"Find → qualify → write → send/reply/book" sounds like a slogan until you watch an agency week. Then it becomes a checklist of who does what, what the system blocks, and which metrics actually mean the client is getting conversations — not just activity.
Here is the day-to-day meaning of each step inside client-isolated workspaces.
1. Find — build a TAM slice the ICP can survive
Day-to-day: pull people and accounts that match the client's ICP, into that client's workspace only.
What good looks like:
- Firmographic and persona filters match the written ICP
- Geography and disqualifiers applied before enrichment spend balloons
- Source provenance retained on fields (so you can answer "where did this come from?")
What bad looks like:
- A shared agency list pasted into every client
- "We'll clean it later" enrichment on obvious non-fits
- Titles that can never take the meeting clogging the top of the funnel
Find is not the same as spray. Find is scoped sourcing against an ICP you are willing to defend on a client call.
2. Qualify — decide who is allowed into a campaign
Day-to-day: score signals and enforce campaign entry rules so weak fits never burn sends.
Qualification inputs typically include:
- Role fit and seniority
- Account fit vs. ICP
- Timing / intent-style signals when you have them
- Suppression and prior engagement history for that workspace
- Hard disqualifiers (competitors, banned verticals, do-not-contact)
Only then does a prospect enter a sequence. Qualify is the cheapest compliance and deliverability control you have — cheaper than apologizing for a bad list after the fact.
3. Write — evidence first, client voice second, clever last
Day-to-day: generate (or review) messages from the prospect profile + signals, in the client's voice.
Operator rules that keep quality high:
- Cite something real (role, company context, signal) — do not invent urgency
- Keep the ask conversational (book a call / explore fit), not a feature dump
- Prefer supervised drafts until edit rates fall
- Store tone and offer context in the workspace so Client A never leaks into Client B
Writing without qualify is fiction. Writing without client voice is agency-branded spam wearing a client domain.
4. Send — channels, windows, limits, gate
Day-to-day: messages leave on connected email domains and LinkedIn for that client, inside send windows and limits, only if the compliance gate passes.
The gate that matters: no send without a compliance verdict and an unsubscribe URL. That is a database constraint in GTM — not a footer toggle. For US commercial email baselines on opt-out behavior, see the FTC's CAN-SPAM compliance guide.
Also day-to-day:
- Warm and pace domains; do not "catch up" volume on Monday by nuking reputation
- Keep LinkedIn inside connection health and platform reality
- Log what sent, to whom, from which identity
5. Reply — shared inbox, stop the sequence, own the thread
Day-to-day: replies land in the workspace shared inbox; the sequence stops; a human (or assisted workflow) continues the conversation.
Non-negotiables:
- A reply is a conversation, not a cue to fire the next templated bump
- Ownership is clear (who works positive replies before end of day)
- Negative / unsubscribe / "not interested" paths update suppression correctly
- Client-facing reporting can show reply outcomes without mixing workspaces
Agencies that automate send but ignore reply staffing create ghost pipeline: activity up, booked calls flat, client trust down.
6. Book — turn intent into a calendar event and CRM truth
Day-to-day: positive threads move to a booked conversation with calendar and CRM follow-through.
Define "booked" with the client up front:
- Is a 15-minute intro enough?
- Must it be with a specific seat (AE vs. founder)?
- Does a no-show count until rescheduled?
Then instrument the path: link scheduling, capture the event, attribute it to the workspace campaign, and feed learnings back into qualify rules.
A realistic agency week mapped to the motion
Monday — refresh ICP notes; review suppressions; set the week's find targets per client workspace.
Tuesday–Wednesday — find + qualify batches; supervised review on new segments; autopilot only where earned.
Thursday — reply-heavy day; protect SLA on positive threads; tighten copy on weak angles.
Friday — booking follow-through; credit and volume review; decide what graduates from supervised next week.
This rhythm beats heroic end-of-month blasts.
Credits touch every step — measure the whole loop
Because find burns data credits and write/send burn action credits, reporting only "emails sent" hides the real unit economics. Tie credit spend to the full chain: qualified entries, conversations started, and booked calls. That is how an agency decides whether next week needs more find depth or more reply capacity.
What GTM is optimizing toward (honestly)
GTM is AI outbound for agencies across that full loop in isolated client workspaces. Packaging is separate from the motion: access in batches, onboarding as a one-time fee ($2,000 or $5,000 agreed with sales, no credits included), then prepaid monthly credits (minimum $250) with data credits at $0.01 and action credits at $0.0025.
The internal optimization goal is on the order of $100 usage per qualified booked call. That is a goal for how we build and tune the system — not a guarantee, not a case study, not a promise for your client mix.
CTA
If you want this motion as the default operating system for agency outbound — not a pile of disconnected tools — join the GTM waitlist.