The question behind a repeat repair is never "is this repair worth it". Almost any single repair is worth it. The question is whether the same opening is going to be back on the list next season, and that is answered by what failed, not by what it cost.
What is actually inside a window service call
Three things: diagnosis, the part, and the labour to reach the part. Diagnosis takes longer than most people expect, because the symptom is downstream of the defect. A sash that will not stay up is a balance. A slider that drags is a roller, or a track worn through. A casement that will not pull tight at the top is usually a hinge arm rather than the lock everybody blames.
The part is where a repeat call is decided. Before we quote a second visit on the same unit we write down the manufacturer, the series, the glass size and any hardware markings stamped into the balance shoe or the operator housing. A repair is priced by whether that part can still be sourced, not by how difficult it is to fit.
The five parts we replace most, and what each one tells you
- Coil and block-and-tackle balances. Double-hung and single-hung units. The sash drops, or racks in the frame. Straightforward while the series is current; a dead end once it is not, because balances are matched to a sash weight and a jamb profile, not sold by the inch.
- Rollers and track. Sliding windows and patio doors. Drag first, then a bound corner that loads the glass edge. A worn roller takes a new roller; a track worn through the aluminium does not. The new roller cuts the same groove within a season.
- Weatherstrip, pile and bulb seal. The draft that "came back". Replacement has to match the fin height and the kerf width, not just the colour. Pile that is too short seals nothing; pile that is too tall stops the sash closing square, which puts load back into the lock.
- Operators, hinge arms and locks. Casement and awning units. Handedness, arm length and stack height all matter, and the wrong operator binds the sash on the first crank. This is the group most often superseded by a redesigned part that almost fits.
- Sealed insulating glass units. Not a repair. The unit is a manufactured assembly (two lites, or panes of glass, a spacer carrying desiccant, a primary and a secondary seal), and it is replaced whole.
Which failures are a part problem and which are an assembly problem
This is the whole distinction, and it is the one a cost comparison usually skips.
A part problem is a component that wore out inside an assembly that is still square, still dry and still sealed at its perimeter. Hardware, pile, glazing bead, screens, adjustments and perimeter sealant all sit here. Fitting the part restores the unit, and the next failure is a different part on a different schedule.
An assembly problem is a frame that has moved, a sash that is deflecting, a sill or jamb that is wet behind the cladding, or a series whose hardware is out of production. Here a new part is consumed by the same defect that ate the last one. A replacement balance in a racked frame carries an uneven load from the day it goes in. A new sealed unit set into a frame that lets water sit under it fails at the edge seal again.
The practical line we use: repair while the failure is in a part; stop when the failure is in the assembly. That line is visible in a log long before it is obvious on site.
Keeping the log so the decision makes itself
Five fields, written the same way every time, are enough:
- A unit identifier tied to the building and the elevation: "north elevation, opening 4", not "the kitchen one"
- Manufacturer and series, read off the sash label or the spacer, once, on the first visit
- The symptom in the words the occupant used, before anyone interpreted it
- What was fitted, with the part number, and whether it was a supersession
- Whether that same opening has appeared on the list before
Two patterns fall out of a log kept that way. Repeats concentrated on a handful of openings are a repair problem, and repairing them is the right answer. Repeats spreading steadily across one elevation are an assembly reaching the end of its service life, and that is the point at which replacing the elevation stops being the expensive option.
On a multi-unit property the log is the bid
For an association board or a property manager the log does one more thing: it turns a maintenance line into a scope. A year of tickets sorted by building and elevation shows which elevation is going first: in Northern Nevada, almost always the one taking afternoon sun and the wind that comes with it. That lets the work be phased by elevation instead of unit by unit, which is one mobilisation and one set of resident notices rather than a dozen truck rolls.
It also settles the question that stalls boards, which is whether the money is coming out of maintenance or out of reserves. A ticket history that shows repeats concentrated in one building is a maintenance answer. One that shows repeats spreading is a capital answer, and the log is the evidence either way.
What a service visit records that a ticket cannot
Wherever a stop or a piece of trim has to come off to reach a part, we photograph the sill and the jamb behind it. The condition of the substrate under a leaking unit is the fact that decides everything downstream: whether a re-glaze is worth doing, whether the frame is sound enough to take new hardware, whether the wall is quietly getting wet. It is invisible from a ticket that says "adjusted sash, no charge".