Skip to content
Shepherd GTM

Targeting

The cockpit for who gets reached, on which channel: slice the broker universe any way the data allows, then point LinkedIn or email at exactly that audience. The unit is the account: an office of 10 to 50 brokers, not a person.

450 brokers behind this audienceStored

What this is

A filter over the same scored universe as Play 1, collapsed down to two things LinkedIn can target on: a company list and a set of job titles. Nobody on this page gets contacted from this page.

What it is not

Not a CRM and not a second copy of Pipeline. There is no per-person record here, no email, no cadence. It exists to produce three CSVs an ad platform can read.

Companies to target
18
LinkedIn company list - the account is the office, not the person
Brokers behind the audience
450
of 450 in the universe
Decision makers
319
titles that own the placement decision
Reachable by email
271
verified address on file, the email channel's ceiling

Describe the audience

The model only sets the filters below. It never contacts anyone.

Build the audience

Wider is your first half: offices inside appointed brokerages that have never sent a risk. Deeper is the second: brokers who already submit and could submit more. Fresh signal narrows to firms the Signal Engine flagged in the last 30 days, so paid spend lands where something just happened.

Point a channel at it

LinkedIn ads

works today

Campaign Manager wants a company list and job titles, not email addresses, which this universe produces without spending an enrichment credit. Upload the company list as a Company List audience, paste the titles into title targeting.

Email, through AgentMail

planned

271 of this audience have a verified address, and AgentMail is the sending layer I would wire: API-first inboxes built for agents, so every draft the Signal Engine writes goes out from a real, warmed mailbox with replies threading back in. Sending stays inside the Sequences approval gate - this page picks who, never fires the send itself.

Draft in SequencesAgentMail

Company list export, preview

  • Locktonfresh signal
    Los Angeles, CA
    51
  • USI Insurance Servicesfresh signal
    office not recorded
    44
  • Marsh
    Atlanta, GA · Philadelphia, PA · New York, NY
    44
  • Gallagherfresh signal
    office not recorded
    39
  • WTW
    Houston, TX · Atlanta, GA · Birmingham, AL
    39
  • Aonfresh signal
    New York, NY
    34
  • IMA Financial Group
    Portland, OR
    29
  • Alliant Insurance Servicesfresh signal
    Dallas, TX
    28
  • Holmes Murphy
    Denver, CO · Waukee, IA · Sioux Falls, SD
    24
  • CAC Group
    Franklin, TN · New York, NY · Houston, TX
    22
  • Marsh McLennan Agency
    St. Louis, MO · Portland, ME · Clearwater, FL
    21
  • McGrifffresh signal
    Birmingham, AL · Houston, TX
    20
  • HUB International
    Chicago, IL
    17
  • NFP
    New York, NY · Houston, TX
    15
  • Brown & Brown
    office not recorded
    9
  • Newfront
    Encino, CA
    7
  • Heffernan Insurance Brokersfresh signal
    Petaluma, CA
    5
  • Woodruff Sawyer
    office not recorded
    2

Title targeting export, preview

  • Client Executive, Surety/Shareholder6
  • Construction Practice Leader3
  • Vice President of Construction3
  • Vice President - Construction Services Group3
  • National Energy Practice Leader3
  • Executive Vice President, Oil & Gas Co-Leader3
  • National Construction Practice Leader2
  • Senior Vice President, Construction Services Group2
  • Senior Vice President - Construction Practice2
  • Senior Vice President, Construction & Infrastructure2
  • Vice President, Construction2
  • Vice President, Construction Practice2
  • Assistant Vice President - Construction2
  • U.S. Energy Practice Leader2

403 more in the export.

What I would actually run

  • One audience, two channels. The same filtered list feeds the LinkedIn flight and the email sequence, so when an office converts there is no argument about which list found it - there was only one list.
  • One campaign, one control. Half the target offices get the paid flight, half do not, and both halves get the same Play 1 email. The gap between them is the whole answer, and without the control there is no answer.
  • $500 to $1,500 a month. Small enough to kill without a meeting. Logged in the Spend tab against Play 02 so cost per submitting office is a real number rather than a feeling.
  • Kill rule. 90 days, no lift over the control, the budget moves into Play 1 data instead.