GHL Prospecting Outreach Cadence SOP
Overview
What: the multichannel outreach cadence that works a cold prospect from a verified audit through to a booked meeting or a clean close.
When: a prospect exists in the sub-account with a verified audit and at least one verified contact point.
Who: the Implementation Lead.
Time required: roughly 10 minutes of active work per prospect across the full cadence, spread over 11 days.
This SOP does not cover: what to say (see Prospecting Cold Call Scripts and Prospecting Outreach Email Templates), how to run a call block (see Prospect Phone and Call Execution SOP), or what is legally permitted (see SC Outbound Calling Compliance Standard).
Architecture
Prospecting runs inside the Sales Pipeline, not alongside it. A prospect enters at New Lead with Source: Prospecting. When the report email goes out, the opportunity moves to Audit Sent, and it stays there for the rest of the cadence. The stage answers one question, whether the audit has been sent. Everything else about outreach (replied, deferred, re-activated, suppressed) is carried by tags.
| Stage | Means | Moved by |
|---|---|---|
| New Lead | Verified prospect, report not sent yet | The prospecting agent, on creation |
| Audit Sent | Report generated and the cadence running, waiting on a reply | C-021.Prospecting - 01.generate audit (WF-1), when it generates the report. ⚠️ Since 2026-11-04 that run sends no email; the email is Email 1A, sent later by the caller. Recorded as the move existed before the change; workflow definitions cannot be read over the API, so confirm in the builder |
| Appointment Set | A specific date and time agreed and booked | A human, by hand |
| Discovery | Passed the ICP screen, and the call was held | A human, per §Qualification |
| Closed Lost | An explicit no, a failed screen, or the cadence ended. Status set to Lost | A human |
⚠ ️ The diagram below is the intended model, not what runs today. Four of the tags it names do not exist in the sub-account, and "run the cadence, touches 1 to 5" is a single box covering one automated touch and four manual ones. For what is actually built, see the as-built map in the C-021 spec §2.1. Read them together: this one is the destination, that one is the position.
Why one stage, and tags for the rest. A prospect that has been called twice and emailed once is not at a different point in the sales process than one that has been called once. Attempt state is attribute data, and attributes belong on tags and fields, so there is no stage per touch. Receiving the audit is different. It is the one hand-over every prospect gets, and "who has had the report" is the question the pipeline is asked most, so it gets a stage of its own. Audit Sent carries a 10% probability, just above New Lead's 5%, so cold prospects barely move the forecast.
The cadence
Email first, then a warm call. The first email hands over the audit report with no ask beyond a reply, so every later touch, spoken or written, follows delivered value instead of interrupting cold. This is an operator-fit decision: the call-first sequence outperforms on paper (80 percent of the highest-performing sequences in the 3.4-million-cadence analysis open with a call), but a cadence executes only as well as its operator sustains it, and the call here is deliberately warm, optional, and skippable. The call-first variant remains available.
| Touch | Day | Channel | Content | Precondition |
|---|---|---|---|---|
| 1 | 0, once the report is checked | Email, sent by rule | The person who checks the report adds prospecting-values-added, and C-021.Prospecting - 02.send first email sends First Email - Offer, First Email - Audit Only or nothing, by the rule in the C-021 workflow spec | Report generated and checked, email on file |
| 2 | 3 | The bucketed finding touch (template 9 or 11-14): the hand-counted finding, for a prospect who has been slotted. It does not re-deliver the report, which arrived at touch 1. This is the touch the slotting work exists for | Audit verified, bucket assigned, finding recorded | |
| 3 | 7 | Call (optional) | Warm follow-up referencing the delivered report (S2 script). Voicemail only on call attempts 1 and 4 (§6) | Verified phone, outside FL/OH/PA, 09:00-20:00 local, weekday. Skippable without breaking the cadence |
| 4 | 10 | Final touch (Template 5). Graceful close | No response | |
| - | 12 | - | Tag prospecting-not-ready-now, clear prospecting-outreach-active | No response |
| 5 | 60 to 90 | Re-activation. Refresh the audit first | ICP fit was strong, one attempt only |
Changed 2026-11-04: touch 1 is sent by the caller, and it carries the offer. Tagging no longer sends the report. Until the rest of this cadence is re-cut for the new first email: touch 2 (the bucketed email) is not sent to anyone who received Email 1A, because it repeats the report and says nothing about the offer; touch 3 is the S2 call; touches 4 and 5 are unchanged. The two paragraphs below record why touch 1 became one email, before it moved to the caller.
Why one email rather than two. Until 2026-10-30 touch 1 announced a report and touch 2 delivered it the next morning. The split was meant to remove a duplicate hand-over and buy a window for a human to record a finding; the window argument was withdrawn within hours of being made, when the gap collapsed from three days to overnight. What remained was a second step that can fail silently, and does: message history to 2026-10-29 shows three prospects who were promised a report and never received one. A single send has no gap to lose anyone in, and zero replies across every send of the pair means there was nothing in the two-step structure worth protecting.
Why immediately rather than overnight. The overnight wait existed to honour an announcement at a respectable distance. With nothing announced there is nothing to honour, and the objection that a fifteen-minute gap "reveals that no human was involved" does not apply to a single cold email that never claimed one. Sending on tagging buys two things: the prospect gets the report while the trigger that selected them is still true, and call situation S3 survives — the caller tags, waits five minutes, dials, and the report is genuinely in the inbox.
Three emails before the final touch, not four. Touches 1, 2 and 4 carry email. The stop rule below returns to three, reversing the amendment the announce-and-deliver pair required.
A reply on any channel gets a same-day response on that channel and moves the prospect to RESPONDED; remaining touches stop.
Excluded-state note: the state exclusion is a calling restriction. Email-first makes Florida, Ohio and Pennsylvania prospects workable as email-only cadences (skip touch 3). New York left the excluded list on 2026-09-25 and is a normal calling state.
The call-first variant
When a prospect warrants opening with a call (or you simply choose to), the previous sequence stands: call with opener A/B/C, voicemail 1 plus the paired audit email within 30 minutes (Template 7), day-2 call, day-5 call with voicemail 2 plus paired email, day-9 final touch. Scripts and voicemail rules are in the Cold Call Scripts. The stop rules below bind both variants.
The re-activation touch
Touch 6 is a nine-word email, not a re-pitch. Subject line is the prospect's first name. Body is one short question, no signature block, no links.
Still worth fixing how you show up on Google?
This format is a reactivation tool and it is the correct use of it. It fails when it is used as a first touch, so never open with it.
Stop rules
| Rule | Limit |
|---|---|
| Emails in the initial cadence | 3 (touches 1, 2 and 4) |
| Emails to the same prospect per day | 1 |
| Voicemails in the email-first cadence | 2 at most, on call attempts 1 and 4 (Cold Call Scripts §6), with nothing paired |
| Dial attempts per prospect | 5, either variant. 98 percent of conversations that will ever happen have happened by then |
| Re-activation attempts | 1 |
| After an explicit no or an unsubscribe | Stop every remaining touch immediately |
| After a do-not-call request | Stop permanently, every channel |
Weekday sends only, best Tuesday to Thursday 9:00 to 11:00 local. Calls between 08:30 and 20:00 recipient local time, weekdays. No dials to Florida, Ohio or Pennsylvania (email is fine).
Tags
Outreach state lives here. Exactly one state tag applies at a time; the marker tags are additive.
State tags
| Tag | Applied when | Cleared when |
|---|---|---|
prospecting-outreach-active | Touch 1 is placed. Applied by the touch-1 workflow since 2026-09-17, not by hand | Cadence ends by any route |
prospecting-responded | Prospect replies or engages on a call | Advanced to Discovery, or closed |
prospecting-not-ready-now | Day 12 with no response, or an explicit deferral | Re-activation begins, or closed |
prospecting-qualified | Passes the ICP screen and advances | Terminal |
prospecting-lost | Closed without advancing | Terminal |
Marker tags
| Tag | Meaning |
|---|---|
prospecting-contact-verified | A human confirmed the phone or email is real. Shared with the discovery SOPs |
prospecting-audit-verified | The audit passed the quality checklist |
prospecting-cadence-start | The deliberate start signal, applied by a human after both verification tags are on (see Audit Accuracy SOP). Whether the cadence runs by hand or by the C-024 workflows, this tag is what starts it |
prospecting-reactivated | The single permitted re-activation touch has been sent. Its presence blocks a second — the stop-rule counter for re-activation |
sc-suppressed | Do not contact on any channel, ever. Set by a do-not-call request. Never removed |
sc-suppressed is checked before every touch on every channel. It is the one tag that overrides everything else in this document.
⚠️ Most of the tags in this table do not exist in the sub-account. Verified live 2026-09-23. Five prospecting tags exist — new-prospect-received, prospecting-contact-verified, prospecting-audit-verified, prospecting-outreach-active and prospecting-audit-email-opened. Everything else above, including sc-suppressed, has never been created.
This table describes the intended model and is not wrong; it is simply ahead of the account. Until the gap closes:
- Create a tag on first use, spelled exactly as written here. A tag typed from memory as
suppressedordo-not-contactis a tag nothing will ever find, and the state model depends on exact names. - A do-not-call gets DND set on the contact as well as the tag. DND is enforced by the platform; a tag is a label that something has to remember to read. Nothing currently reads this one.
- Nothing machine-checks suppression today. The published report workflow has no suppression node and no tag for one to query. The First Touch SOP step 2d is the only enforcement point, and it is a human reading a list.
Tracked as S1-1 and S1-2 in the follow-up sprint; the full live-vs-canon tag comparison is in the structure inventory §4.1.
Custom fields
All on the contact — five fields, per the Prospecting Outreach Automation PRD data model.
| Field | Type | Purpose |
|---|---|---|
Last Disposition | Dropdown | The thirteen values in the call execution SOP |
Opener Used | Dropdown: A / B / C | Which opener was used, so performance can be compared |
Audit Finding Used | Text | The specific finding led with. Keeps call and email consistent |
Audit PDF URL | Text | The shareable audit-report link, delivered by touch 1. Must be non-empty before the send — the workflow guards on it, because an email whose only purpose is this link is worthless without it |
Call Attempt Count | Number | Against the cap of 5 |
Field availability (verified 2026-09-08). All five fields now exist on the live sub-account. Four of them were missing that morning and were created the same day:
Audit Finding UsedN9WJ3EXcTVb5aQO04bBf,Opener UsedVbjA2mjolA06RsprJeqZ,Call Attempt CountVoh0LwzP8BVOCtvvUT4T,Last DispositionbEXnOD6DOmRJVzkUvqTb(the twelve SOP values).Audit PDF URLisvvqfkt1kTpf7Qfad2Zpa. Ids and the sandbox gap are in the structure inventory §5.Tradeexists but is populated on zero prospects, and Template 8 requires it - see S2-5 in the channels backlog.
State the cadence needs but no field carries:
| State | Where it lives |
|---|---|
| Prospect state and timezone | The native contact State and Contact Time Zone fields, set at verification. The excluded-state guard and the calling window read these |
| Voicemail count | Structure. Only call attempts 1 and 4 script a voicemail, so the cap of 2 holds when the scripts are followed |
| Audit freshness | Checked by a human against the audit itself before re-activation (Audit Accuracy SOP) |
| Re-activation used | Marker tag prospecting-reactivated |
| How the contact point was found, and the niche | The contact's verification note, using the Contact Identified Via tier vocabulary shared with both discovery SOPs |
| Deferral reason | The body of the re-engagement reminder task |
Advancing a prospect
A prospect leaves prospecting and becomes a real opportunity when it clears the ICP screen.
Qualification screen
| Check | Pass |
|---|---|
| Revenue | $100,000 to $2,000,000 |
| Headcount | 2 to 50 |
| Vertical | Home services, legal, real estate, or an adjacent professional service |
| Value of one new customer | $3,000 or more |
| Lead handling today | A real gap exists, after hours or in follow-up |
| Authority | Owner, or a named decision maker |
Full criteria and the disqualifier list are in the Ideal Customer Profile.
On pass
- Move the opportunity from Audit Sent to Appointment Set once a date and time are agreed (or from New Lead, if the prospect replied before the report went out). It moves on to Discovery only when the meeting is actually held
- Apply
prospecting-qualified, clearprospecting-outreach-active - Set opportunity value from the expected tier
- Book the discovery call and prepare using the Discovery Call Prep Template
- The Symphony Core Sales Pipeline SOP governs from here
On fail
Move the opportunity to Closed Lost, set its status to Lost, and record the reason in loss_reason. Apply prospecting-lost. Do not re-engage a prospect that failed the ICP screen; that is different from one that simply did not respond.
GHL configuration
| Setting | Value |
|---|---|
| Sub-account | Symphony Core LLC (vOyZYmChRuyjFM0yUtba) |
| Pipeline | Sales Pipeline |
| Entry stage | New Lead. Moves to Audit Sent when 01.generate audit (WF-1) generates the report. Since 2026-11-04 no email goes out at that point (see the Architecture table) |
| Opportunity source | Prospecting, set automatically when a prospect is added from the Prospecting tab |
| Outbound calling number | The caller's assigned number — see the SC Phone Number Register |
Voice trust registrations, verified 2026-08-18: SHAKEN/STIR, CNAM, and Voice Integrity are all Approved, as are the A2P messaging brand and campaign. Outbound calls carry authenticated, branded caller ID.
Any automation of this cadence is gated behind the Prospecting Outreach Automation PRD.
Touch 1 is now automated and published. As of 2026-09-17 a workflow generates the audit report and emails its link when a prospect is tagged; the human surface is that one tag, and the procedure for applying it is the Automated First Touch SOP. As of 2026-10-30 that one email carries the report rather than announcing it. Everything from touch 2 onward still fires only when a person acts. The node design is the C-021 Prospecting Workflow Spec, which supersedes the C-024 spec.
Verification
Per prospect
- Audit verified before the first touch
- Contact point verified and
prospecting-contact-verifiedapplied -
Audit Finding Usedis populated and matches what was actually said - Exactly one state tag applies
-
Call Attempt Countis within the cap of 5, and no more than two voicemails were left - No archived after-call or after-voicemail template was sent
Weekly
| Metric | Target |
|---|---|
| Call connect rate | 12 to 22 percent |
| Conversation to meeting | 5 to 10 percent |
| Dials per booked meeting | 100 to 370 |
| Email open rate | 40 percent or better |
| Email reply rate | 5 to 10 percent |
| Prospects advanced to Discovery | Tracked, no target while the motion is new |
| Suppression requests | Under 2 percent of connects |
Review these weekly during an active calling period. Change one variable at a time; changing the opener and the target list in the same week tells you nothing.
Related Documents
- Prospecting Cold Call Scripts - the spoken touches
- Prospecting Objection Handling - responses
- Prospecting Outreach Email Templates - the written touches
- Prospecting Email Outreach Strategy - why the email touches are shaped as they are
- Prospect Phone and Call Execution SOP - phone verification, call blocks, dispositions
- Prospect Email Discovery SOP - finding an email when one is needed
- GHL Prospecting Audit Accuracy SOP - the upstream quality gate
- SC Outbound Calling Compliance Standard - legal constraints on the call touches
- Symphony Core Sales Pipeline SOP - the pipeline this runs inside
- KB-015: GHL Prospecting Tool - how prospects are generated
- Prospecting Outreach Automation PRD - the automation gate
- C-024 Prospecting Outreach Workflow Spec - the workflow design that implements this cadence
Document History
| Version | Date | Changes |
|---|---|---|
| 3.7 | 2026-11-12 | Calling window is 08:30 to 20:00, following the compliance standard. The Implementation Lead set the first dial at 8:30 am on 2026-11-12 |
| 3.6 | 2026-11-09 | Touch 1 is the first email sent by rule once the report is checked, not the caller's Email 1A. Touch 3 is the S2 call; voicemails follow the scripts (attempts 1 and 4, cap 2, nothing paired) instead of zero |
| 3.5 | 2026-11-04 | Touch 1 moves to the caller and carries the Results-First offer. WF-1 is renamed 01.generate audit and generates the report without sending it; Email 1A (Template 20) goes out from 02.send email when the caller adds prospecting-values-added, minutes before an S3 call. Audit Sent's mover is restated with that caveat. Until the cadence is re-cut, touch 2 is withheld from anyone who received Email 1A and touch 3 becomes the S2 call |
| 3.2 | 2026-10-27 | Stage map gains Appointment Set between Audit Sent and Discovery, and Discovery is redefined as held rather than booked — the two were the same column, which hid show-rate. The qualification step now moves a prospect to Appointment Set on booking and to Discovery only once the meeting happens. Last Disposition is thirteen values, not twelve |
| 3.1 | 2026-10-09 | Touch 2 becomes the automated report delivery at 09:00 the next business day, and the bucketed finding email moves to touch 3 on day 3, where it no longer re-delivers a report the prospect already has. Later touches renumber. Records why overnight rather than three days or the fifteen minutes considered. The three-email cap becomes four, because the announcement obliges the delivery — touch 2 is not a discretionary touch and the cap is stated so that nobody spends it elsewhere. One email per prospect per day is unchanged and absolute |
| 3.4 | 2026-10-30 | Touches 1 and 2 merge into one email, sent on tagging. The second step could fail silently and had — three prospects were promised a report and never received one — and the personalisation window that justified the split had already been withdrawn. Touches renumber 3→2, 4→3, 5→4, 6→5, and the email cap returns to three. Sending on tagging rather than overnight keeps call situation S3 alive, because the caller can tag, wait five minutes and dial against a report that has genuinely arrived |
| 3.3 | 2026-10-29 | The configuration table's two number rows collapse into one that names the role and links to the SC Phone Number Register. The Testing number row was removed outright: that number became an inside-sales caller's live prospecting line on 2026-10-29, so the row was telling readers never to call prospects from the number every prospect is now called from |
| 3.0 | 2026-10-09 | Touch 1 announces the report; touch 2 delivers it. The old shape handed the same report over twice — once by the automated day-0 email and again inside the bucketed day-3 intro — which no document had noticed. The split removes the duplicate and creates the window in which a human reads the report and writes Audit Finding Used, so touch 2 can satisfy the consistency rule this SOP already states. Touch 2 now has two rows: the bucketed intro for slotted prospects, and a generic automated delivery for unslotted ones, which exists so an announced report always arrives rather than depending on a draft workflow. Stop rules unchanged — three emails, one per day. Copy and node design in the C-021 spec v1.7 |
| 2.9 | 2026-10-07 | New stage, Audit Sent, between New Lead and Discovery. C-021 WF-1 moves the opportunity there when the report email goes out, so "who has had the report" is a filter on stage. The stage answers only that; replied, deferred, re-activated and suppressed stay on tags. The "why tags rather than a second pipeline" rationale is narrowed to match. Closed Lost now sets status Lost and loss_reason. Follows the touch-1 send automation PRD, which authorised the sent stage as the record |
| 2.7 | 2026-09-24 | The Architecture diagram is marked as the intended model rather than what runs. Four of the tags it names do not exist, and its single cadence box covers one automated touch and four manual ones. Points at the as-built map in C-021 §2.1: this one is the destination, that one is the position |
| 2.6 | 2026-09-23 | The §Tags table gains a live-state warning: verified 2026-09-23, five prospecting tags exist in the sub-account and most of the table does not, including sc-suppressed. The model is not wrong, it is ahead of the account. Until the gap closes a do-not-call sets DND as well as the tag, because DND is platform-enforced and nothing currently reads the tag — the published report workflow has no suppression node and no tag for one to query |
| 2.5 | 2026-09-17 | Touch 1 is automated and published, so the cadence renumbers: the workflow's report delivery becomes touch 1 and the bucketed intro that carries the hand-counted finding becomes touch 2. "Nothing in this document fires without a human today" was true until 2026-09-17 and is replaced with what is now automated and what is not. prospecting-outreach-active is written by the workflow rather than by hand |
| 2.4 | 2026-09-08 | Call-first leftovers removed from email-first canon: Audit PDF URL is delivered by touch 1, not touch 2; the voicemail-count row now states the email-first cap of zero before the variant's cap of two. Terminal day aligned to 12 in the flowchart and the tag table (the cadence table, checklist and PRD already said 12). Added the field-availability warning: four of the five custom fields do not exist on the live sub-account |
| 2.3 | 2026-08-27 | Cadence switched to email-first (operator-fit decision by Rohit Anand): touch 1 hands over the report by email (Template 8), day-3 follow-up, optional warm call day 7 (no voicemail), day-10 final touch, day-12 close, reactivation unchanged. Call-first sequence retained as a variant; stop rules re-stated for both (0 voicemails email-first, email/day cap, unsubscribe stop); FL/NY noted as call-only exclusion |
| 2.2 | 2026-08-27 | Field-model simplification (PRD v2.2, 19 → 5): custom fields reduced to five contact fields; added the "state without a field" table (native State/Contact Time Zone, structural voicemail cap, prospecting-reactivated marker tag — added to the tag list — verification-note provenance, task-body deferral reason); verification checklist updated to match |
| 2.1 | 2026-08-22 | Added the prospecting-cadence-start control tag (human start signal, applied after both verification markers) and cross-linked the C-024 workflow spec |
| 2.0 | 2026-08-18 | Re-based onto the single Sales Pipeline, with outreach state carried by tags rather than a separate four-stage pipeline. Cadence changed from email-only to call-first multichannel across 6 touches. Added stop rules, the tag state model, expanded custom fields including the disposition field, the qualification screen, and the automation gate |
| 1.0 | 2026-03-02 | Initial version. Four-stage prospecting pipeline, email-only cadence at days 0, 3, and 7 |
| 2.8 | 2026-09-25 | New York opened and Ohio and Pennsylvania closed, per the outbound calling compliance standard v1.1; recording disclosure now starts the recorder |