How to catch a sponsor inflating a self-reported completion percentage

A draw request shows up: 62% complete, per the sponsor's pay application. Three photos attached, concrete looks poured, steel looks up. You're supposed to sign off and release capital on that number. Nobody at the table independently measured anything.

Where the percentage actually comes from

On most construction loans, the completion percentage on a draw request traces back to an AIA G702/G703 or the sponsor's own schedule-of-values breakdown. The general contractor's project manager estimates a percentage against each line item, and the sponsor rolls it up into one topline figure for the lender or fund. Nobody along that chain has to show you an independent measurement. An inspecting architect or a third-party construction monitor might sign off, but only if the deal structure calls for one, and even their visit is often weeks behind the draw date itself.

That gap between reported and built is where inflation happens. A sponsor under pressure to hit a return hurdle, or trying to avoid tripping a covenant, has every reason to round 54% up to 62%. The photos attached to the package are supposed to back the number. In practice they often just support the vibe of progress.

What the photos can and can't tell you

Sponsor photos are selected by the sponsor. Someone on the GC's team walks the site with a phone, picks the three or four frames that look best, and attaches them to the draw request. A photo of poured footings doesn't tell you whether they were poured this month or last quarter. A wide shot of steel framing doesn't tell you whether the east wing, which the schedule of values says is 70% done, has had a crew on it in weeks.

A few things worth checking before you take the percentage at face value:

• Pull the EXIF data off the actual photo files, not a PDF screenshot. Timestamp and GPS tags get stripped the moment someone pastes an image into a draw package, so ask for the originals.
• Check the construction sequence against the claim. Rebar visible with no slab poured on top doesn't match a line item that says foundation is complete.
• Compare shadow length and direction against the claimed date and site latitude. It's a rough check, but a photo dated for July that reads like a low winter sun deserves a second look.
• Match the footprint in the photo against the recorded plat or site plan. Sponsors have been known to round up acreage or shift a building envelope when staked lot lines don't match the pro forma.

None of these catch everything, and none of them replace an independent image. They're just the fastest way to flag a package that deserves a harder look before the draw clears.

When the gut check says something's off

If the photos and the percentage don't line up, or if the sponsor's submission is the only visual record anyone on your side has ever seen, the fix isn't arguing with the GC's project manager over email. It's getting a look at the site that didn't come from the sponsor's phone. An independent before-and-after image, pulled from satellite or aerial capture rather than a visit the sponsor scheduled, confirms whether the footprint matches what's stated and whether the structure on the ground reads like 62% or like 40%.

That's the gap Site Verification is built to close: an on-demand image pair and progress read that doesn't depend on which photos the sponsor chose to send.

Before you release the next draw, it's worth pulling a look at the site from a source the sponsor doesn't control.