Most contractors find new construction projects the same way: a bid board, an invitation, a tip from a rep. The problem with all three is timing. By the time a project is on a bid board, the owner has a GC and the GC is assembling a list. By the time you're invited, you're one of six. By the time the rep mentions it, so does every other rep. You're not finding projects — you're finding projects everyone else already found.
The projects worth finding announce themselves much earlier, in writing, in public records almost nobody reads. The work of getting in early isn't secret information — it's reading the timeline in order and acting before the crowd shows up. Here's the timeline, and the playbook for each stage.
The signal timeline
Every commercial project moves through the same sequence of public filings before a shovel hits dirt. Each one is a signal, and they get progressively later — and more crowded:
- Plat filing
- 6–12 mo out
- Plan review
- 3–9 mo out
- Permit issued
- Building now
- 01 ·Plat / replat filing — 6 to 12 months out. Land has to be subdivided or replatted before most projects can proceed, and that goes in front of the local Planning Commission as public record. This is a developer committing capital to a site. Almost nobody competing for the work is watching here.
- 02 ·Plan review — 3 to 9 months out. The design enters the city's plan-review queue with a use, a valuation, and an applicant attached. The project is real, the budget is forming, and the team is still being assembled. A handful of sharp BD people watch this stage.
- 03 ·Permit issued — building now. The job is happening. The bid list is closed, the GC is picked, the trades are bought out. This is where bid boards and permit lists live. Everyone is watching — which is exactly why it's too late.
Stage 1: Monitor plat filings (the earliest signal)
A plat filing is the first hard evidence a project exists. In Houston, the Planning Commission meets every other Thursday, and its agendas list every plat and replat up for a hearing — subdivision name, applicant, general location. Read those agendas and you can see, a year ahead, where capital is moving.
The manual version: pull the agenda every two weeks, cross-reference each item against the parcel and owner, flag the ones in your trade's wheelhouse and territory, and start a tracking sheet. It works. What fails is doing it every single cycle without skipping — which is the entire reason most contractors never get the early lead. The discipline beats the difficulty.
What you're looking for at this stage isn't a job to bid — it's a relationship to start. A developer who just platted 40 acres is going to need a GC, and the GC is going to need subs. Being the firm that reached out before anyone else called is worth more than being the lowest of six bids nine months later.
Stage 2: Watch the plan-review queue
Plan review is where a platted intention becomes a designed building. The application carries a project use, a valuation, and an applicant — often the architect or the owner's rep — which means it carries the two things you need most: a sense of scale and a name to call.
This is the sweet spot for most GCs and subs. It's late enough that the project is definitely real and the program is known, but early enough that the team isn't locked. If you reach the owner or architect while their drawings are still in routing, you're in the conversation before the formal bid list exists. By the time that list goes out, you want to already be on it — ideally as a name they asked for, not a name they found.
Cross-reference back to Stage 1 here: when a plan review shows up on a parcel you flagged at plat stage, that's your highest-confidence signal in the whole pipeline. The project moved from "land committed" to "building designed," and you saw both. The mechanics of pulling the underlying permit record are covered in how to pull a building permit in Houston.
Stage 3: Use permits to read the market, not to find work
By permit stage the specific job is gone — but the data still earns its keep, just for a different purpose. Permit history tells you which builders and developers are the most active in your zip codes, which ones repeat, and who's expanding into your territory. That's a target list for relationship-building, not a list of bids to chase.
Treat permits as market intelligence: the active developer you keep seeing in the permit record is the same developer whose next plat you want to catch at Stage 1. The timeline loops — today's permit-puller is tomorrow's plat filer, and now you know to watch them.
Filtering by trade and territory
Raw agendas are noise. The whole pipeline only becomes useful once you filter it down to the projects you can actually win, which comes down to two axes:
- Territory. Define the zip codes, council districts, or corridors you'll actually drive to. A perfect lead 50 miles outside your range is a distraction, not a lead.
- Trade and job size. Filter for project types and valuations that match your book. A $400K tenant finish and a $40M mixed-use tower are different businesses; chasing both dilutes your pursuit.
- Developer profile. Some owners self-perform, some always competitively bid, some negotiate with a short list. Knowing which is which — from permit history — tells you which early signals are worth acting on.
Once the filter is set, the workflow is simple: every cycle, a short list of projects that match your trade and territory, sorted by how early in the timeline they are, each with an owner to contact. To pressure-test the idea on a single address before you build the whole system, run Sightline — our free tool that pulls the formation signals and parcel context for any Houston address.
Reaching the GC while the list is open
Finding the project early only matters if you act early. The move at plat or plan-review stage isn't "please add me to the bid list" — there often isn't one yet. It's a specific, informed outreach: you know the parcel, you know the program, you know the owner did similar work before, and you're introducing your firm as the right fit before they've started looking. That's a different conversation than competing on price at the back of the line.
Concretely: a short, specific note to the owner or architect referencing the actual project, a portfolio of comparable work, and an ask to be considered when the team comes together. Do that consistently, three to nine months ahead, and you stop bidding against six firms and start getting invited as a name they already trust. The full method — channels, sequencing, what converts — is in how to get construction leads.
The takeaway
The bid list isn't where you find projects — it's where the projects you should have found months ago go to get competed down to a price. Read the timeline in order: plats a year out, plan reviews mid-flight, permits as market intelligence. Filter to your trade and territory. Reach out while the list is still open. Do that every cycle and your pipeline stops depending on who happened to invite you.
You can run the whole playbook by hand — the records are public. The reason most firms don't is that it's several hours of reading agendas every two weeks, forever, and the consistency is what breaks. If you'd rather have the pipeline read for you, see what that looks like and what it costs on the pricing page.