We have onboarded more than 275 companies into our outbound agency. The campaigns are the visible part of that work. The part that decides whether a campaign launches in eight days or eight weeks is everything that happens between the moment a client pays and the moment the first email sends.
Early on, that gap was where deals went to die. A signed client would sit waiting on a Slack channel nobody had created, inboxes nobody had ordered, and an ICP document nobody had written, while the account manager chased four people for information the client had already given us once.
The fix was not more project management. It was removing every step that a human had no reason to be doing. Payment fires a webhook, the workspace assembles itself, and by the time a human joins the process, the only thing left for them to do is think.
Here is the full sequence, in the order it runs.
Payment Is the Trigger, Not the Signature
Stripe handles the payment and sets up the subscription, and that payment is what starts everything downstream via APIs and webhooks.
This ordering matters more than it looks. Plenty of agencies kick off provisioning when the contract is signed, which means the build starts before the money lands and the ops team carries the risk of a deal that never funds. Tying the trigger to the payment event removes that entirely. Nothing gets built for a client who has not paid, and everything gets built the second they have.
It also means there is exactly one source of truth for "is this client live". No one has to ask.
One Form, Filled Once
Sales fills a single new client form: retainer, scope, industry, specific requests. It lands in Airtable, which becomes the command center for the whole build.
The rule we hold to is that the client's information is captured once and then read by every system that needs it. The alternative, which we ran for longer than we should have, is a world where the retainer lives in the CRM, the scope lives in a proposal PDF, the industry lives in someone's memory, and the account manager reconciles all three by hand on the kickoff call.
Airtable is not doing anything clever here. It is holding one record per client and making that record addressable by everything downstream.
The Workspace Builds Itself
n8n picks up that Airtable record and fires the build.
A Slack channel is created and the client is invited. A shared Drive folder appears with the ICP matrix, copy templates, and collateral already in it. The client dashboard is created and the invite goes out through Slack.
None of these steps is hard. All of them are easy to forget, and forgetting one is what turns a smooth first week into an apologetic email. Automating the boring provisioning is worth more than automating the interesting work, because the boring provisioning is the part that gets skipped under pressure.
Email Infrastructure Ordered on Day One
Domains and inboxes need weeks of warm-up before they can send at volume without landing in spam. That warm-up is the longest lead time in the entire onboarding, and it is not compressible.
So we order them from Hypertide automatically, the same day the payment comes through. Not after the kickoff call, not after the ICP is signed off, not after the copy is approved. The clock on deliverability starts before anyone has decided who we are emailing, because the warm-up runs in parallel with every other step and there is no version of the plan where we need fewer inboxes than we ordered.
This one change moved our average time to first send more than any copywriting improvement we have made.
Deliverability is the thing that quietly decides whether any of the rest of this matters, and it is worth checking your existing setup before you scale sending on top of it. We built a mini tool that scores a domain and flags what will send you to spam.
You can check your sending setup here, for free:
Spam Checker Tool
Fields marked with * are required
The Onboarding Form, and What Happens When It Comes Back Empty
The client gets the onboarding form the day they sign.
The realistic part of this step is the fallback. Forms come back empty, or half-filled, or filled by someone who was not on the sales call and does not know the answers. When that happens, the account manager fills it out with them on a call rather than sending a reminder and waiting.
Treating the empty form as an expected outcome rather than a client failure is what keeps the timeline intact. The information is the point. How it gets captured is negotiable.
The Kickoff Call Becomes Structured Data
The kickoff call is recorded inside Claap, which is built around GTM conversations and catches the nuances that generic transcription misses. On the call, we brainstorm campaign angles, ideas, and priorities with the client in the room.
Then the recap does real work. From it, AI writes the ICP matrix and the next steps straight back into the workspace, along with the campaign angles worth prioritising.
The reason this is worth automating is not the time saved on note-taking. It is that the ICP matrix written from the recording reflects what the client said, in their words, instead of what the account manager remembered three days later. Every downstream decision, the list, the copy, the tiering, inherits that document. Getting it wrong at the source is expensive and hard to detect.
Campaign angles are the other output of that call, and the quality of the first three angles sets the tone for the entire engagement. We built a mini tool that generates them from an ICP.
You can generate campaign angles for your own ICP here, for free:
Campaign Ideation Tool
Enter your email to generate your campaign ideas.
TAM Mapping Before Anyone Writes Copy
We map the total addressable market against the ICP matrix in GetLeads before anyone writes a line of copy.
The sequencing is the whole argument. Writing copy first produces messaging built for a market you have not sized, and the failure mode is discovering after launch that the ICP as written contains 400 companies rather than the 40,000 the retainer assumed. At that point the copy is fine and the campaign is finished.
Mapping first also changes the conversation with the client. Instead of asking them to approve an ICP in the abstract, we can show them the size of it and let them decide whether to widen the criteria, tier the list, or accept a smaller and sharper target.
Once the TAM is mapped, the next question is which adjacent companies look like the accounts already converting. We built a mini tool that takes your best accounts and returns their nearest matches.
You can pull a lookalike list from your strongest customers here, for free:
Lookalike Finder Tool
Fields marked with * are required
Quick examples:
Build, Launch, and the Weekly Sync
Once the angle is approved, we review the list and the copy with the client together, then get the campaign live in Instantly for email, and lemlist as well when the motion goes multichannel.
The reviews do not stop at launch. On weekly syncs, we go back through what is working and what is not, using the Claap recordings as the record of what was agreed. That gives us room for planning and improvement, and it surfaces client sentiment and early signs of churn while there is still time to act on them.
A churn signal in week six is a problem you can solve. The same signal read in week sixteen is a renewal you already lost.
What This Buys You
None of the eight steps above is individually impressive. Stripe fires a webhook, n8n creates a folder, a form gets filled, a call gets recorded. Any of them could be done by hand.
Doing them by hand is where the time goes. Across 275+ clients, the compounding win is not that automation does the work faster. It is that the work happens at all, in the same order, every time, without depending on how busy the account manager was that week.
The humans in this process do two things: they run the kickoff call, and they decide which campaign angles are worth building. Everything else is provisioning, and provisioning is what software is for.
