Section 1
Key takeaways
• The baseline reply rate across 12 million outreach emails is just 8.5%, your first touch is a coin flip, and sending one more follow-up boosts replies by 65.8%. • Follow-up steps 2 through 6 carry 58.6% of all replies, and Step 3 alone drives 35.6% of email-sourced meetings, the deal lives in the messages most founders never send. • A second email with no new content replies at 1.8%; one carrying a fresh angle replies at 8.4%, nearly five times better. The variable is content, not courage. • The operating range is 3 to 5 touches, spaced tight early and wider later (Day 0, 3, 7, 12, 18+), because that's how attention decays. • One rule governs the whole thing: one new value angle per touch, proof, reframe, resource, break-up, never a naked "circling back."
Section 2
The first email is a coin you flip, then walk away from
Start with the uncomfortable baseline. Backlinko analyzed 12 million outreach emails and found that only 8.5% of them get any response at all . So nine out of ten messages you send into a cold inbox vanish. That's not a copywriting problem you can fully fix, it's the physics of a busy person's inbox. People are in meetings, on holiday, buried, or simply weren't ready the day you showed up. Part of that is timing you don't control, most of any market isn't in-buying-mode when you arrive, which is why 95% of your market isn't buying right now and a single send can't be the whole test. Now hold that next to a second number from the same study. Sending just one additional follow-up boosts replies by 65.8% . As Backlinko's 12-million-email study frames the follow-up gap bluntly: "a single additional follow-up message can lead to 65.8% more replies" . Read those two findings together and the conclusion is almost violent in its simplicity. The single highest-ROI action available to most service founders is not a better subject line, a fancier tool, or a new channel. It's sending message number two. The founder who stops at one is leaving the majority of their potential pipeline sitting in drafts that never got written. Belkins' 2025 sequence data makes the same point from the other direction. In their numbers, follow-up steps 2 through 6 together account for 58.6% of all replies . Sit with that. The majority of responses do not come from your opener. They come from the messages most founders never bother to send. The first email isn't where the deal happens. It's the cover charge that gets you into the room where the deal happens. If you run a consultancy, an agency, a fractional practice (a senior operator rented part-time rather than hired full-time), a managed-services shop, anything where the sales motion is "reach a specific human and earn a conversation", this is your single biggest leak. Not your offer. Not your positioning. The fact that you send one email and call it a market test.
Section 3
Why founders quit: the real failure is content, not courage
Here's where most advice goes wrong. It tells you to "be persistent" and "follow up more," as if the problem were a willpower deficit. It isn't. The reason follow-up feels gross, the reason you personally hesitate to hit send on touch number three, is that you've got nothing new to say, so the only honest version of the message is "did you see my last email?" And you know that's annoying because it annoys you when you receive it. So you face a fake choice: be a pest, or give up. Most founders, being decent people, give up. The escape from that trap is to change what a follow-up is. A follow-up is not a reminder that you exist. It's a new piece of value delivered to someone who hasn't bought yet. When you reframe it that way, the willpower problem dissolves, because now every touch has a reason to exist that has nothing to do with your need for a reply. It's the same logic that makes nurturing the 95% pay off rather than pester, value, repeated, to people who simply weren't ready yet. The data backs this hard. Gangly studied B2B cold emails and isolated the one variable that matters most: whether the follow-up carried something new. A second email with no new content replied at 1.8%. A second email carrying a fresh angle, a different proof point, a different framing of the pain, a genuine "is this still relevant?", replied at 8.4% . That's the gap, and it's driven entirely by what was inside the message, not the fact that a message was sent. Same slot in the sequence, same sender, and the version with a new angle pulled replies nearly five times higher. That's the whole game in one comparison. Two emails, same prospect, same sender, sent at the same point in the sequence. The empty one barely moves. The angle-carrying one outperforms it nearly five to one. Persistence sent both. Only one of them worked. This is why "just follow up more" is incomplete advice. More empty touches don't compound, they decay, and they burn goodwill on the way down. A ladder where every rung climbs is what compounds. A ladder where every rung sits at the same height is just a tall, annoying way to get ignored.
Section 4
How many touches, and where to stop
There's a fear lurking under all of this: if I keep going, won't I cross the line into harassment? Yes, eventually. So let's bound it. Belkins puts the practical sweet spot for B2B sequences at 3 to 5 steps . That's the operating range. Below three, you're leaving the 58.6% of replies that live in the follow-up steps on the table . Far beyond five, you're scraping diminishing returns and risking your reputation for marginal gain, and at that volume, the boring mechanics of deliverability start to bite as much as the manners. Three to five is where the math and the manners agree. Inside that range, one step does disproportionate work. In Belkins' data, Step 3 alone books more appointments than Steps 1 and 2, driving 35.6% of all email-sourced meetings . The third touch is the peak. Not the first, where most founders put all their energy, and not some heroic twelfth touch. The one that lands after two prior messages, when you've established you're not going away, and you arrive with a sharp reframe rather than a nudge. Almost nobody gets to Step 3 with anything fresh to say, which is exactly why Step 3 is so productive for the people who do. The bar is on the floor. Most of your competitors sent one email and vanished. A few sent two "circling back" notes and gave up. You showing up a third time with a genuinely new angle isn't competing against a crowd, it's competing against silence. That's the quiet advantage of this whole framework: the discipline it demands is exactly the discipline almost no one else is willing to supply, so the field thins out at precisely the touch where the meetings get booked.
Section 5
Spacing: tight early, wider later
Touch count is half the design. Spacing is the other half. Get it wrong and you either crowd the inbox (and read as desperate) or vanish so long they forget who you are. The pattern that works: tight early, wider later. Early in a sequence, the prospect has the most context, they just saw your name. A 2-to-3-day gap keeps you in active memory without stacking unread messages on top of each other. As the sequence ages, widen the gaps. By the back half, you're spacing touches out by a week or more, because the job has shifted from "stay in their working memory" to "catch them on a day when the timing finally works." This front-loading is the same force behind signal decay: the relevance of a touch fades fast, so your sharpest windows are the early ones. This isn't just my preference. Practitioner cadence guides land on the same shape, early touches a few days apart, then widening intervals as the sequence ages . The convergence is the point: independent operators, working from independent data, keep arriving at front-loaded spacing because that's how attention decays. Concretely, the cadence I run and recommend stretches across roughly three weeks: Day 0, Day 3, Day 7, Day 12, and Day 18-plus. Notice the shape, the first three touches land inside a week, then the gaps open up. That front-loading respects how attention actually works. You're most likely to convert someone while you're still a fresh memory, so you don't waste your best windows by spacing them a month apart. But you also don't carpet-bomb, because the gaps widen exactly as the risk of irritation rises. This shape also keeps you honest about quality. Five touches in three weeks, each requiring a distinct angle, forces you to actually have five things to say before you start. If you can't fill the rungs, that's a signal your offer or your research isn't deep enough yet, better to learn that before you hit send than after you've annoyed a good prospect.
Section 6
The BGA framework: The Follow-Up Ladder
One rule governs the whole thing: one new value angle per touch, never a naked "checking in." If a rung doesn't carry something the prospect hasn't seen, it doesn't go out. Here's the ladder, with the angle, the timing, and the rule of thumb for each rung. (Building five distinct rungs is the kind of swipe-and-adapt work a template pack is for, a ready set of angles beats inventing them at the moment of hesitation.) Rung 1, The Ask (Day 0). State who you help, the specific outcome, and the single small next step. No life story. One clear ask. This is the cover charge, expect it to underperform, because the baseline reply rate across millions of emails is just 8.5% . Metric: if you're getting meaningful replies from Rung 1 alone, your list is unusually warm; for cold outreach, treat a non-reply here as the start of the sequence, not the end. Rung 2, New Proof (Day 3). Bring a result they haven't seen. A specific outcome for a comparable business: "We took a 12-person IT firm from a 3-week proposal backlog to same-week turnaround." Don't restate the ask, evidence it. This is the rung that separates the 1.8% from the 8.4% in Gangly's data : same slot in the sequence, but the version carrying a fresh proof point replies nearly five times better. Rule of thumb: one concrete, named-shape result beats three vague claims. Rung 3, New Angle (Day 7). Reframe the problem. Shift from "here's what I do" to "here's the cost of leaving this unsolved." If Rungs 1 and 2 sold the upside, Rung 3 names the downside of standing still, the slow leak they've stopped noticing. This is where you make the cost of inaction concrete, because the status quo, not a competitor, is what's actually winning the deal. This is statistically your peak: Step 3 drives 35.6% of all email-sourced meetings in Belkins' data . Put your sharpest reframe here. Rule of thumb: if a rung is going to get a real reply, it's most likely this one, so don't phone it in. Rung 4, New Resource (Day 12). Give value with zero ask attached. A template, a teardown of something in their world, a relevant introduction, a two-paragraph diagnosis of a problem you can see from the outside. No "so can we chat?" This rung does two jobs: it proves you're useful before you're paid, and it breaks the pattern of every other vendor in their inbox who only ever takes. Rule of thumb: the resource should be useful even if they never reply, if it isn't, it's just a disguised ask. Rung 5, The Break-Up (Day 18+). Close the file, out loud. "I'll assume the timing's off and stop here, want me to leave the door open, or should I close your file?" This is not surrender. It's the rung that reliably resurrects dead threads, because it removes all pressure and quietly triggers loss aversion, people who ignored four value-packed emails will suddenly reply to keep a door from closing. The mechanics of writing this one well are worth their own study, see the break-up message. Rule of thumb: make the exit genuinely easy and face-saving. The break-up only works because most prospects, given a clean way out, choose not to take it. The compounding logic is the whole point. Gangly proved an empty second touch replies at 1.8% while an angle-carrying one replies at 8.4% . Stack that effect across five rungs that each climb, and you capture the 58.6% of replies that live in the follow-up steps, instead of flipping one 8.5% coin and walking away. A worked example Take a fractional CFO prospecting a founder-led e-commerce brand. The lazy version: five emails, all some flavor of "circling back on my note below." Dead by email three. The ladder version. Rung 1 (Day 0): "I help DTC brands, direct-to-consumer, selling straight to shoppers rather than through retailers, doing $3-10M get to a clean monthly close in under five business days. Worth a 20-minute look?" Silence. Rung 2 (Day 3): "Quick proof point, last quarter I cut a $6M apparel brand's close from 19 days to 4, which surfaced a margin leak they'd been eating for a year." Silence. Rung 3 (Day 7): "The real cost here usually isn't the slow close, it's making Q4 inventory bets on numbers that are three weeks stale. That's the part that gets expensive." Now you're talking about their exposure, not your service. Rung 4 (Day 12): "No ask, here's the 8-line close checklist I give every client. Even if we never work together, this'll tighten your month-end." Rung 5 (Day 18): "Haven't heard back, so I'll assume the timing's off and close your file. Want me to leave the door open for Q1?" Five touches. Five distinct reasons to reply. Not one "checking in." That's a ladder. The other version was a guy knocking on the same door five times. The difference between them isn't effort, both sent five emails, it's that the ladder version did its homework once, up front, and earned the right to keep showing up.
Section 7
You're running The Follow-Up Ladder right when…
You can open any prospect's thread, count three to five touches, and find that every single rung says something the previous one didn't, a new proof point, a sharper reframe, a free resource, a clean break-up, with not one naked "circling back" in the chain. Your spacing tightens early and widens late, your third touch carries your best angle because that's where the meetings live, and you'd be comfortable receiving every message you sent. If you can't fill five rungs with five real angles, you don't have a persistence problem, you have a content problem, and that's the better problem to have, because it's the one you can actually solve before you hit send.