Harborline Advisors — Campaign Brief

The read

Five things about the positioning. Two of them will decide whether this campaign works at all.

Four unrelated verticals off one brochure is a firm description, not a wedge. Crypto, startups, real estate and Gulf Bay small business, with accounting, CFO, tax and planning inside each. For this campaign exactly one line survives: a CFO team that has sat inside venture-backed companies. Everything else comes off the sender profile before a connection goes out.
The Bayside address is a liability everywhere except one cell. A founder in SF, New York or Austin reads a Florida CPA firm as regional and cheap. Suppress geography everywhere except Gulf Bay issuers, where the local line is true, speakable and earns trust.
The startup vertical has zero published proof. No logos, no named case studies, no stage playbooks, while the firms this founder will Google publish all three plus pricing. Either get written permission to name two venture-backed clients, or accept the handicap knowingly. Until that exists the artefact is the proof, so the artefact has to be genuinely good.
Reid's short-form video already drives most of their top-of-funnel and it isn't wired into outbound at all. The sending profile has to be his personal account. A founder who gets the DM and then finds three videos of Reid explaining runway converts at a different rate to one who finds a services page. Keep the post cadence running through the campaign window.
Their intake form gates on annual revenue, which is exactly wrong for this buyer. A just-raised seed company has $0 revenue and $4M in the bank, so it lands in the bottom bucket or gets dropped. If that routing isn't changed before send, the campaign dies at handoff and the reply rate will look like a messaging problem when it's a plumbing problem.

Worth knowing: the LinkedIn company page under "Harborline Advisors" belongs to a different business - an unrelated CRM consultancy in another country. Anyone searching them on LinkedIn lands on the wrong company. That's a real problem for a campaign that sends people to look them up.

Why this offer

Three directions were designed, steelmanned and argued against each other. Judged on cold reply rate, how little the prospect has to do, whether the work can be done at volume, proof support, and reachable segment size.

	Direction	Verdict
1	Form D runway read Just-raised founders, sourced from the filing	Chosen. The only direction whose buyers actually live on LinkedIn, whose source is machine-readable enough to run at volume, and whose trigger has a fixed date that turns a founder into a buyer. Sourced from the filing rather than the press, so M1 carries a public fact the founder didn't know was public.
2	On-chain event map Crypto treasuries and token projects	Strongest single message of the three and the smallest reachable list. After the self-published-address guardrail it collapses to DAOs, protocols and token projects, and the treasury signer is frequently pseudonymous or absent from LinkedIn. Kept as a high-value cell inside the Form D machine, so the crypto division still gets used at no extra sourcing cost.
3	Public-record portfolio sheet Florida real estate investors	The deepest free data of the three, and it fails on channel. Landlords and small syndicators are the softest LinkedIn segment Harborline Advisors has, and an offer nobody accepts a connection request to is worth zero. County appraiser sites also don't scale. Its entity-resolution discipline was kept.

The offer is a dangle, not a giveaway. Updated 12 Aug. We say the page exists and offer to walk through it. We do not attach it, link it or email it. The artefact is the reason for the meeting, not a substitute for it.

READTheir Form D on EDGAR. Free, public, filed by them.

BUILDA one-page runway curve with every assumption printed on it.

WALKTwenty minutes. They can argue with the assumptions.

Honest trade-off: asking for twenty minutes rather than permission to email a file will sit below a pure give-it-away offer on raw reply rate, and convert better per reply. A founder who argues with your assumptions on a call is already in the engagement.

The messages

Blank connection request. Sends from Reid's personal profile. The raise figure is the only number taken from the lead, and it's named as public. Everything else is a stated assumption the sender owns.

M1 · the filing55 frame · ~59 filled

Hi {{first_name}}, saw {{company}}'s Form D from {{filing_month}}. It's a public filing, so I used the number in it to put together a one page runway read for you. Every assumption is written on the page so you can argue with mine. Easier to walk you through it than email it over. Want me to?

"So you can argue with mine" is the line that makes this safe. It hands the founder authority over the model, which stops the message asserting anything about a cash position the sender has never seen.

M2 · the second signal46 frame · ~51 filled

{{first_name}}, one more public thing, and it's not from the filing. {{company}} has a live post for {{role}}. A hire like that moves the burn line, so I added a second version of the page with it in. Twenty minutes and I'll take you through both.

"It's not from the filing" is deliberate. An earlier draft opened with "second thing from the same filing" and then delivered a job post, which lied about its own provenance. One observation only, and it's a statement about the model rather than about their actual burn.

M3 · the close38 words

{{first_name}}, our CPAs and CFO advisors come out of public accounting and venture-backed startups, and clients typically save 50 to 80 hours a quarter. Say the word and I'll find twenty minutes to walk you through the numbers.

Both claims are lifted straight from Harborline Advisors's own published copy. There is deliberately no slot here: an open authority slot is where an invented statistic gets in at fill time. The crypto cell swaps in the October 2025 crypto division, the Gulf Bay cell swaps in the local line.

The recon

Six free lookups per lead, about twelve minutes, all before a word is written. No paid vendor at any step.

1 SOURCE THE RAISE FROM THE FILING, NOT THE PRESS SEC EDGAR daily index every business day, filter form types D and D/A, fetch primary_doc.xml. Free, no key, no vendor. It is the raise itself rather than someone's coverage of it. No filing but a publicly announced round -> NO_FILING_L1 cell, lower priority, capped at 25% of weekly volume. 2 READ THE FIELDS Item 7 date of first sale Item 13 offering / sold / remaining Item 14 investor count Item 3 related persons + address Item 4 industry group Item 5 issuer size Amount sold outside $1M-$15M -> drop. Pooled investment fund -> drop. Remaining > 0 -> flag ROUND_OPEN, changes the M2 finding. 3 RESOLVE THE ENTITY TO A HUMAN, HARD GATE Item 3 executive officer -> LinkedIn, confirmed by a second independent source: team page, state registry, or a press mention. If the match is not certain the lead is DROPPED, not guessed. 4 GET THE HIRING PLAN - the second signal Careers page and public ATS board. Open finance reqs, and posted salary bands where a pay-transparency state applies. 5 STATE REGISTRY STATUS Sunbiz for Florida, the equivalent Secretary of State elsewhere. Entity status, formation date, annual report standing. If delinquent, do NOT use it - fall back to role mix. 6 CELL ROUTING FORM_D_CORE default, full artefact FORM_D_CRYPTO web3 issuer, adds an on-chain treasury read, crypto division becomes the M3 authority FORM_D_LOCAL_GULF BAY Pinellas / Hillsborough / Pasco / Manatee, adds the local firm line NO_FILING_L1 announced round, no findable filing

Targeting

TITLES Founder, Co-Founder, Founder and CEO, CEO, President, COO Preference: the person named in Form D Item 3 with an executive officer flag, matched to LinkedIn. EXCLUDE CFO, VP Finance, Head of Finance, Finance Director, Controller, Accounting Manager, Bookkeeper. NOT just excluded as recipients: if ANY current employee holds one of these on the company page, the WHOLE COMPANY is dropped. The offer is built for a company with no finance leader. SIZE 5-60 employees on LinkedIn. Form D amount sold $1M-$15M. Below $1M there is no budget for CFO work and the funnel fills with the cheapest product. Above $15M they hire a full-time VP Finance within the quarter. GEOGRAPHY United States only - Form D is a US filing. SF Bay · NYC · Austin · Boston · Seattle · Denver · LA · Chicago · Miami-Fort Lauderdale · Gulf Bay Bias ~60% of volume to California, Colorado, New York, Washington, Illinois, Hawaii and Maryland: posted salary bands are law there, so the burn side is built from real numbers rather than benchmarks.

Signals - the two-key rule

Captured, primary. A Form D or D/A with date of first sale between 14 and 70 days ago. Dated, public, machine-readable, with a fixed clock attached.
Created, the second key. An open req for Controller, Head of Finance, Accounting Manager or first finance hire, or a finance hire whose start date falls in the last 60 days.
The stack rule, hard. A closed round alone is not a trigger and never sends. Raise plus finance signal equals confirmed intent. This single gate is what separates the campaign from the congratulations wave that hits every founder the fortnight after a round.
Timing edge. Date of first sale frequently precedes press coverage by two to six weeks. Where the filing leads the announcement, flag PRE-ANNOUNCEMENT and move to the front of the queue.
Anti-signal, drops the lead. An existing CFO, VP Finance or Controller on the company page. The forcing function has already been answered internally.
Anti-signal, drops the lead. More than ten weeks past first sale. The lead investor's preferred-vendor list has been worked and a cold DM is now late.

Before it sends

Fix the intake routing first. This one kills the campaign silently.Replies must bypass the Speak With An Advisor form entirely and land with Reid directly. That form gates on estimated annual revenue, and a just-raised seed company with $0 revenue and $4M in the bank gets misrouted or dropped. Confirm it's changed and tested before the first send.

Nothing is attached, linked or emailedThe page exists to earn the twenty minutes. If a reply asks for it by email, the answer is a time offer, not an attachment. Brief this before launch or the campaign quietly becomes a PDF drop and the meeting rate collapses.

The page exists before M1 goes outBuilt and saved per lead, with every input printed on it: the cash figure and its Form D source, headcount, each open role and the band used, and whether that band was posted by law or level-benchmarked. M1 says it exists. A yes with nothing to show ends it.

Two signals, or it does not sendA Form D with first sale 14 to 70 days ago AND a created finance signal. A closed round on its own is the congratulations wave, and every other service provider is already in that inbox.

Identity certaintyThe Form D officer must resolve to a LinkedIn profile confirmed by a second independent source. If the match is not certain the lead is dropped, not guessed. A wrong name in message one is a fabricated observation.

Proof exclusionThe $25K to $100K tax and inefficiency saving is blocked on any issuer declaring no revenue. The founder's obvious next question is "saved on what tax?" and there is no honest answer for a company with no profit.

Sender is Reid, personallyNot a firm page, not a junior. Headline and About rewritten to lead on the venture-backed CFO line, Bayside suppressed outside the Gulf Bay cell, video cadence continuing through the window.

Volume honestyForm D coverage is thin after this much filtering and the list will be small. It does not get backfilled with demographic leads to hit a number. If qualified volume falls under roughly 40 a week, the answer is more markets or a second cell, never a looser gate.