Section 1
Key takeaways
• No cold channel wins in isolation: cold email runs ~0.45% replies-per-send, cold calls 2.3% success, and LinkedIn ~10% response on average . • The lever isn't the channel, it's the trigger: a concrete reason to reach out right now, plus whatever relationship already exists. • Match one channel to the signal: DM for warm threads and named humans, call for urgent or high-stakes hard-to-reach decision-makers, email for cold, scalable, tightly-targeted lists. • Spraying all three at everyone hits the floor of every channel and burns the list, including your domain reputation, the asset you can't see. • Reserve the full stack, sequenced, never simultaneous, for the few named accounts whose value justifies real orchestration.
Section 2
The benchmarks, read honestly
Start with the numbers everyone quotes wrong, because the denominator is where the lies live. Belkins analyzed 7.5 million cold emails sent in 2025 and reported an average reply rate of 0.45% . That sounds catastrophic until you notice the method: they measured replies divided by total emails sent, not replies divided by opens . That's a deliberately strict denominator. Plenty of vendors quote replies-per-open instead, which is why you'll see "good" cold-email reply rates that look several times healthier . Both can be true. One measures the whole funnel, including everyone who never opened the message; the other measures only the people who cracked the envelope. When a founder tells you their cold email "gets a healthy reply rate," the first question is: of what? Of opens or of sends? The gap between those two denominators is the gap between a campaign that prints meetings and one that quietly burns your domain reputation while you congratulate yourself on a vanity number. Here's the figure inside the Belkins data that actually matters for service founders: founders and owners were the single most responsive seniority group, replying at 0.57%, higher than any other title . If you sell peer-to-peer, you're a founder emailing a founder, you're addressing the most responsive audience in the dataset. That's not a license to spray. It's the opposite. It tells you the person most likely to reply is also the person most allergic to obvious automation, because they get a hundred of these a week and recognize every template on sight. The responsiveness is real, but it's conditional: it rewards the sender who reads like a peer and punishes the one who reads like a sequence. Cold calling tells a similar story from a different angle. Cognism reports the average cold-call success rate fell to 2.3% in 2025, roughly half of 2024's 4.82% . Their own data-led teams hit 6.7%, nearly triple the average, which is the whole point: the channel's yield is mostly a function of data quality and targeting, not the channel itself . But the real bottleneck on calls isn't receptiveness. It's reach. A study of more than 55,000 dials found only a 16.6% connect rate even with quality data . You dial six people to talk to one. Yet of the people you do reach, 69% of B2B buyers say they're open to cold calls from new providers, and 82% have accepted a meeting from strategic cold outreach . Read those two facts together and the call stops looking like a numbers game and starts looking like a trigger game: the bottleneck is getting them on the phone, and the only thing that justifies the dialing is a reason live conversation beats text. Then LinkedIn. Expandi analyzed more than 70,130 real campaigns and found a 29.61% connection-acceptance rate and a 10.3% average response rate, which it frames as more than 101% more replies than cold email . On its face, LinkedIn wins. But look closer at what gets replies. Conversational, post-connection messages, Expandi calls them Messenger campaigns, reply at 16.86%, versus 8.62% for templated pitches . Roughly double. The lift on LinkedIn isn't the platform. It's the human-to-human thread after someone has already accepted you. A templated pitch fired the moment a connection lands performs about as well as decent cold email. The conversation is what converts, not the logo on the app. Notice the pattern across all three channels. The number that moves isn't the channel, it's the presence or absence of a real reason, a real relationship, and a real human on the other end. That's the lever. Everything else is logistics.
Section 3
Why "spray all three" feels safe and isn't
The reflex, once you accept that every channel underperforms in isolation, is to run all three at everyone. More surface area, more shots, more coverage. It feels like diversification. It's actually a confession that you don't know your trigger. Consider a real shape: a 12-person fractional-CFO firm, a firm that rents out senior finance leadership without the full-time cost, selling to founders of $2–10M e-commerce brands. They buy a list of 4,000 "ICP-matched" contacts, where ICP means ideal customer profile, load them into a sequencer, and fire email plus a LinkedIn connection plus a call task on every single one, same week, same generic message. On paper that's 12,000 touches. In practice it's 4,000 people getting pitched three times by a firm that has done zero work to know any of them. The email gets the 0.45% base rate or worse, because nothing in it earns the entrance fee . The LinkedIn note is a template, so it lands near the 8.62% floor, not the 16.86% ceiling . The calls connect 16.6% of the time and, with no trigger to open on, convert near the 2.3% average . Three channels, three floors, and a burned list you can never warm up again. And the burned list is the real cost, the one founders discount because it doesn't show up on this quarter's dashboard. A prospect who deletes your email, ignores your connection request, and screens your call in the same week hasn't just declined once, they've filed your name under noise. The next time you reach out, even with a genuine trigger, you're climbing out of a hole you dug yourself. Worse, the email volume itself degrades the asset you can't see: send thousands of unwanted messages and mailbox providers begin routing you to spam, so the few well-researched emails you send later never even arrive. Spraying doesn't just waste this campaign. It taxes every campaign after it. Steffen Hedebrandt, co-founder of Dreamdata, said it plainly in Belkins' 2026 outreach blueprint: "If you're not willing to 'pay the entrance fee' and do the actual work of getting to know me enough to write a relevant email, stay out of my inbox!" That's not a complaint about cold outreach. It's a pricing model. The entrance fee is the research, the relevance, the genuine reason. Spraying all three channels is what it looks like when a founder refuses to pay it and tries to make up the difference with volume. Volume doesn't cover an unpaid entrance fee. It just spreads the rejection across more channels, and it does it faster, three times faster, in the three-channel spray, because you exhaust the prospect's tolerance on three surfaces at once instead of one. The data backs the alternative, but read it carefully. Belkins reports that orchestrated multichannel outreach outperforms single-channel, over 20% higher close rates, 20% lower CAC, and 25% shorter sales cycles . That looks like an argument for using every channel. It isn't. The operative word is orchestrated: sequenced, account-by-account, with each touch building on the last. Orchestration is the opposite of spraying. Spraying is simultaneous and identical; orchestration is sequential and earned. The lift comes from the orchestration, not the number of channels, and you only orchestrate accounts worth the labor. The moment you try to orchestrate 4,000 accounts you stop orchestrating and start spraying again, because nobody can research 4,000 accounts well enough to sequence them by hand. There's a deeper reason orchestration beats spray that the close-rate numbers only hint at. A sequenced touch carries memory: the email references the comment, the call references the email, and each step proves you've been paying attention. That accumulated evidence of effort is exactly the entrance fee, paid in installments. Spray carries no memory, every touch arrives as if it were the first, because to the sender it basically was. The buyer feels the difference immediately. One reads like a relationship forming; the other reads like a machine that happened to find their address three times.
Section 4
The BGA framework: Trigger-Channel Fit
Pick the channel by where the relationship and the buying signal already live, never by what's easiest to automate. A trigger is any concrete event or condition that gives you a real reason to reach out now: a funding round, a new hire in a role you serve, a job posting that exposes a gap, a public complaint, a warm intro, a content interaction, a renewal window. No trigger, no message. With a trigger, the trigger itself tells you the channel. Most founders struggle here not because triggers are rare but because they're not watching for them. Triggers hide in plain sight: a job posting that names the exact pain you solve, a press release, a leadership change, a product launch, a comment thread where a prospect describes their problem in their own words. The discipline is to build a habit of collection, a saved search, an alert, a five-minute scan each morning, so that when a signal fires you already have the named human and the reason attached. A founder who waits to invent a reason at send time will always default to the template. A founder who collects triggers continuously always has something true to say. 1. DM when there's a warm thread or a real signal and a named human. Use LinkedIn or social DM, direct message, when the person can already half-place you: they accepted your connection, engaged with a post, were introduced, or operate in a circle where your name carries. The channel rewards conversation, not pitching: Expandi's Messenger replies hit 16.86% precisely because they come after a connection and read like a human, where templates sit at 8.62% . Rule of thumb: never pitch in the connection note, and never send a DM you couldn't defend as relevant if the person screenshotted it to a peer. For the fractional-CFO firm, the DM play is the founder who just posted about a painful month-end close or a fundraise, you have a named human and a live signal, so you reference the specific post, not your service. 2. Call when the decision-maker is hard to reach and the trigger is urgent or high-stakes. Reserve the phone for situations where a live conversation actually changes the outcome, a complex problem that needs back-and-forth, a time-sensitive trigger, a deal big enough to justify dialing six times to connect once. Connect rates are brutal at roughly 16.6% , but receptiveness isn't: 69% of buyers are open and 82% have taken meetings from strategic outreach . The metric to protect is data quality, Cognism's teams hit 6.7% versus a 2.3% average mostly by calling the right people with a real reason . Rule of thumb: if you can't state the trigger in your opening sentence, you're not ready to dial. The CFO firm calls when a brand just closed a Series A, capital in, finance function not built yet, and a fifteen-minute conversation beats any email. 3. Email when it's cold, scalable, and there's little prior relationship. Email is your only true scale channel, so accept the low base rate and win on two things: a tight ICP and a real reason to write. Founders out-reply every other group at 0.57% , so peer-to-peer email is viable, but only if you pay the entrance fee Hedebrandt names . Rule of thumb: measure replies-per-send like Belkins does , not the flattering replies-per-open number vendors prefer , and hold yourself to beating the 0.45% baseline by a multiple, not a hair. For the CFO firm, email is the right tool for the 300 brands that match the ICP perfectly but have no warm thread and no acute event, segment tightly, write one genuinely researched line per prospect, and keep volume low enough that every send is defensible. 4. Stack all three only on your highest-value named accounts, sequenced, never simultaneous. This is where Belkins' 20%-plus close-rate lift, 20% lower CAC, and 25% shorter cycles come from : a small list of accounts worth real orchestration, touched in a deliberate order, comment on their post, then a connection DM referencing it, then a researched email, then a call when a trigger fires, each step earning the next. Rule of thumb: if you can name fewer than 25 accounts on this list, you're doing it right; if it's 2,500, you're spraying and calling it strategy. The CFO firm runs this on the ten dream logos whose lifetime value justifies a partner spending an hour per account per month. The rule in one line: one trigger, one best-fit channel; reserve the full stack for accounts worth orchestrating. Spraying all three at everyone isn't coverage. It's the tax you pay for not knowing your trigger.
Section 5
Working it through: three founders, three correct answers
A boutique brand-design studio gets a notification that a past client's VP of marketing just moved to a new company. That's a warm thread and a named human. Trigger-Channel Fit says DM, reference the move, the relationship, no pitch. Email would waste the warmth; a cold call would be strange. One channel, correct. A commercial-insurance broker learns a regional manufacturer just had a public safety incident. High stakes, time-sensitive, a decision-maker who won't read a cold email this week. That's a call, and the opening line is the trigger. The 16.6% connect rate means persistence, but the 82% who've taken meetings from strategic outreach exist because someone called with a real reason . One channel, correct. A managed-IT provider wants to reach 400 dental practices in three metros. No relationships, no acute events, but a clean, narrow ICP. That's email, cold, scalable, low base rate accepted, won on segmentation and one researched line each . DMs don't scale to 400 named humans you've never met; calls would take a month. One channel, correct. In all three, the founder did the same thing: read the trigger and the relationship, then matched. None of them sprayed. And the one dream account each of them would orchestrate across every channel? That's the short list, the place the full stack earns its 20%-plus lift .
Section 6
You're running Trigger-Channel Fit right when…
You can look at any prospect and name two things before you pick a tool: the trigger and the existing relationship, and the channel falls out of those two facts instead of out of whatever your sequencer makes easiest. Your "stack everything" list is short enough to recite from memory and every name on it has a reason to be there. You measure email replies-per-send and still beat the baseline, your DMs read like a human who paid the entrance fee, and you only dial when you can state the trigger in your first sentence. If you're firing all three channels at a 4,000-row list and calling it pipeline coverage, you're not running the framework, you're paying the tax.