CAMPAIGN BRIEF - WORKED EXAMPLE (v3)

READ THIS FIRST

This is a real brief, in the exact shape yours should take. It is here so you can see the
target structure, the level of detail, and the tone - not so you can ship it.

Two things are deliberately not yours to keep:

1. THE PROOF FIGURE IS A PLACEHOLDER. Every mention of "{proof}" is where YOUR verifiable
   result goes. The filled examples further down use an invented figure so you can see how the
   sentence reads when it is complete. Replace it with something a sceptical prospect could
   check. If you do not have one yet, the campaign carries a question instead of a claim, and
   that is fine. Inventing one is not.

2. THE NAMES AND COMPANIES ARE INVENTED. Every person and company in the approved examples is
   made up.

Everything else - the message shapes, the locked phrasings, the slot rules, the QA list - is
the part worth studying.

-----

Shared across three campaigns targeting founders and principals, broad influencers, and B2B
sales leaders. All three run the same "Free Trial - 3 free sales meetings" offer on the same
compressed skeleton. Slot fills change by segment; the locked phrases do not.

Replaces v2 "Curiosity-led Pattern Interrupt" brief.

-----

ANGLE: Simplified outcome-led DM in a compressed format. Direct outcome promise (3 free sales meetings) anchored by a social proof PS and a pre-pay trial structure (prospect doesn't pay anything until meetings are booked). Deliberately simplifies the previous signal-intelligence angle in favour of a direct outcome framing. Acknowledged risk: this shape pattern-matches to other agency DMs the ICP has been burned by. The hedge is the PS credential and the pre-pay trial structure.

ANGLE TYPE: Outcome-led, compressed DM.
OFFER TYPE: Free Trial - book 3 sales meetings free before any contract or payment.

CONNECTION REQUEST: Blank (higher acceptance rate).

PERFORMANCE GATES: 50 connections before judging. Target opportunity rate: set this from your own
prior campaigns. If you have no prior number, write "no benchmark on file" rather than borrowing
somebody else's.

-----

FORMATTING RULES (HARD):
- No em dashes. No en dashes. Hyphens, commas, and periods only.
- Double line breaks between M1 sections (3 visually distinct sections, never a wall of text).
- First name in every message. Company name at least once in M1.
- No emojis.
- Lowercase "free" always. Never "Free" or "FREE" mid-sentence.

-----

M1 STRUCTURE - LOCKED. 3 sections with blank lines between. Only variable text is {first_name}, {observation}, {sub_niche}.

{first_name} - {observation}.

If I offered to book you 3 sales meetings with {sub_niche} for free, would you be open to chatting?

PS - This is a free trial for the same service that helped {proof}.

LOCKED: The offer sentence and the PS structure are locked. Do not paraphrase. The phrasing "book you 3 sales meetings with {sub_niche} for free" must appear verbatim, and the PS must carry your proof point unaltered once it is filled.

M1 WORD COUNT: 50-85 words (soft).

-----

M2 STRUCTURE - LOCKED. List reveal + re-ask. Variable text: {first_name}, {their_target_market}, {sub_niche}, {their_services}.

{first_name} - had a quick look at {their_target_market} & found ~50 {sub_niche} showing buying intent for {their_services} right now.

These are the type of people we'd book you into sales meetings, for free.

Interested?

LOCKED: The second and third sentences are locked. Do not paraphrase. "These are the type of people we'd book you into sales meetings, for free." appears verbatim.

M2 WORD COUNT: 30-60 words.

NOTE ON "~50": that number is a claim. It must come from a search you actually ran for this lead's
market. If you did not run one, the sentence does not go out.

-----

M3 STRUCTURE - LOCKED. Cheeky breakup one-liner. Variable text: {first_name}, {sub_niche}.

{first_name} - so I guess you don't want 3 free sales meetings with {sub_niche}?

LOCKED: Entire message is locked. M3 WORD COUNT: 15-25 words.

DEVIATION NOTE: This M3 does NOT follow the standard M3 architecture (verifiable authority + restated offer + direct CTA). It is a deliberate breakup line. Loss aversion through implied accusation, daring the prospect to correct. If reply rates on M3 underperform in calibration, fall back to the standard M3 pattern.

-----

SLOT GUIDE

{first_name}: The lead's first name, exactly one word, no title prefixes. If profile shows only an initial or name is ambiguous, skip the lead.

{observation}: One specific, discoverable observation about THIS lead or company. 10-25 words. One observation only. Must include company name at least once in M1. Must pass the Swap Test (if you swap in another lead's name, does it still work? If yes, rewrite). Open with "saw", "noticed", or similar natural DM opener. Never fabricate. Never parrot their headline verbatim. No filler comments ("that's impressive", "sharp positioning play"). Research source: LinkedIn profile, company website, recent funding/hiring news.

{sub_niche}: The specific buyer type the prospect sells to. 2-5 words. Must be a person or role description, not a vertical or industry. Extracted from the lead's research notes, company site, or campaign segment definition. If you can't name a specific sub-niche, skip the lead.
  Good: "SMB owners", "healthcare procurement leaders", "healthtech RevOps and sales leaders"
  Never: "decision-makers", "leaders", "companies in your space", "your ideal customers", "your ICP", placeholder text like "[ideal customer]"

{their_target_market} (M2 only): Geographic or vertical descriptor for where the prospect's buyers are. 2-6 words. Paired with {sub_niche} in M2.
  Good: "UK SMBs", "US healthcare systems", "Series A-C healthtech"
  Never: "the market", "your market", "your space"

{their_services} (M2 only): How the prospect describes what they sell. 2-6 words. Must be how THEY would describe their service, extracted from their website/LinkedIn about section.
  Good: "group health and commercial coverage", "language access services", "AI sales intelligence"
  Never: vague descriptors ("business consulting", "B2B services", "tech solutions")

{proof} (M1 PS only): YOUR verifiable result, written once and reused verbatim across the campaign. It sits inside the fixed sentence "This is a free trial for the same service that helped {proof}." It is not a per-lead slot - it is the one thing you fill in before the campaign runs at all.
  Shape that works: a named outcome with a number and a timeframe, client kept anonymous.
  Rule: if a prospect replied "which client?", you must have an answer ready. If you do not, you do not have the claim, and the PS comes out of the message entirely.
  Never: a rounded-up figure, a projection, an average across clients presented as one result.

-----

NICHE NOTES - TONAL GUARDRAILS

- This ICP (founders, principals, sales and marketing leaders at 11-200 employee B2Bs) has been pitched by a great many lead gen agencies this year. The outcome promise format is familiar to them. The differentiator is the pre-pay structure. Lean on that.
- Write as if the sender is typing this himself in a LinkedIn DM. Conversational, natural, direct. Not programmatic. Not advertorial.
- Never use declarative statements about their situation. Use "I'd imagine" framing if elaborating on pain.
- The observation in M1 must feel like a real human wrote it. If it reads like a database output, rewrite.

LANGUAGE TO USE:
- "book you 3 sales meetings" (locked outcome framing)
- "buying intent" (specific, research-coded)
- "for free" (lowercase)
- "open to chatting" (warm, low-commitment CTA)

LANGUAGE TO AVOID:
- Never use: "lead gen", "leverage", "synergy", "best practices", "cutting-edge", "revolutionary", "game-changing", "best-in-class", "hop on a call", "jump on a quick call"
- Never use your own internal positioning phrases in M1/M2/M3. The words you use to describe your buyer to yourself are not customer-facing, and they read as a category label being applied to the reader.
- Never leave merge tags or placeholders literal: {first_name}, [ideal customer], [sub-niche]. HARD FAIL if any reach the prospect.

-----

SOCIAL PROOF RULES

- One approved credential for this angle, filled into {proof} before the campaign starts.
- Must appear verbatim in the M1 PS every time. Do NOT paraphrase the figure or the timeframe.
- Client name stays vague ("another client") to protect their identity.
- Do not stack multiple proof points in one message. One credential per sequence.
- If a prospect asks in reply "which client?" - have a prepared answer. Name them if you have permission; otherwise describe the vertical and result anonymously.
- If you cannot fill {proof} with something real, delete the PS. A campaign with no proof line still works. A campaign with an invented one is a claim the prospect can check and you cannot.

-----

PRE-SCRIPT EXCLUSIONS

- No lead gen agencies or B2B outbound consultancies (competitor territory - they know the playbook)
- No one already running campaigns like this one (check their recent posts)
- No companies under 11 employees or over 200 employees
- No companies outside the campaign's geography
- No leads where you cannot identify a specific {sub_niche} (generic fills kill the message)
- No leads at a company you have contacted before on another campaign. Check first, exclude, and flag for manual review.

-----

HANDLING REPLIES

IF THEY REPLY "send me info": Do NOT send a deck. Send a one-screen explanation of how the pilot works (we run 10 days of outreach, target ~200 {sub_niche} matching their buying signals, book 3 meetings before any payment). End with one qualifying question: "What's your current approach to reaching [sub-niche]?"

IF THEY REPLY "is this a guarantee?": Soften the framing. "Not a contractual guarantee, but our typical campaign in this segment books 2-4 meetings in the first 200 outreaches. Happy to show you the case study." (Only use that sentence if those are your real numbers.)

IF THEY REPLY "how much does it cost after the trial?": "Our standard engagement is a quarterly retainer after the 3 free meetings, but we can scope that on a call once you see the quality of the meetings we book."

IF THEY REPLY WITH A NO: Grace-out. "No worries - door's open if that ever changes. Happy to share signal patterns from your market if it's ever useful." Move to closed.

-----

QA CHECKS (MECHANICAL)

- Word count M1: FAIL if <50 or >85 words
- Word count M2: FAIL if <30 or >60 words
- Word count M3: FAIL if <15 or >25 words
- CR is blank: FAIL if populated
- No em dashes: FAIL if em dash found anywhere
- No en dashes: FAIL if en dash found anywhere
- M1 contains company name: FAIL if company name absent from M1
- First name present: FAIL if {first_name} unfilled in any message
- Locked phrasing verbatim: FAIL if the offer sentence or PS is altered
- PS present in M1: FAIL if PS missing, unless {proof} is deliberately empty and the PS was removed for the whole campaign
- {proof} filled: FAIL if the PS ships with the slot name still in it
- M2 locked sentences verbatim: FAIL if "These are the type of people we'd book you into sales meetings, for free." is altered
- M3 verbatim shape: FAIL if M3 deviates from "{first_name} - so I guess you don't want 3 free sales meetings with {sub_niche}?"
- Placeholder leakage: FAIL if a brace, a bracket, or a literal slot name survives into the output
- No kill-list words: FAIL if "lead gen", "leverage", "synergy", "best-in-class" appear
- Lowercase "free": FAIL if "Free" or "FREE" appears mid-sentence

QA CHECKS (CONTENT)

- SWAP TEST: Remove {first_name}, {observation}, {sub_niche} from M1. If the remaining message still applies to 5+ other leads in the batch, FLAG the observation for rewrite.
- SUB-NICHE SPECIFICITY: Verify {sub_niche} is role-based and specific, never generic.
- OBSERVATION QUALITY: Verify {observation} is anchored in something publicly discoverable. Flag fabricated or generic observations.
- COMPANY-NAME ONCE: Company name appears once in M1, never more. Use "your team" after first mention.
- CROSS-MESSAGE REPETITION: M2 must introduce "~50 {sub_niche} showing buying intent" as genuinely new info, not a restatement of M1's offer. M3 must not repeat M2's list claim.
- READ-ALOUD TEST: Read M1 aloud. Would you say this to someone you had just met? If it sounds like a cue card, rewrite. M1 must read naturally across the three paragraphs, with no abrupt tonal shifts.

-----

APPROVED EXAMPLES (from calibration)

INVENTED. The people, companies and the proof figure below are all made up. They show the shape
of a correctly filled skeleton, nothing more.

Marcus Hale (Brightsure, B2B insurance, UK SMBs - tripled sales team in 6 months across group health, commercial and comp coverage):

M1:
"Marcus - saw you've roughly tripled Brightsure's sales team in six months across group health, comp and commercial coverage.

If I offered to book you 3 sales meetings with SMB owners for free, would you be open to chatting?

PS - This is a free trial for the same service that helped another client book 23 meetings in their first 30 days."

M2:
"Marcus - had a quick look at UK SMBs & found ~50 SMB owners showing buying intent for group health and commercial coverage right now.

These are the type of people we'd book you into sales meetings, for free.

Interested?"

M3:
"Marcus - so I guess you don't want 3 free sales meetings with SMB owners?"

-----

Dana Kessler (Lingua Bridge, VP CX and Marketing, 9k+ employees after three acquisitions, sells language access into US healthcare and legal):

M1:
"Dana - saw Lingua Bridge has had a step-change year with three acquisitions and headcount past 9k, selling language access into healthcare and legal.

If I offered to book you 3 sales meetings with healthcare procurement leaders for free, would you be open to chatting?

PS - This is a free trial for the same service that helped another client book 23 meetings in their first 30 days."

M2:
"Dana - had a quick look at US healthcare systems & found ~50 healthcare procurement leaders showing buying intent for language access services right now.

These are the type of people we'd book you into sales meetings, for free.

Interested?"

M3:
"Dana - so I guess you don't want 3 free sales meetings with healthcare procurement leaders?"

-----

Priya Raman (Kestrel Signal, Founding AE, Series A AI sales intel platform, sells into healthtech RevOps and sales teams):

M1:
"Priya - noticed you're the first growth hire at Kestrel Signal, selling AI sales intel into healthtech RevOps and sales teams.

If I offered to book you 3 sales meetings with healthtech RevOps and sales leaders for free, would you be open to chatting?

PS - This is a free trial for the same service that helped another client book 23 meetings in their first 30 days."

M2:
"Priya - had a quick look at Series A-C healthtech & found ~50 healthtech RevOps and sales leaders showing buying intent for AI sales intelligence right now.

These are the type of people we'd book you into sales meetings, for free.

Interested?"

M3:
"Priya - so I guess you don't want 3 free sales meetings with healthtech RevOps and sales leaders?"

-----

CAMPAIGN STATUS

Written to the campaign block in campaigns.md:

status: paused (a human sets it active)
fill method: search
offer: Free Trial - book 3 sales meetings free before any contract or payment
band: (blank, no data yet)
last pulse: (blank until PULSE runs)

-----

OPEN FLAGS FOR CALIBRATION

1. PROMISE VS DELIVERY GAP: M1 promises "book you 3 sales meetings" but the actual deliverable is "we run the campaign, you work the inbox." Watch for replies asking "so you book the meetings or I do?" during calibration. If this surfaces, rewrite M1's offer sentence to "If we ran a free trial campaign that booked you 3 sales meetings...".

2. M3 AUTHORITY GAP: M3 deviates from the standard M3 architecture (no verifiable authority claim, no direct CTA). If M3 reply rates underperform, test a fallback variant carrying a real number you can back up, then a direct ask.

3. PROOF TIMEFRAME MISMATCH: If your {proof} is a long-horizon outcome paired with a short trial, sharp prospects will notice. Have a shorter-horizon version of the same result ready for the follow-up.

4. PATTERN-MATCH RISK: This format looks similar to other agency DMs. Monitor whether reply sentiment skews sceptical or engaged. If sceptical, the pre-pay PS isn't landing hard enough and needs to move up into the body.
