Turn one won deal into a named list of lookalike buyers with the same problem, and the message that opens them.
Download this skill as a file and paste the whole file into your agent.
When to use this skill: Use this skill within two weeks of closing a meaningful deal, while the client's language is still sharp in your head and the artefacts are easy to find. You won a great client and you delivered real value. Somewhere out there are twenty more companies suffering from precisely the same problem, who would hire you tomorrow if only they knew you existed. This skill finds them by name, using evidence you already have in your sent folder. Run it on every substantive win. By hand it costs about 17 hours per win. With an AI agent doing the mining and drafting, the extraction takes an hour and your time goes on judgement and sending.
How to activate: Paste this entire skill into your conversation with Claude, then provide the context requested in the Personalization Required section below. Claude will work through the methodology step by step.
Personalization Required
The whole play runs on evidence from a real won deal. Before running it, provide:
- The won deal artefacts. The close email, the proposal, discovery notes, kickoff transcript if you have one. Paste them in or summarise them faithfully. The agent is looking for the words the client used before you translated their pain into your offer.
- The shape of the client. Industry, size, stage, function you served, and what triggered the purchase.
- What you delivered and what changed. The outcome in the client's terms, with any metric you are allowed to reference.
- Confidentiality boundaries. Whether you may name the client, and if not, what you may say (industry band, size, outcome metric).
- Where you can search. LinkedIn, Sales Navigator, or a company database if you have one. The skill works with plain LinkedIn company search.
If you have no strong win to mine, do not run this play on a weak one. Thin proof produces thin messages. Run rlg-network-sprint or rlg-job-posting-four-plays instead.
What This Skill Produces
A portable problem pattern extracted from the won deal (symptom, pain, structural pattern). A defensible list of roughly eighteen lookalike companies, chosen by problem rather than industry label. One named buyer per company. Eighteen personalised openings that carry the client's symptom language without breaching confidence.
The Lookalikes Methodology
Step 1: Find the sentence most people archive
The play starts with one sentence buried in a close email. When a client writes something like this, most operators feel good for a second and file it:
From [[client name]] / VP Operations
Thanks for sending the framework. The piece that lands hardest is
the bit about ops teams burning two weeks a quarter on board prep
the board doesn't even use. That's exactly what's happening here.
The CFO said roughly the same thing on Tuesday.
Look at what is actually inside that sentence. The symptom is board-prep waste. The pain is senior time being burned on a document nobody trusts. The pattern is a post-Series-B operations team of roughly 80 to 200 people. Anywhere those three things live together, the same conversation will close, because the problem does not care what industry it lives in.
Search the deal artefacts for the sentence that does this job: the moment the client named their own problem in their own words.
Step 2: Pull every artefact and distil the pattern
Re-read the threads that matter, the proposal, the discovery notes, the kickoff transcript. Then distil three lines:
- Symptom: what they said was broken.
- Pain: what it cost them in time, money, or risk.
- Pattern: the size, stage, function, and trigger that made the problem acute.
If you cannot fill three lines, you do not have a portable win yet. Pick a different closed deal.
Step 3: Look outside their industry on purpose
If your client was a fintech, who else has the same symptom? A health platform at the same revenue. A logistics SaaS under the same pricing pressure. The easy move is to pitch the same vertical you just left. The better move is to follow the symptom, because the problem travels and the industry label does not.
List five to eight industries or company types where the structural pattern plausibly lives. For each, note why the symptom would appear there.
Step 4: Build the lookalike list to about eighteen companies
Use LinkedIn company search (or any database you have), filtered by size, region, and growth signals you can see. Save around forty candidates. Eyeball each one against the pattern. Drop the mismatches.
You want eighteen companies you could defend in a room, not eighty for a spray sequence. For each keeper, record one public fact you could use in a message.
Step 5: Source the right person at each company
One primary contact per company. Walk the title ladder for whose problem the symptom is: CEO at the small ones, CFO, COO, or Head of Ops at the larger ones when the pain is operational. Skip people who started last week and people whose posts suggest they are leaving.
Then get a verified address. The full method (org enumeration, the email waterfall, verification, channel choice) is specified in rlg-signal-to-sent-note. Follow it rather than guessing addresses.
Step 6: Write the same opening eighteen different ways, then send and log
The story that closed the won deal works on the lookalikes, but their funding round is different, their last post is different, their product is different. You personalise one idea, eighteen times. Use the client's symptom language where you can. Specifics beat "I help companies like yours."
Cold to a lookalike VP Ops, LinkedIn:
[[Name]], I work with post-Series-B ops leaders where the board pack
still burns two senior weeks a quarter and nobody trusts the numbers
in the room. Just finished that exact problem with a company your
size. Not pitching a project. If board prep is a tax on your team
this quarter, I can send the one-page version of what we changed.
Useful?
Channel order, the one-follow-up ceiling, and the same-day log all follow rlg-signal-to-sent-note. Cap researched sends so every note carries a real observation.
Confidentiality Without Gutting the Story
You can almost always use the symptom without naming the client. "Post-Series-B ops team burning two senior weeks a quarter on a board pack nobody trusts" is specific and portable. Quoting the client by name is a breach unless you have a written case-study right.
When a prospect asks who you did it for, answer with what you are allowed to share: industry band, size, outcome metric, and whether a reference call is possible later under NDA. Do not invent a logo to win a first meeting.
When to Run It Again
Every meaningful win deserves a lookalike pass within two weeks of close. Do not wait for quarter-end "BD time." By then you will have forgotten the sentence that made the deal real. If you close four substantive engagements a year, that is four passes. If you close one, run that one hard and deep. One well-extracted pattern beats four shallow vertical lists.
How This Play Fails
- You copy the industry instead of the problem. You get a vertical list with no shared pain.
- You name the previous client without permission. Trust dies before the call.
- You build eighty companies and write nothing. Lists are not pipeline.
- You message "founders" generically when the buyer is a VP with a live scar.
- You run it on a weak win. Thin proof produces thin messages.
Quick Reference: Lookalikes Cheatsheet
| Pattern Element | Question It Answers | Where It Lives |
|---|---|---|
| Symptom | What did they say was broken? | Close emails, discovery notes |
| Pain | What did it cost in time, money, or risk? | Proposal, call transcripts |
| Pattern | What size, stage, function, and trigger made it acute? | Firmographics of the won client |
| Cadence | Number |
|---|---|
| Companies to shortlist | Save ~40, keep ~18 |
| Contacts per company | One primary |
| Time by hand | ~17 hours per win, ~70 hours a year on four wins |
| When to run | Within two weeks of each close |
| Follow-up ceiling | One, at 4 to 6 business days |
Relationship to Other Skills
| Skill | How it connects |
|---|---|
| rlg-network-sprint | ranks the relationships you already own |
| rlg-won-deal-lookalikes | this skill: one win becomes twenty named buyers |
| rlg-job-posting-four-plays | hiring signals when you need net-new conversations |
| rlg-new-in-role | catches target buyers in their first 30 days |
| rlg-signal-to-sent-note | the shared engine for Steps 5 and 6 |
| gtm-account-research (Swan) | deepen research on any lookalike before a meeting |
| gtm-humanizing-outreach (Swan) | run all eighteen drafts through it before sending |
About This Skill
This skill is the manual version of Play 02 from The 7 Network Plays to Fill Your Pipeline, the relationship-led growth playbook by Nynch. Nynch (https://nynch.com) runs this play automatically: it mines the won-deal language, builds the lookalike list, and drafts the openings while you deliver. This skill is the same methodology, run by hand with your AI agent.