"A reliable person replies quickly"

You reply fast because that's what reliable looks like. People know they can count on you. If someone needs something, they don't have to wait until morning and wonder whether you saw it. That reputation took years to build, and every fast reply is a small deposit in it.

Look at what the reputation is actually for. Reliability is doing what you said you'd do, by when you said you'd do it. It isn't doing everything the moment it's asked. A colleague who always delivers on Thursday when they said Thursday is reliable. A colleague who answers at ten at night and then misses Thursday isn't. Speed and reliability get run together because speed is easy to see and reliability takes weeks to show, and in the moment the buzz arrives, speed is the only one on offer.

There's research on the gap between what receivers think is expected and what senders actually expect. Laura Giurge and Vanessa Bohns ran eight pre-registered studies with just over four thousand participants, looking at non-urgent work emails sent outside normal working hours. They found what they call the email urgency bias. People receiving such an email assumed the sender wanted a faster reply than the sender, asked separately, actually did. The receivers were putting themselves under a pressure the senders hadn't applied. And the bias wasn't harmless. In their first study, with public sector employees, the people who overestimated what was expected of them reported more stress and lower wellbeing.

So the fast reply isn't the deposit you think it is. The sender wasn't checking the clock. When Priya's message arrives at ten past nine, she's already decided she'll use the old numbers if she doesn't hear back, and she'll be mildly surprised to get an answer, and mildly pleased, and she'll adjust. What she adjusts is her sense of when you're available, which is chapter 13. What she doesn't adjust is her sense of whether you're reliable, because that was never in question at ten past nine.

There's an honest objection. Some senders do expect a fast reply, and some workplaces do run on it. Chapter 9 takes that on. For now, notice that you've never asked. You've assumed the expectation and met it, and the assuming is the part the research is about.

The practical implication is a question. The next time a buzz arrives, before you answer, ask what you said you'd do and when. If the answer is "the deck by Thursday", the deck by Thursday is the reliable thing, and ten past nine has nothing to do with it.