Southeast where two project schedules overlap in July.
Constraint map
Function
Committed
Window
Precon estimating
88%
Through August
Field supervision (SE)
94%
July overlap
Project management
71%
Headroom
Evidence
Built from assignments and effective hours — adjusted for cross-work-type proficiency — not raw headcount.
Recommendations
- Sequence the two Southeast starts to break the July overlap.
- Protect estimating hours from non-bid work through August.
Keep digging
Which subcontractors become a problem if steel prices spike?
Two first: CentiMark and Bonitz.
Bottom line
Both carry steel-heavy scope and thin workforce depth on the same bids — a price spike hits two correlated failure modes at once.
The overlap
Sub
Commodity
Workforce
CentiMark
Steel-heavy scope
Thin crew depth
Bonitz
Steel-heavy scope
Single-crew reliance
Evidence
Commodity mix and crew depth cross-referenced bid-by-bid across the sub network.
Recommendations
- Pre-qualify a second roofing sub before the fall bid cycle.
- Add steel-escalation language on bids carrying either sub.
Keep digging
Which investments are actually paying off?
BIM LOD 400 is earning its keep. LEED AP has gone table-stakes — the market has normalized around it.
Bottom line
Double down on BIM LOD 400. Stop paying premiums for LEED; the market now treats it as baseline.
Investment scoreboard
Capability
Win-rate lift
Verdict
BIM LOD 400
+20pp · $132M
WINNING
LEED AP
51% → 51%
TABLE-STAKES
+4 more capabilities scored.
Evidence
Win-rate lift measured against matched pursuits — same sector, size band, and delivery method — not raw averages.
Recommendations
- Renew BIM LOD 400 staffing — it carries $132M of influenced revenue.
- Redirect LEED premium spend toward the next differentiating capability.
Keep digging
Did Healthcare GMP actually make us money?
Clearly, yes. Win rate is up 7 points since certification, with $197M in post-cert revenue.
Bottom line
Payback landed in month four. Everything beyond month four represents incremental return.
Before / after
Pre-cert
Post-cert
Win rate
54%
61%
Healthcare revenue
—
$197M
Evidence
Attribution runs pursuit-by-pursuit across every post-cert bid.
Recommendations
- Hold the Healthcare pricing premium — the win rate supports it.
- Apply the certification playbook to the next capability decision.
Keep digging
What should our operating bias be right now: growth, margin protection, or de-risking backlog?
The current read supports margin protection, with selective growth where coverage is thin.
Bottom line
Backlog coverage is adequate but concentrated; pipeline replenishment is on pace. That combination favors margin protection over expansion.
The posture read
Current
Backlog coverage
1.6× next 12 months
Pipeline replenishment
1.2× — on pace
Backlog concentration
Elevated — top 3 customers
Evidence
Backlog coverage, replenishment, and concentration recompute as positions move — a standing read, not a one-time analysis.
Recommendations
- Hold BD spend flat.
- Prioritize margin on new pursuits.
- Reduce concentration in backlog.
Keep digging