The question shows up in every DM sales conversation sooner or later. How many times do I write before I let this one go? People want a number because a number ends the argument they are having with themselves, somewhere between not wanting to be that guy and not wanting to leave money sitting there.
The honest answer is that one number cannot cover the four very different situations people lump together. A lead who never said a word and a lead who told you they were waiting on a client payment are not the same lead, even though both are sitting in your inbox not replying.
So here is a count for each situation, the spacing that goes with it, the one message that does the most work, and what should happen to the name afterward. Because the real mistake is not following up too many times or too few. It is treating the end of a follow-up sequence as the end of the relationship.
Why no single number can answer this
A count with no clock attached means nothing. Two messages in two days reads as pestering. The same two messages over two months barely registers as contact. Any answer to how many has to come with over what stretch, or it is advice you cannot use.
Then there is the number itself. You have probably seen the claim that eighty percent of sales take five follow-ups, usually with no link. I went looking for the study behind it. Depending on who is repeating it, it gets credited to a sales trade association, to a British marketing site, or to a pair of authors, and none of those trails ends anywhere you can read a method, a sample or a date. That is not a reason to stop following up. It is a reason to stop quoting the number, including to yourself.
An average would be the wrong tool anyway. Averaging across the four situations below produces a figure that fits none of them, in the same way that an average house temperature tells you nothing useful if the oven is on.
| What they last did | What you actually know | How many more | Over what stretch |
|---|---|---|---|
| Never replied | Nothing at all | Two | A few days, then a few weeks |
| Replied, then went quiet | Their words, their problem | Three or four | Several weeks |
| Asked for time | A reason, often a date | One, then a closed question | Whatever timeline they gave |
| Said no | An actual decision | None for now | Months, and only on a real event |
They never replied: two messages, then the list
This is the pile people burn the most energy on, and it deserves the least. Either you opened the conversation and nothing came back, or they sent one message, you answered, and they vanished before saying anything of substance.
The thing to notice is that you have no information. Your third message cannot be smarter than your second, because nothing new has arrived in between. You are not following up at that point. You are repeating yourself with different punctuation.
There is also a failure mode you never see. On Instagram, a message to someone who does not follow you lands in their requests rather than their inbox, which Instagram's own Direct Messaging help pages describe plainly, and an unaccepted request looks exactly like a deliberate snub from where you are sitting. Counting to five against someone who never got your first message is not persistence. It is shouting at a wall.
Two messages is enough. The first gives them a reason to reply that is about them, not about your calendar. The second closes the loop and makes a no cheap to send. Then the name goes to the slower list, and you spend the attention somewhere it can compound.
Worth separating from this pile: people who clearly read your message and chose not to answer. That is a signal, not silence, and it is worth different messages, which we covered in leads who read and don't reply. And if the conversation never really started because you opened it from a list, you are doing something else entirely, closer to cold outreach than to following up.
They replied, then went quiet: three or four over weeks
Here the count goes up, because the quality of what you can write goes up. They told you something. Maybe what they are trying to fix, maybe when they want to start, maybe what went wrong with the last person they paid.
Each message has to spend one piece of that. This is what makes three or four reasonable and ten absurd: by the fifth, you have run out of things they gave you, and anything further is filler. Compare these two:
Just following up! Let me know if you're still interested.
You mentioned you wanted to start before your season picks up. Is that still the plan, or has it moved?
The first one could have been sent to anyone in your inbox, which is exactly how it reads on the other end. The second one could only have been sent to them, and it asks something they can answer while standing in line for coffee.
The spacing matters as much as the count. A few days for the first, a week or so for the second, two or three weeks for the third. The gaps grow because their silence is telling you something, and the sequence should reflect that you heard it.
One constraint here comes from outside the sale. Working through a long list of these in one sitting runs into platform behavior that has nothing to do with your intentions, and Meta's spam policy is written around frequency and repetitive content rather than any published number of messages. The pacing side of this is its own subject, and we went through it in how often to follow up on Instagram.
I will not pretend three or four is measured. It is a judgment from working through a lot of these, and the reasoning is the part you should keep: the count should stop where your material stops.
They asked for time, or they said no
These two get confused constantly, and they call for opposite moves.
A stated delay is not silence. They gave you a reason and usually a rough date, which means the timing is already decided and you are not choosing anything. One message on the date they named, then a closed question if they push it again. That is the whole sequence, and we went deeper into it in what to do with a lead who stalls.
A no is a decision, and the count resets to zero. Not forever, though. A no to this offer, at this price, in this month is not a no to you in general, and plenty of people who declined in March buy in November because their situation moved. What changes is the nature of the next contact. Months later, on a real event (you launched something different, they posted that they are hiring, their situation visibly changed), you open a new conversation. That is not follow-up number five. It is a first message that happens to have history behind it.
And there is one line that is not a judgment call. When someone asks you to stop contacting them, in any wording, you stop, and you make a record of it so nobody on your team walks into it six months later. That is worth more than any individual sale, and treating it as negotiable is how a personal account becomes a reported one.
The breakup message, and what makes it work
The last message in a sequence does more work than the three before it, and most people waste it. The common version is some flavor of last chance, usually with a deadline that both of you know is invented.
A closing message has two jobs. It makes saying no cheap, and it leaves the door visibly unlocked. Something like this does both:
I don't want to keep popping up in your inbox. Should I close this out, or is it still something you want to do this year?
Two things happen when you send that. People who were never going to buy answer, because you made it easy and polite to end it, and a clear no gives you your attention back. And a surprising number of people answer the other way, because being asked whether to close the file is the first thing in weeks that required a decision from them.
What to stay away from: guilt in any form, a fake deadline, and the passive aggressive classic of I guess you're not interested. All three ask the person to manage your feelings, which is not a thing a stranger will do for you. If you want more wording to work from, there is a set in what to write to a lead who left you on seen.
What "giving up" should actually mean
This is the part that is usually missing from the answer, and it is the part that makes the count much less dramatic.
Stopping follow-up should mean moving the lead to a slower channel, not deleting them. The name stays searchable, the context stays attached, and you stop spending active attention on it. In practice there are two slower channels. One is your own inbox, where the conversation is tagged and set aside with a reason. The other is whatever broadcast you already run: a newsletter, a weekly post, a community.
If that second channel is email, be aware that the rules change the moment you move someone there. The FTC's CAN-SPAM compliance guide requires you to honor an opt-out request within 10 business days, and to keep the opt-out mechanism working for at least 30 days after the message goes out. Nobody in DMs thinks about this until they import a list of people who never agreed to be on one.
There is a technical reason the slow channel cannot just be more automated DMs, too. Tools built on Meta's official business messaging tools get a limited window to write to someone after their last message: the platform's messaging policy says that for most user actions it lasts 24 hours, stretching to as much as 7 days for a first message after a click-to-message ad. A conversation you set aside in March is far outside that. Reopening it is a human act, which is also why those old threads are the ones that pay, as we argued in following up on old conversations.
The most reliable version of this is the one everyone skips. People who said no, or bought once and drifted, are usually a warmer list than anyone new, and winning back a past client tends to beat chasing a stranger for the fifth time.
What Ficiel does with the leads you set aside
Most of this falls apart for a boring reason. Nothing in a normal inbox remembers that you decided to stop at message three, or brings the name back in six weeks when something changed. The decision is sound and the memory is not.
That is the part Ficiel takes over. Once your accounts are connected, it rereads what you already have and puts the conversations where you spoke last and got nothing back in front of you, with the context attached, instead of asking you to scroll back through weeks of threads. Stopping is a state rather than a deletion: a conversation you set aside keeps its reason on the record and stays searchable, and you can push a name forward to a later date so it comes back to the top of that day's list rather than relying on yourself to remember.
Kord is Ficiel's AI assistant: it rereads your conversations, finds the leads who went quiet, drafts follow-ups in your own voice, and a human always approves before anything is sent. For the question in this article, the approval is the whole point, because only you know that message three was the last one you wanted to send to that particular person.
What it does not do is decide the count for you. There is no sequence that keeps firing on a schedule, and no cadence that escalates because a timer ran out. One message, written and read by a person, sent when that person says so. Ficiel's Essentiel plan is $117/mo, with money back within 10 days.
FAQ
What is Ficiel?
Ficiel is an AI inbox for coaches and setters: it surfaces old Instagram, WhatsApp and LinkedIn conversations worth following up, drafts the message, and nothing is sent without your approval. On this question, it mostly removes the memory problem: the conversations you stopped following up stay findable with their context, and you decide whether a name comes back.
How many times should I follow up before giving up?
It depends on what they last did. Two messages for someone who never replied, three or four spread over weeks for someone who replied then went quiet, one on the agreed date for someone who asked for time, and none for now for someone who said no. The count should stop where your material stops: when you have nothing new of theirs to write about, another message adds nothing.
When should I stop following up completely?
Permanently, only when someone asks you to stop contacting them. Every other stop is temporary. You end an active sequence, move the name to a slower channel, and the next contact comes from a real event rather than from a timer.
Does a lead who said no ever come back?
Some do, and their reasons usually have nothing to do with you: the budget showed up, the person they hired instead did not work out, their season changed. What makes the difference is that you kept the context and came back on an occasion rather than a schedule. Treat it as a new conversation, not as the next message in an old one.
What should the last follow-up say?
Ask a question they can answer in one word and make a no easy to send, for example whether you should close this out or keep it open. Skip the invented deadline and anything that sounds like guilt. The point of a closing message is to get a decision either way, because a clear no gives you your time back and some people answer the other way once something finally requires a choice.
Is it bad to follow up twice in the same week?
For one person in an active conversation, no, especially if they asked you to check back. The problem is doing it across a long list on the same day, which is where platform behavior starts mattering more than your sales judgment. Spread the work out and let the gaps grow as the silence does.
Picking the date is the easy half, and remembering it on the day is the hard one, which is why it helps to know what scheduling a WhatsApp message or Instagram DM actually does.
On WhatsApp the count runs into one extra doubt, since a block and an ordinary silence look exactly the same and the platform states that it will not tell you which.
Deciding the count per situation is one half of the job, and knowing which stage your leads actually stop at tells you which of those four piles is eating your month.
When all four situations land in the same week because a launch just closed, the count matters less than the sort, which we cover in following up every maybe.
