How to Follow Up on a Document You Know Was Read

There is a particular awkwardness to this one. You know they opened it on Thursday, came back on Friday, and spent real time on it. They do not know that you know. And now you have to write something that is informed by all of it without being about any of it.

Most advice about follow-ups assumes you are working blind. This is about the other situation, which is harder to write well and much easier to get wrong.

The rule: never quote the telemetry

“I saw you opened it twice on Thursday” is a true sentence and the fastest available way to turn a useful signal into a conversation about surveillance. The moment you narrate the data, the subject of the exchange stops being your proposal and becomes your software.

This is not about concealment. Every document shared through Quixli carries a visible notice telling the reader that the owner can see the view — they were told, before they read a word. But being told a thing is happening and having it recited back to you are different experiences, and only one of them is comfortable. The notice is what makes this analytics rather than surveillance; not reciting it is what makes you normal to deal with.

What the signal is actually for

Two things, and neither of them is a sentence you write down anywhere.

• Timing. Which day you send. The identical message reads as attentive on the day somebody looked at your document and as needy on a day they have not thought about you in a fortnight.

• Framing. What you lead with. A reader who spent a long time on it needs a different first paragraph from one who glanced.

Everything below is an application of those two. Nothing below asks you to reveal how you knew.

Four situations and what to send

1. Opened once, briefly, and nothing since

They glanced and shelved it. This is the most common outcome and the one most often mishandled, because it feels like the moment to chase and is precisely not.

Wait. Then come back with something new rather than a reminder — a reminder asks them to feel bad about a thing they have already decided to deprioritise. A relevant example from other work, a change in scope, a piece of thinking that was not in the original: any of these is a legitimate reason to write, and none of them is “just checking in”.

Do not open with “Did you get a chance to look at the proposal?” You know the answer, and asking a question whose answer you have is the one move that makes the tracking obvious.

2. Opened once, and read properly

A single long visit is the strongest thing a single session can tell you. Somebody sat with your work.

Follow up quickly and specifically, and offer the decision rather than the summary. They have the detail already — repeating it back suggests you do not know that they read it. What they usually lack is a low-cost way to move: a short call, a named next step, a decision framed small enough to say yes to.

3. Opened, and then opened again

The return visit is the highest-intent signal available, and it is the one with the shortest shelf life. People go back to a document when it has become relevant to something happening now — a conversation, a meeting, a budget.

Move the same day if you can. And here the instinct to add information is wrong: somebody re-reading your proposal does not need more of it. Remove obstacles instead. The things that stop a read proposal are almost never the content — they are approval, sequencing, timing, and who else has to see it. Ask about those.

4. Opened by somebody new

Your document has travelled. Write for the person it travelled to, not for the person you sent it to — shorter, and leading with the decision rather than the detail. What a forwarded link can and cannot tell you covers how to read this properly, including why you should hold the conclusion loosely.

The move that works without revealing anything: offer a short call for anyone else who needs to be across it. It is a normal sentence, it is useful whether or not you are right, and it invites the new reader into the conversation without either of you naming them.

What not to change

When a proposal is read and nothing happens, the reflex is to change the price. Resist it for at least one cycle.

Nobody has rejected your pricing. Somebody has failed to act, which is a different thing with many more causes than cost. Discounting unprompted answers an objection that was never raised, and it tells the reader something worse than the discount is worth: that your first number was not real. If price genuinely is the obstacle, you will find out by asking, and then the discount buys you something.

Cadence

Two follow-ups keyed to what you can see, then a change of approach. Beyond that you are not following up, you are accruing.

The third message should not be a third follow-up. Change the channel, change the person, or close the loop — a short note saying you will stop chasing and they should come back when the timing is right. It is the message people actually reply to, and it costs you nothing you still had.

For reading the signal itself rather than writing the reply, what a second opening actually tells you is the companion piece. If the problem is that nothing was opened at all, silence is a different diagnosis entirely. And what Quixli shows you, limits included, is what all of this is written against.