A punch list is the running ledger of small items that need to be fixed, completed, or touched up before a residential project is fully accepted. It starts as a handful of items at substantial completion and grows into the single most likely place a project loses momentum, because the items are too small to schedule like real work and too numerous to ignore. The mechanic that keeps a punch list moving is structure: every item attached to a location, a trade, a photo, and a due date the moment it is filed.
The reason punch lists bog down is rarely the work itself. A door that does not latch is fifteen minutes for the framing sub. A paint touch-up is ten minutes. Multiply by sixty items across eight trades and the calendar problem becomes obvious: nobody is on site for fifteen-minute jobs. The punch list closes when the workflow forces items to route to the right sub the day they are filed, not the week the sub happens to come back for something else.
The three punch list cycles
On a residential build, three distinct punch lists run roughly in sequence, sometimes overlapping. They are filed by different people with different urgencies and need to be tracked separately even when the items end up routing to the same sub.
- Pre-CO contractor walk. The superintendent or project manager walks the house room by room before requesting the certificate of occupancy inspection. This is internal quality control. Items are filed by the builder against the builder’s own subs.
- CO inspector walk. The municipal building inspector walks the house against the code checklist. Failures here block the certificate of occupancy. This list is short, code-driven, and time-critical.
- Buyer walk. The buyer (or their inspector, on a custom build) walks the house at close-out. Items range from legitimate defects to aesthetic preferences that may or may not be in scope.
The buyer walk is the one that benefits the most from structure, because the buyer is the originator most likely to file vague items. “Floor looks scratched” is not actionable. “Living room hardwood near the fireplace, scratch about 3 inches long, photo attached” is.
The structure that works
Every punch item, regardless of origin, needs four attributes filed at the moment of entry, not later.
| Attribute | What it captures | Why it matters |
|---|---|---|
| Location | Room, area, or floor plan pin | The sub finds the item without a tour |
| Trade | Which sub owns the fix | Routing happens automatically |
| Photo | Visual of the issue | Removes ambiguity about what needs to happen |
| Due date | When the fix is expected | Creates accountability and allows escalation |
The discipline is at filing, not at triage. If the originator does not name a trade and attach a photo when they file the item, the item enters the queue as a problem rather than a task. Someone has to interview the originator, re-inspect the location, and add the missing fields. That is the moment most punch items stall.
Inspector punch list vs. buyer punch list
The building inspector’s punch list and the buyer’s punch list are different documents with different urgency. They share format but they do not merge.
The inspector’s list is code-driven, short, and time-critical. A failed GFCI in the kitchen blocks the certificate of occupancy. The fix is mandatory, the timeline is immediate, and there is no negotiation about scope. These items route directly to the responsible sub with a hard deadline tied to the next inspection date.
The buyer’s list is acceptance-driven, often long, and rarely time-critical in the same way. Items range from clear defects (cabinet door misaligned) to aesthetic preferences (a different paint sheen would have been better) to scope-creep attempts (could you add a hose bib here?). The PM has to triage each item before it routes to a sub: is this a defect, a change order, or out of scope entirely? Subs do not get pulled to fix a buyer’s change of mind on someone else’s clock.
The single biggest workflow win
The biggest single improvement to punch list throughput is forcing the originator to name a trade and add a photo at the moment they file the item. Not at the next walkthrough. Not when the PM gets to it. At entry.
The reason is friction asymmetry. The originator is standing in the room with the issue in front of them. Adding a photo costs five seconds. Naming a trade costs five more. After they leave the room, recovering both costs ten minutes and a return trip. Multiply by the volume of items on a sixty-item list and the workflow either runs in real time or it does not run at all.
Systems that allow vague entry (“the bathroom is wrong”) produce backlogs that compound. Systems that require structured entry produce lists that route themselves.
Routing on the day of filing
Once an item is filed with location, trade, photo, and due date, the next question is when the sub finds out about it. The answer should be the same day, not the same week.
The mechanism is push notification to the assigned sub on filing, with the item visible on a per-sub queue rather than on a shared spreadsheet the sub has to remember to check. Subs check their own queue. They do not check the builder’s spreadsheet. A workflow that depends on the sub finding the item is a workflow that produces aged items.
On a sixty-item punch list with a weekly cycle, items close in roughly four weeks because each sub gets a small batch each week and works through them during regular site visits. On the same list with a monthly cycle, items close in roughly four months because each sub waits for the cumulative batch to be worth a separate trip. The math does not change. The cadence does.
Sign-off mechanics
An item is not complete because the sub says it is complete. It is complete because the originator (or their delegate) signs off on the fix. Digital sign-off per item with timestamp is the cleanest mechanic: the sub marks the item ready for review, the originator inspects, signs, and the item closes with both timestamps preserved.
The financial lever that makes sign-off matter is payment. The sub does not get paid for the rework until the item is signed off. On a builder running retainage on the subcontract, this is straightforward: the held retainage does not release until punch is closed, and punch is closed when every item is signed off. On a builder paying subs in full each draw, the lever is weaker, which is one reason pass-through retainage on subs is worth keeping.
Buyer-side punch list management
On builds where the buyer files punch items directly through a portal, the flow needs a triage layer between the buyer and the sub. The buyer sees their list. The PM sees the buyer’s list plus a workflow to approve, reject, or reclassify each item before it routes to a sub.
Approval routes the item to the responsible trade with a due date. Rejection sends the item back to the buyer with a reason (out of scope, change order required, normal settlement). Reclassification routes to a different trade than the buyer suggested, because buyers often attribute items to the wrong trade (paint vs. drywall, cabinet vs. trim). The triage step takes a PM five minutes per ten items if the items are well-filed, and an hour per ten if they are not.
Warranty walks at 30, 60, 90, 365 days
Most residential builders run a warranty walk at 30 days post-close, often a 60 and 90 day check on specific systems, and a comprehensive 365 day walk before the one-year warranty period closes. Items filed during these walks are punch items in everything but name, and they should run on the same infrastructure.
The 365 day walk is the one most likely to produce real items, because that is when seasonal changes have stressed every system in the house. Settlement cracks, doors that stick when humidity changes, HVAC issues that only show in cooling season. A punch list system that rolls into warranty tracking natively avoids the alternative, which is starting a new spreadsheet for warranty and losing the trade attribution and photo history that already existed.
Time math on punch close
On a 60-item list with a weekly cycle, the math is roughly:
- Week 1: 15 items closed (the easy ones, fast subs hit them first)
- Week 2: 18 items closed (the meaty ones, sub returns for batched work)
- Week 3: 18 items closed (cleanup, rework on items that failed sign-off)
- Week 4: 9 items closed (the stragglers, often dependent on parts or trades)
The same list on a monthly cycle stretches to four months because the items only get touched when the sub is on site for something else, which is to say rarely. The compounding effect is brutal: as the project ages, the subs are further along in their next jobs, their willingness to come back for ten minutes diminishes, and small items that should have been a week of work consume a quarter.
Tools and what works
Floor plan markup beats spreadsheet, and structured per-item entry beats floor plan markup for everything except buyer walks. The reason floor plan markup wins on buyer walks is the buyer can stand in the room and pin the location visually without thinking about which room ID to type. The reason structured entry wins everywhere else is the data is searchable, sortable, and groupable by sub, by trade, by status.
A spreadsheet works for projects with under twenty items and one trade. It falls apart at thirty items across five trades because nothing routes automatically and the per-item history disappears into a column nobody updates. The standard failure mode is a column called “status” with values that drift between “done,” “Done,” “DONE,” and “d.”
Common pitfalls
Punch items that are actually change orders are the most common confusion. The buyer files “please add an outlet here.” That is not a punch item. It is a change order, and it needs to be priced, signed, and scheduled as new work. Routing it as a punch item produces a sub who shows up expecting fifteen minutes of touch-up and finds out they are doing an hour of electrical with no payment authority.
Punch items with no trade assignment go nowhere. They sit in a queue that no sub owns, and the PM is the only one who can route them, which means they wait for the next PM review cycle. Forcing trade attribution at filing eliminates the unassigned queue.
Items that do not get marked complete until the buyer asks again is the third pitfall. The sub fixes the item, moves on, and never marks it ready for review. The originator assumes nothing happened. Two weeks later the buyer asks why the item is still open. The fix is a default reminder: any item with a due date past today and no completion timestamp escalates to the PM automatically.
How BuilderGrid handles punch list
BuilderGrid uses floor plan markup for buyer walks (drop a pin, attach a photo, name the trade) and structured per-item entry for inspector and contractor walks. Items route to the assigned sub the same day they are filed, with notification to the sub’s queue rather than to a shared spreadsheet. Sign-off is digital and timestamped per item, with retainage release tied to punch close. Items roll automatically into 30, 60, 90, and 365 day warranty tracking, so the punch list infrastructure becomes the warranty infrastructure without a separate spreadsheet on the side.