Every framework, principle, and mechanic that goes into a LeadVolley message — structure, persuasion, and execution, laid out in full. This is documentation, not a pitch: the same reasoning we apply to every client's campaigns, written out so it can be checked, referenced, and reused.
Most cold email advice mashes three separate questions into the single word "framework." Splitting them apart is the single most useful move for both writing and diagnosing messages — each layer answers a different question and fails in a different way.
| Layer | Question it answers | Examples | How it fails |
|---|---|---|---|
| 1. Structure | What order do ideas arrive in — the skeleton | Trigger-led, PAS, BAB, proof-led, question-led, AIDA, plain/blunt | Wrong arc for the buyer's awareness level — agitating a pain they don't feel yet, or pitching a vision to a skeptic |
| 2. Persuasion elements | Which psychological levers get layered onto that skeleton | Personalization, trust, scarcity, urgency, reciprocity, and more | Overstacking — every lever at once reads as an infomercial; or a fake lever (manufactured scarcity) that destroys trust |
| 3. Execution craft | How it's actually rendered on screen | Subject line, opener, length, CTA design, formatting | A great message nobody opens, or one that opens well and dies on a high-friction CTA |
One caveat before any of it: messaging is the third input to results, not the first. List quality and offer strength dominate — the best-written email to the wrong person still fails, and a mediocre email to exactly the right person at exactly the right moment often works. When a campaign underperforms, copy is where everyone looks first, and where the problem least often lives. See Diagnostics.
Looking for plain definitions instead of the full framework? See the Glossary.
The skeleton decides what the reader encounters first, second, and third. The right choice depends mostly on the buyer's awareness state — how conscious they already are of the problem — and their skepticism level.
| Framework | Skeleton | Best for | Failure mode |
|---|---|---|---|
| Trigger-led / trigger | Observed fact about them → why it implies a problem or opportunity → low-friction ask | Skeptical, over-pitched buyers; any time a real signal exists | A generic "signal" ("I saw you're in SaaS") reads as fake research — worse than no personalization at all |
| Problem–Agitate–Solve | Name the pain → make its cost vivid → position the offer as relief | Pain-aware buyers who feel the problem daily | Agitating a pain the reader doesn't actually feel reads as manipulative |
| Before–After–Bridge | Current state → desired future state → offer as the bridge | Vision-driven buyers; category-new offers | Vague future-state language ("imagine 10x growth") with no proof behind it |
| Proof-led | Result achieved for a peer → mechanism in one line → "want to see how?" | Risk-averse buyers who need comparables | Unnamed or implausible proof — a big number with zero specifics |
| Question-led | One sharp question about their situation, optionally one line of context | Opening a conversation, not closing one | A rhetorical or leading question ("Want more sales opportunities?") reads as fake and gets deleted |
| AIDA | Attention → Interest → Desire → Action | Longer-form or warm audiences; landing pages more than cold email | Too long for cold — four full beats rarely fit in 100 words |
| Plain / blunt | Who I am → what I do → is this relevant? No dressing. | Technical buyers allergic to marketing; refreshing in over-copywritten niches | Blunt without relevance is just spam with fewer words |
Two practical notes. First, a reader-context opener is less a competing framework than a universal opener — a real observed signal can lead into a PAS body, a proof body, or a question. That's why it's the default we lead with: it's compatible with everything and earns the read the other frameworks then spend. Second, no framework survives the wrong awareness level. The same outbound-system pitch is PAS to a founder already drowning in unpredictable outbound, and BAB to one who hasn't yet clocked that inconsistent outbound is a solvable problem. Framework choice is really a targeting decision wearing a copywriting costume.
The levers layered onto the skeleton. These seven form the working core rubric — each maps to a well-studied persuasion principle.
| Element | Underlying principle | What it looks like in copy | Misuse risk |
|---|---|---|---|
| Reader-centric writing | Relevance — people scan for "is this about me?" | "You/your" dominates; the product is barely named; sentence one is about them, not the sender | None, really — close to universally safe |
| Pattern interrupt | Attention — breaking the expected shape of spam | A subject or opener that names their exact situation instead of a pitch shape they've seen a hundred times | Weirdness for its own sake — interrupting the pattern and then having nothing relevant to say |
| Personalization | Costly signal — proof a human (or a good system) did homework | A detail only true of them: their tech stack, a job post, a funding announcement | "Personalization theater" — fake-specific flattery ("love what you're building!") that signals automation, not effort |
| Trust / social proof | Authority — others already vouch | A named client, a count, a specific rating or result | Unverifiable or padded claims that can't survive a quick check |
| Scarcity | Loss aversion — limited supply | A real, limited capacity — "only taking on a couple of pilot clients this quarter" | Fake scarcity is the fastest trust-destroyer on this list — only use real constraints |
| Urgency | Loss aversion — limited time | Time-sensitivity tied to a real trigger — a funding window, a new-hire's first 90 days | A manufactured deadline with no reason behind it reads as a pressure tactic |
| Low-friction CTA | Commitment gradient — small yeses before big ones | A yes/no question ("want me to send it over?") instead of "book 30 minutes" | So soft it's unanswerable ("thoughts?") — friction isn't the only variable, clarity matters too |
| Element | Underlying principle | What it looks like in copy |
|---|---|---|
| Reciprocity / value-first | Give before asking | Attach the useful thing itself — a breakdown, a benchmark — with no gate and no ask beyond "want more?" |
| Specificity | Numbers beat adjectives; detail signals truth | "Teams your size typically add 15–20 qualified conversations a month" vs. "dramatically improve your sales opportunities" |
| Similarity / liking | People trust people like them | "We work with other [their industry] companies" — peer framing, not aspirational framing |
| Loss framing | Losses loom larger than equivalent gains | "Every week without outbound is conversations a competitor is having instead" vs. "grow your sales opportunities" |
| Curiosity gap | Open loops demand closing | "Noticed something about how you're targeting X — 30-second explanation if you want it." Powerful, easy to abuse into clickbait |
| Consistency | Small agreements predict big ones | A permission-based CTA is this in disguise — a "yes" to receiving info makes the next "yes" easier |
The most important thing about Layer 2 is restraint. A message deploying every lever at once — personalized, urgent, scarce, proof-laden, curiosity-hooked — reads exactly like the infomercial it structurally is. Strong cold messages typically run two or three elements deliberately and skip the rest.
A useful discipline: pick one primary lever — the thing the message is "about" psychologically — and at most two supporting ones, and let the specific prospect's real situation choose which. A prospect with a public pain signal wants personalization plus loss-framing; a prospect in a proof-hungry industry wants social proof plus specificity; a genuinely time-bound situation earns urgency honestly.
A well-chosen real situation often generates several elements at once without any writing tricks. "We're taking on two pilot clients this quarter and have one slot left" is simultaneously scarcity, urgency, and specificity — and it works precisely because it's just true.
| Approach | Example | Notes |
|---|---|---|
| Specific-relevant | "outbound for post-Series-A teams" | Names the exact thing the reader already wants; strongest general-purpose choice |
| Signal-referencing | "your Series A" | Ties directly to the observed trigger; naturally curiosity-generating without clickbait |
| Internal-style / casual | "quick one re: sales opportunities" | Mimics a colleague's email; opens well but risks feeling deceptive if the body is clearly a pitch |
| Curiosity | "noticed something about your outbound" | High open rates; must be paid off in the first line or trust dies immediately |
| To avoid | ALL CAPS, "FREE", emoji, any claim the body doesn't honor | Spam-filter triggers and trust-killers — the subject is a promise the body must keep |
The first line does two jobs: prove the message is about the reader, and prove it isn't a blast. A real signal or hook does both simultaneously, which is why it should almost always be sentence one — not after an introduction. "Hi, my name is..." and "I hope this finds you well" fail both jobs and burn the only guaranteed-read line in the email.
On mobile, the subject plus roughly the first ten words of the body are all a prospect sees before deciding — write those words as if they're the whole email, because for most recipients they are.
| CTA type | Example | Friction | When to use |
|---|---|---|---|
| Hard meeting ask | "Do you have 30 minutes Thursday?" | High | Warm audiences, referrals, or very strong signals — usually too much for touch one |
| Calendar link | "Grab a time here" | High | Convenient but presumptuous cold — better mid-conversation, once interest exists |
| Soft interest ask | "Worth a quick look at this?" | Med | The workhorse — clear, answerable, low pressure |
| Permission-based | "Want me to send over the breakdown?" | Low | Best-in-class for touch one — a yes costs seconds and starts a consistency chain |
| Question CTA | "How are you handling outbound right now?" | Low | Opens dialogue rather than selling; great for conversation-first sequences |
| Binary yes/no | "Want a 2-minute rundown of how this would work for you?" | Lowest | One decision, one thumb-tap to answer — the entire email exists to set up this single question |
One CTA per message, always. Two asks halve the response to both — and the CTA should match the email's actual goal: touch one's job is usually to start a conversation, not book the meeting.
"Personalized" isn't binary — it's a spectrum with real cost/scale tradeoffs, and choosing the right depth per campaign tier is a strategic decision, not a copywriting one.
| Depth | Example | Cost / scale | Honest assessment |
|---|---|---|---|
| Mail-merge tokens | "Hi {FirstName}, I see you work at {Company}" | Free, infinite | Fools no one — arguably worse than nothing, because it advertises automation |
| Firmographic | "...for companies like {Company} in {industry}" | Cheap, near-infinite | Table stakes — establishes broad relevance only |
| Persona-based | Different pains/angles per role (founder vs. VP Sales) | One-time work per persona, then scales freely | The highest-leverage tier — most of the benefit of personalization at template cost |
| Signal-based | "Saw you just closed your Series A..." | Per-prospect research, automatable with good data | The sweet spot for quality-tier campaigns — the signal doubles as the qualification |
| Bespoke research line | A genuinely specific observation about their business, written for them alone | Minutes per prospect; doesn't scale | Reserve for high-value targets — bland output here is the tell of AI-drafted copy pretending otherwise |
The practical architecture: persona-based templates as the floor for everyone, signal-based parameters for the qualified tier, bespoke lines for the handful of accounts that justify minutes of human attention. The failure mode to avoid: spending bespoke effort on unqualified volume, or blasting mail-merge at accounts that deserved research.
Each follow-up is its own message with its own job — the classic failure is a follow-up that just repeats touch one with "bumping this up." A working sequence changes the angle, not just the timestamp.
| Touch | Job | Framework note |
|---|---|---|
| 1 — Hook | Earn the read; open a conversation | Trigger-led opener; permission-based or binary CTA — the persuasion strategy lives here |
| 2 — Bump | Resurface without nagging | Shortest message in the sequence; add one new, concrete detail — never just "any thoughts?" |
| 3 — Value / new angle | Give before asking again; switch persuasion levers | If touch one was urgency-led, go proof-led here — this is reciprocity's slot |
| 4 — Breakup | Close the loop with dignity | Often the highest reply rate in the sequence — no guilt, the door stays open |
Vary the lever across touches the way you'd vary the angle across test cells — a prospect unmoved by urgency may still move for proof. A four-touch sequence that runs four different levers is effectively an in-flight test on that one prospect.
Four original openers, annotated against the layers above. All four begin with reader context; what changes is the persuasion levers, chosen by how much real signal exists.
| Weak | Strong | What changed |
|---|---|---|
| "I hope this email finds you well. My name is Alex and I work at..." | "Saw [Company] closed a Series A last month —" | Burned line → signal opener |
| "We help businesses dramatically improve their sales opportunities" | "Teams your size typically add 15–20 qualified conversations a month once outbound is running" | Adjective → number; us → them |
| "We're the leading outbound provider trusted by many companies" | "Built specifically for [their industry] companies — not a generic one-size-fits-all sequence" | Unverifiable claim → checkable specificity |
| "Let me know if you'd like to schedule a 30-minute demo" | "Want me to send over the 2-minute breakdown?" | High-friction meeting ask → permission-based micro-yes |
| "Just bumping this to the top of your inbox!" | "One thing I didn't mention last time — setup takes a week, not a quarter." | Empty bump → new, concrete information |
| "Act now — limited time offer!" | "Taking on a couple of pilot clients this quarter — a few slots left." | Fake urgency → true, specific constraint |
This is the highest-leverage table in this whole reference, because "rewrite the copy" is only the right prescription for some of these symptoms.
| Symptom | Likely layer | First things to check |
|---|---|---|
| Low delivery / spam placement | Not messaging at all — infrastructure | Warm-up, authentication, volume ramp, links/images in copy. No copy rewrite fixes a deliverability problem. |
| Opens but no replies | Structure, elements, or the offer | Is the CTA answerable in one thumb-tap? Is the message about them or about us? If multiple angles all die, it's the offer, not the words. |
| Replies, but "not relevant" | Targeting, not copy | The message reached someone whose situation doesn't match the premise — fix the segment or signal, not the sentences. |
| Replies, but wrong persona engages | Persona mapping | The message's pain resonated with the user, not the buyer (or vice versa) — re-split angles per persona. |
| Positive replies that stall pre-meeting | Sequence / reply handling | Examine the reply-stage messages and response speed, not the cold email itself. |
The discipline this table enforces: diagnose before rewriting. In practice, targeting and offer problems outnumber copy problems — they just look identical from the inbox.
The frameworks above are what goes into every sequence we run. Happy to walk through exactly how they'd apply to your ICP.
Apply for the pilot