I've looked at 500+ founder inboxes. In almost every one of them, about 22% of the conversations are real potential buyers and the other 78% are noise — compliments, pitches, free-call requests, the occasional bot. The founders who own those inboxes can't tell which conversation is which in real time, and by my math that blind spot costs them 2–4 good leads every month.
This guide teaches you how to find your 22% by hand. It's the same method behind ForesightHQ, the app I built, but you don't need the app to run it. You need a spreadsheet, about an hour for the first pass, and a willingness to stop trusting Instagram's sort order.
One thing before we start. Nothing in this guide asks you to get more leads, post more content, hire a VA, or install a chatbot. Every dollar this method finds is already sitting in conversations you've already had.
The sales doctrine you never agreed to
Somewhere along the way, the online coaching world settled on a default sales model: every lead deserves your full attention, and skipping anyone means risking the one that mattered. Nobody teaches it as a doctrine. It just gets absorbed, the same way you absorbed "reply quickly" and "be helpful," until showing up for every single DM feels like the definition of doing sales right.
I call it Chase-Everyone Sales, and the math underneath it doesn't work. If 78% of your inbox will never buy, then giving every conversation equal attention means most of your selling energy goes to people who were never going to pay you. The buyers get whatever is left over, whenever you happen to get to them.
Here's the assumption holding the whole doctrine up: that you can't tell who's ready until you've engaged with everyone. That assumption is wrong. Buying intent leaves fingerprints all over a DM thread, and once you know what they look like, you can sort your inbox by intent instead of by whoever messaged last.
That's the whole method. Three phases: Audit, Score, Scan. The rest of this guide walks you through each one.
Why you can't see your buyers right now
Open your Instagram inbox and look at how it's ordered. Newest message on top, oldest on the bottom, and the conversation that pinged your phone two minutes ago jumps the line ahead of the one that's been quietly waiting since last Tuesday.
Recency is a UX decision Meta made for engagement. You inherited it as a sales decision by accident.
The consequence shows up the same way in inbox after inbox. The hottest lead you have — the one most likely to buy this week — is statistically not at the top. She messaged eight or nine days ago, asked a specific pricing question, got your answer, then went quiet because $3,000 is a real decision and her life kept happening while she sat with it.
While she was sitting with it, a dozen newer messages stacked on top of her thread. By day four you'd have to scroll to find her, and by day nine you don't remember she exists. When she doesn't hear from you again she quietly concludes you weren't that interested in working with her — she didn't go cold so much as quiet, and your inbox buried her.
Founders almost always read this as a personal failure — bad memory, bad discipline. I'd push back on that. You were running a sales operation on an interface Meta built to maximize engagement, and the interface won.
Phase 1: The Audit
The first phase is a one-time excavation. You're going to read your own inbox as a dataset, which sounds obvious until you realize you've probably never actually done it. Most founders have only ever reacted to the top of their inbox, one notification at a time — almost nobody has read the whole thing.
Here's the process:
- Open a spreadsheet. Four columns: name, date of last message, what they last said (short version), and a blank column for the score we'll fill in during Phase 2.
- Go back 90 days in your DMs. Yes, all of it — including the Requests folder, which is where IG quietly routes messages from people who don't follow you. I've found buried buyers in there more times than I can defend.
- Log every conversation where the other person showed any interest in what you sell. They asked about your offer, replied to a story about your work, or even just mentioned the problem you solve. When in doubt, log it. Skip the obvious noise — spam, people pitching you, and chat that was never about the work.
- Mark every thread that went quiet after money came up. A pricing question that got an answer and then silence, a "let me think about it" that never resumed. These get a flag in the spreadsheet. They're the most valuable rows in the whole audit.
Plan for about an hour. It's tedious, and somewhere in the middle you will find a conversation that makes your stomach drop a little — someone who was clearly ready, asked a real question, and got buried before you answered. That moment is the entire point of the audit, because the revenue was never missing, just unsorted.
When I ran this for my first client, the audit pass surfaced 75 hot leads sitting in an inbox she checked every single day. Within 48 hours she'd recovered $7,500 from conversations she'd already had. No new ads, no new content, same follower count on Friday as on Wednesday.
Phase 2: The Score
Now you sort the spreadsheet by buying likelihood, which means you need to know what a buying signal actually looks like. Conversations aren't equal, and the cues that predict closing are more specific than "they seem interested."
These are the signals that predict a close, based on the inboxes I've audited:
- A pricing ask. Any form of it — "what's your rate," "do you have payment plans." Someone doing wallet math is past curiosity.
- Return engagement after silence. They went quiet for a week and came back on their own. Re-engagement after a gap is one of the strongest signals there is, and it's the one founders miss most because the returning thread looks like just another message.
- Specific pain, specifically described. "I had three discovery calls this month and closed zero" is a buyer describing the problem your offer solves. Vague relatability ("ugh, DMs are so much") is conversation, not intent.
- Logistics questions. Start dates, time commitment, how onboarding works, whether it's live or recorded. Nobody asks how the thing works unless part of them is already inside it.
And these are the signals that look warm but don't convert: compliments without a question attached, "can I pick your brain," generic enthusiasm, and free-resource requests dressed up as interest. Look, those people aren't useless — some of them become buyers in six months, and a few are just kind humans being kind. But they are not where your best selling energy goes today.
Score every row in your spreadsheet with a simple three-tier system. A means money or logistics entered the conversation — these close. B means specific pain or return engagement, real intent that needs developing.
C is everything else. Sort the sheet so the A's sit on top, and you're now looking at something Instagram has never once shown you: your inbox ranked by revenue instead of recency.
Your follow-up order writes itself. A-flagged threads that went quiet after pricing get a message today — something low-pressure that reopens the door without apologizing for the gap. "Hey, was thinking about your question on the payment plan — want me to walk you through how it works?" beats "sorry for the delay!" every time, because it picks the thread up where the intent was instead of making the silence the subject.
The energy math nobody taught you
Before the third phase makes sense, you need one piece of cognitive math: after 8–10 real DM conversations in a day, your brain is done making quality decisions. Decision fatigue is a measurable thing, and it takes something like 15 minutes to mentally reset between genuinely demanding conversations. You cannot willpower your way past it — I've tried, my ADHD brain has receipts.
Now put that next to Chase-Everyone Sales. If you reply in the order Instagram presents, your freshest brain goes to whoever happened to message most recently — which, per the 78%, is usually a compliment or a brain-pick. By the time an actual buyer gets your attention at 2pm, you're cooked, and you show up flat in the one conversation where flat costs you $3,000.
So you feel "bad at sales" in exactly the conversations that matter most, and you conclude you're not a natural closer. The data says otherwise: you've been spending your best cognitive hours on the 78% and serving the 22% from the bottom of the tank. Your energy is a finite sales asset, and the score from Phase 2 tells you where to spend it.
Phase 3: The Scan
The first two phases find the buried money. This phase keeps it from getting buried again, because the audit is a one-time dig and your inbox starts re-burying things the moment you close the spreadsheet.
The Scan is a daily operating rhythm with three rules:
- Open your scored list before you open Instagram. The order matters more than it sounds. If IG opens first, the notification stack sets your agenda and the whole sort collapses back to recency.
- Top-ranked conversations get your first hour. A-threads first, fresh brain, before anything new gets a reply. New messages get scored before they get answered — a quick A, B, or C as they come in, which takes seconds once the signal map is in your head.
- C-threads wait for low-energy hours. The 4pm version of you can handle "love your content!" just fine. She should never be the one handling a pricing conversation.
Run it daily if you can, every two or three days at minimum. The whole pass takes 20–30 minutes once it's a habit, and the follow-up windows that used to slip — the nine-day silences that killed deals — stop slipping, because quiet A-threads sit at the top of your list instead of fourteen scrolls deep.
Here's the hard part: the rhythm is the phase where people relapse. Knowing the method and running the method on a busy week are different things, and the default behavior — open IG, reply to whatever's on top — is wired in deep. The spreadsheet only beats the algorithm on days you actually open the spreadsheet, and that's worth saying out loud before you start rather than three weeks in.
So treat the manual version as a 14-day experiment instead of a new habit you're promising yourself. If you're still running the Scan on day 15, you've built the rhythm and you don't need anything else from me. If the spreadsheet quietly died somewhere in week two, write down which part died first, because that's the exact spot where willpower was holding the system up — mine died at the daily re-score, which is why the last section of this guide exists.
What 60 days of this looks like
Run the math on your own numbers. If your inbox has been losing 2–4 good leads a month and your offer runs $2,000–$3,000, bad sorting has been costing you roughly $5,000 a month in revenue you'd already earned. That's the number in this guide's title, and for most founders I've audited it's conservative, because the first audit digs through 90 days of backlog at once.
The deeper change is harder to put a number on. Somewhere in the first few weeks, the story you tell about yourself shifts — "I'm bad at sales and I can't keep up" stops surviving contact with the evidence, because the evidence says you were selling fine. You just couldn't see who you were selling to.
The part where I tell you what I built
Everything above works manually. I taught it as consulting before any software existed, and founders ran it with nothing but a spreadsheet and the signal map. If you're getting a handful of DMs a week, the spreadsheet is honestly all you need — close this tab and go run it.
The founders who break the manual version are the ones with real volume, which is exactly who the method matters most for. More conversations means more buried buyers the sort is costing you, and more daily re-scoring demanded from a brain that's already at its decision limit. Volume makes the method valuable and the spreadsheet unsustainable at the same time.
Here's what happened to me, though: I'm a chronic overthinker with ADHD goodness on top, and the manual version of my own method kept losing to my own brain — the audit went stale, busy weeks ate the Scan, and re-scoring by hand every morning turned out to be exactly the kind of repetitive sorting task my brain quits doing after week two. I break things like this, so I build things that don't depend on me. it's a whole pattern.
So I automated the sort. ForesightHQ connects to your Instagram and Facebook DMs, scores every conversation by buying likelihood using the same signals you just learned, and keeps the ranked view current every day so the Scan stops being a spreadsheet you maintain and becomes an app you open. It reads your conversations to score them — it never posts or replies as you, and you can disconnect it anytime.
The part worth underlining: the audit you just spent an hour on by hand, it runs in minutes on your full 90-day backlog, the day you connect — Requests folder included. The buried-buyer list shows up before you've done anything but log in.
The Standard plan is $39 a month. One recovered $3,000 sale pays for six-plus years of it, which is the kind of wallet math I can say with a straight face precisely because you now know how many of those sales are sitting in your backlog. The first audit will show you — it's foresighthq.app when you want to run it.
Either way, run Phase 1 this week. The 22% has been in your inbox the whole time, and the only thing that's ever been missing is the sort.