An RFI by email vs. one with a timestamp: what actually goes missing

A Request for Information — a question raised against a project, addressed to whoever can actually answer it, like an architect or an engineer — is one of the most ordinary pieces of paperwork on a construction site. Most of the time it’s handled the same ordinary way: an email, a reply somewhere in the thread, and a copy-paste of the answer into whatever record the GC is keeping, if anyone remembers to do that step at all. It works, most of the time. The times it doesn’t are the times it actually costs something.

The gap isn’t asking the question — it’s proving what came back

Nobody really struggles to send an RFI by email. The failure mode is later: a delay dispute, a defect that traces back to a drawing detail, a disagreement about whether a change was actually approved — and the answer is somewhere in an inbox, unsearchable across a project, un-timestamped in any way that survives being forwarded, and dependent on someone having manually copied it into the project file at the time. If that step got skipped — and on a busy week, it usually does for at least one RFI — the answer effectively doesn’t exist as project evidence anymore, even though someone genuinely did answer it.

The retyping step is also where wrong answers quietly enter the record. A reply gets summarised from memory, or copied with a transcription slip, and what ends up in the project file is someone’s paraphrase of what was said rather than what was actually said. It’s rarely a big enough error to notice at the time — it’s exactly the kind of small drift that only matters months later, when the RFI answer is the thing being relied on.

Why "just save the email" doesn’t really solve it

Saving or forwarding the email thread feels like it closes the gap, but it doesn’t address the two things that actually matter later: whether the answer is easy to find against the right project when you need it, and whether the timestamp on it is trustworthy. An email client’s send/receive time is real, but it’s not tied to the project record in any structured way — finding the right RFI among months of correspondence means searching by memory, not by project.

How Novato handles this

An RFI in Novato is addressed to a saved, confirmed contact — architect, engineer, client, whoever — reused from a project’s own contact list rather than retyped from memory each time, which closes off the "sent to the wrong address" failure mode before it starts. Composing the question sends it by email with a no-login reply link; whoever answers responds directly through that link, optionally attaching a file, and the answer lands back in the project record automatically, with a timestamp, the moment they submit it. Nobody copies it in by hand, so there’s no paraphrase step for an answer to drift through. Every RFI can be downloaded as a PDF with the question, who it was sent to, and the answer once received — a document that exists the same way months later as it did the day it came in.