--- license: MIT name: rlg-network-sprint skill: 01-05 description: Turn the network you already own into a ranked list of warm conversations, with a real angle for each one. --- # RLG Network Sprint **When to use this skill:** Use this skill when your pipeline has gone quiet and you have years of relationships sitting unranked in LinkedIn and your inbox. You are not short of conversations. You are drowning in them. Perhaps sixty people in a network of twelve hundred are ready for a real conversation this quarter, and you have no way of knowing which sixty they are. This skill turns the address book you already own into a live ranking of who matters right now, why they matter, and what you would open with. Run it quarterly. Done by hand, the first full pass on 1,200 connections costs 80 to 96 hours. With an AI agent running the scoring and drafting, it fits in an afternoon plus your judgement. **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 This skill ranks your real network, so the output is only as good as the inputs you bring. Before running it, provide: - Your LinkedIn connections export. Go to Settings > Data privacy > Get a copy of your data > Connections, wait for the file, and share Connections.csv. Expect mess: half the rows will be missing titles and a third will list a company that was true two years ago. That mess is the point. Your memory has been carrying a cleaner fiction than the file. - Who you sell to. Title patterns, company size band, industry or problem domain, and the trigger events that make someone ready to buy. If you have a documented Ideal Customer Profile, share it. If not, describe your three best clients and why they were great. - What you sell. Your service, the problems it solves, and what an engagement looks like. - Anything you remember about the relationships. Past clients, people who have referred you, people you would call friends. Even rough notes raise the quality of the relationship scoring. Without this, the output is a generic contact list. With it, the agent builds a ranking calibrated to your actual history. --- ## What This Skill Produces A scored and ranked network sheet: fit score, relationship strength, last real contact, and a predicted problem for every contact worth keeping. A Monday top-ten list of the people to contact this week. Drafted outreach for the top contacts that names a real problem, brings a fresh angle, and asks for nothing. --- ## The Network Sprint Methodology --- ### Step 1: Load the export and face the real network Parse the Connections.csv. Count the rows, flag the rows with missing titles or stale companies, and deduplicate obvious repeats. Do not clean it to perfection. The goal is a working list, not a beautiful one. The relationships you already have are worth more than the ones you have not built. A continuous audit pays you in conversations that never had to be cold, which is the only kind that protects your status instead of spending it. Most operators drop this play not because they are careless but because the default tooling was a sales CRM and a vague instruction to stay in touch, which turns into admin with no signal and no sentence worth sending. ### Step 2: Write the ideal client as a scoring rubric, not a vibe Four fields on one page: - Title patterns you sell to - Company size band - Industry or problem domain - A trigger event (funding, new leader, expansion, cost pressure) Score each field 0 to 2. A perfect fit is 8. This is the work you have carried in your head for years and never written down. Until it is written, you will re-litigate it on every row and the afternoon will die. Agree the rubric with the user before scoring anything. ### Step 3: Score every contact for fit Apply the rubric to each row: green (6 to 8), amber (3 to 5), red (0 to 2). Working by hand, do this in batches of 40, because by contact 80 the rubric starts to drift and by 200 the energy is gone. Working with an agent, score everything, then have the user spot-check one batch of 40 before trusting the rest. Do not write any outreach yet. Red rows can be archived without guilt. The play is a ranking, not a reunion tour. ### Step 4: Score the greens and ambers for relationship strength Four questions only: 1. When did you last speak? 2. Have they hired you or referred you? 3. Would they take a call from you without wondering who you are? 4. Was the last contact substantive or performative? Band each contact as warm, cool, or polite-stranger. This number lives in the user's head and nowhere else, so ask, do not guess. A contact who only ever liked your posts is a polite stranger, not a warm relationship. ### Step 5: Predict the problem each high-fit person is probably facing Use public crumbs: a hiring freeze post, a board that wants a number, a peer who just got cut, a reorg nobody has announced yet, a product launch that will break their process. Check their recent LinkedIn activity and company news where possible. Write the predicted problem in one sentence per contact. If you cannot name a plausible problem in one sentence, move them down the list. You are not ready to message them without becoming "just checking in." ### Step 6: Write the outreach. One person, one angle, no ask for work A real message names the problem, brings an angle they had not considered, and asks for nothing more than a short conversation or permission to send something useful. Different every time. Prefer the channel you already use with that person. LinkedIn is fine when you have no email. Cap sends at five to eight per working day so the quality holds. One follow-up only, four to six business days later, shorter than the first note. Then stop unless they reply. Two reference patterns, drawn from real sends: Past client, gone quiet 11 months, email: Subject: The board-prep waste we fixed at [[their company]] [[Name]], I still think about the line you used when we closed out: ops burning two weeks a quarter on a pack the board never opens. I have seen that pattern three more times since. Curious whether it is still true on your side, or whether the problem moved. No pitch. Twenty minutes if useful. If not, ignore this and I will not chase. Warm peer, possible referrer, LinkedIn note: [[Name]], saw you posted about the reorg. When we last spoke you were still protecting the advisory bench from becoming a delivery factory. If that fight is live again, I have a one-page pattern from two similar transitions that might save you a week. Want it? ### Step 7: Build the sheet that keeps the play alive Keep one sheet with these fields per contact: - Fit score (0 to 8) from the rubric - Relationship (warm / cool / polite) - Last real contact (date and channel, not a like) - Predicted problem in one sentence - Status and next date Sort weekly by fit descending, then by relationship strength, then by how stale the last contact is. Your Monday list is the top ten you have not contacted in ninety days. That is the whole operating system. Anything fancier is procrastination dressed as infrastructure. ### Step 8: Run the reply properly A yes to a short conversation is not a pitch meeting. Open with the problem you named. Ask what is true for them now. Only then, if it still fits, describe how you have solved it and what a first engagement would look like. If they say "not now," ask when it would be sane to check back, put that date in the sheet, and keep the relationship warm with occasional substance rather than calendar spam. --- ## How This Play Fails - You score everyone as green because saying no feels rude. The list becomes unusable. - You send "hope you're well" notes. Reply rates die and you blame the play. - You try to finish 1,200 rows in one weekend. Quality collapses after row 150. - You never re-run after quarter one. The ranking rots while you deliver. - You message people who only ever liked your posts. Polite strangers are not warm. --- ## When Not to Start Here If your honest first-degree graph is under a few hundred people and almost none of them buy what you sell, do not punish yourself with this play. Start with the hiring-signal plays instead (rlg-job-posting-four-plays), where the budget is already approved and the need is public. Come back to the Network Sprint when there is something worth ranking. --- ## Quick Reference: Network Sprint Cheatsheet | Rubric Field | Score 0 | Score 1 | Score 2 | |---|---|---|---| | Title pattern | Not a buyer or influencer | Adjacent to your buyer | Exactly who you sell to | | Company size | Outside your band | Edge of your band | Inside your band | | Industry / problem domain | No overlap | Partial overlap | Your home ground | | Trigger event | Nothing visible | Possible trigger | Live, visible trigger | | Relationship Band | Meaning | Action | |---|---|---| | Warm | They would take your call without hesitation | Message first, highest priority | | Cool | Real history, but recency has died | Message with a strong, specific angle | | Polite-stranger | Connected, never substantive | Only with a genuinely useful gift | | Cadence | Number | |---|---| | First full pass, 1,200 connections, by hand | 80 to 96 hours | | Re-run | Quarterly, maintenance only if the sheet stays alive | | Monday list | Top ten not contacted in 90 days | | Daily send cap | 5 to 8 researched messages | | Follow-up ceiling | One, at 4 to 6 business days | --- ## Relationship to Other Skills rlg-network-sprint → start here when you have any history at all rlg-won-deal-lookalikes → clones your best win into buyers you have never met rlg-job-posting-four-plays → start here instead when the network is thin rlg-new-in-role → catches target buyers in their first 30 days rlg-signal-to-sent-note → the shared engine: find the person, verify the address, send and log gtm-icp-definition (Swan) → a full ICP build that feeds Step 2's rubric gtm-humanizing-outreach (Swan) → run every draft through it before sending --- ## About This Skill This skill is the manual version of Play 01 from The 7 Network Plays to Fill Your Pipeline, the relationship-led growth playbook by Nynch. Nynch (https://nynch.com) runs this play continuously in the background: it scores the network, watches for decay and triggers, and puts the prepared conversations on your desk every Monday. This skill is the same methodology, run by hand with your AI agent.