Get started
Reply rate tells you a message got read. It doesn't tell you it made money. A campaign at 12% replies and $0 in deals is worse than one at 4% replies and $40k from eight deals, but on most dashboards the first one looks like the winner. So you scale it, and let the campaign that actually pays the bills run quiet.
This skill starts from the deals, not the messages. Connect HubSpot over the MCP, or drop in a CSV of the deals your campaigns generated from any other CRM. Your AI assistant matches every deal back to the campaign that touched the contact, by email first, falling back to name and company, and to your HubSpot lead ID field if you've set one up, ranks your campaigns by the pipeline behind them, and hands each one a verdict with the reason behind it. You climb from revenue back to the campaign, and finally know why it worked.
- Rank every campaign by the pipeline it actually generated, not opens or replies
- Bring deals in live from HubSpot over the MCP, or from a CSV export from any other CRM, with no manual matching
- Get one verdict per campaign, Continue, Stop, Adapt or Investigate, with the reason behind it
- Trace any deal back to the exact campaign and message that created it
Live data, if you're connected
Connect HubSpot and your deals sync automatically. Every campaign writes back to your CRM, so the AI can read your deals straight from HubSpot with nothing to export. It matches each deal to the campaign behind it by contact email first, or by name and company when the email differs. Add an LGM lead ID as a custom property on your deals and that becomes the strongest match of all, but it's optional: the matching still works without it.
HubSpot · native two-way sync
How to connect HubSpotNothing connected? Paste a CSV instead. Drop in your deals and campaigns as an export, any CRM, no LGM or HubSpot account required. Same ranking, same verdicts, just typed in instead of pulled live.
Rank your campaigns by pipeline
Bring your deals and campaigns together
Point the AI at your revenue. Connect HubSpot through the MCP, or paste a CSV of the deals your campaigns generated: name, stage, amount, and the contact behind each one. Any other CRM comes in the same way, as a CSV. Your La Growth Machine campaigns come in live over the MCP, or as an export.
It comes back ranked, with a verdict on each
The AI matches every deal back to the campaign that touched the contact behind it, by email, lead ID, or name and company, sums the pipeline behind each campaign, and ranks them. Every campaign gets a verdict and a one-line reason citing the numbers behind it: leads touched, deals, pipeline, conversion. The deals that match no campaign are kept aside, so you see what your outreach can't take credit for.
The winners to feed, the dead weight to cut, and the near-misses worth fixing, in one view.
Climb into the why
The ranking is the start, not the end. Reply in plain words to ask about a single campaign, and the AI pulls its full numbers: leads touched, deals, pipeline, conversion, and the reason behind its verdict. Put the winner next to the loser and the gap between the two numbers explains itself. You start from the deals and go as deep as the data lets you.
Act on the verdict, your way
The ranking is set and every campaign has a call. Now move on it, in the chat from a CSV, or live in your workspace if your CRM and La Growth Machine are connected.
- Continue the winners, cut the dead campaigns, keep the near-misses and anomalies on your radar
- Hand any "Adapt" campaign to the Campaign Challenger to fix the copy
- Works from a CSV, nothing to connect
- Reallocate on the spot, pause the losers and feed leads to the winners from one workspace
- The native HubSpot sync writes the campaign behind each deal back, so the next run is automatic
- Requires the La Growth Machine MCP
Bonus: bank the plays that actually generate pipeline
Your winners are telling you what to write next. As a follow-up ask, have the AI pull the live message content behind every campaign flagged Continue, the angle, the opener, the channel mix, the proof that landed, and save them in a winning_plays.md. Then the next campaign starts from what already converted, not a blank page. Hand the file to the Multichannel Campaign Builder and it writes the sequence from your own proven plays.
- campaign: RevOps · HubSpot stack
pipeline: €61k · CONTINUE
angle: cost of a broken attribution stack
opener: LinkedIn, one line on the pain, no pitch
channel_mix: LinkedIn → email → LinkedIn
proof: "cut reporting from 2 days to 10 minutes"
- campaign: Founders · Seed–Series A
pipeline: €39k · CONTINUE
angle: pipeline visibility before the next raise
Every new campaign inherits what already paid.
Go further: keep a weekly pulse on what pays
This skill is the deep, on-demand read: you run it when you want to know which campaigns generated pipeline and why. To catch a campaign slipping before the quarter is spent, pair it with the Weekly Performance Dashboard. Every Monday it reads your live campaigns and inbox and hands you the week's moves: the replies to answer first, and the only campaigns that actually need a fix. Connect your CRM and it ties campaigns to the deals they generated, so nothing drifts between two deep reviews.
Needs the LGM MCP connected. The deep autopsy on demand, a light pulse every week.
Prompts to try on your campaigns
Copy one, point it at your own campaigns and deals, and send it to your AI assistant. In plain English, French, any language.
One ranking, every source
- →Swap the source: HubSpot's native two-way sync, or a CSV from any other CRM, the ranking works the same
- →Bank the winners: save the plays behind your Continue campaigns and reuse them in your next brief with the Campaign Builder
- →Keep the pulse: pair it with the Weekly Performance Dashboard so nothing drifts between two deep reviews
See which campaigns actually paid
Get the skill and rank your own campaigns by the pipeline they generated. It starts from your deals.
Get the skill