All articles · May 8, 2026 · 7 min read

How to actually use your LinkedIn Connections.csv: warming your existing 1st-degree network

Most recruiters and sales reps treat LinkedIn's Connections.csv export as an artifact — something they exported once for a backup, a CRM migration, or a spreadsheet analysis, and then forgot about. That's the wrong frame. The CSV is the highest-leverage list you have, because it represents the people who already accepted your invite — and because their reply rates on warm DMs are 5-10× higher than any cold list you can buy.

Here's how to actually export it, what's in the file, and the warming workflow that turns a 500-row CSV into 15-30 conversations a month.

Why the existing connections list is undervalued

Recruiters spend hundreds of dollars a month on Hiretual, SeekOut, Apollo, and LinkedIn Recruiter for sourcing data on people they don't know. Meanwhile, the same recruiters typically have 500-2000 1st-degree connections — past candidates, ex-colleagues, conference acquaintances, people they connected with at a previous company — sitting unused in their LinkedIn network.

Two assumptions are wrong about this list:

"They're already warm because they accepted my invite." Mutual connection acceptance gives you access (the DM panel is unlocked), not warmth (recent context). A connection from 2 years ago who you haven't engaged with since is functionally as cold as a stranger from a sourcing tool. Conversion rates on cold-restart DMs to dormant connections are roughly 5-8% — better than cold InMail, but nowhere near the 40-45% that warm DMs hit.

"If they were going to reply, they'd have replied by now." This treats the connection as a one-shot event rather than a relationship that decays and can be reignited. A recruiter who comments on a dormant connection's recent post 2-3 times before DMing reactivates a 5-30% conversion rate from a list that's currently sitting at 0%.

Run the math: 1,000 dormant 1st-degree connections × 20% reactivation rate × 30% reply rate when DMed = 60 conversations from a list you already have. That's typically 2-3× what a recruiter pulls from cold sourcing in the same month, at zero data cost.

How to export your Connections.csv

LinkedIn's native export, no third-party tools needed:

1. On linkedin.com, click Me → Settings & Privacy.
2. Data privacy → Get a copy of your data.
3. Select Want something in particular? then check only Connections (not the full archive — that takes 24h).
4. Click Request archive. The smaller Connections-only export is usually ready in 5-15 minutes; you'll get an email with the download link.
5. Unzip the file. You get Connections.csv with one row per 1st-degree connection.

The columns: First Name, Last Name, URL, Email Address (only if the connection has shared their email; usually 30-50% of rows), Company, Position, Connected On.

If the headers look slightly different from the above (LinkedIn occasionally tweaks the export), the order is consistent and any importer worth using auto-detects regardless.

Recency-classifying the list

The single most useful thing to do with the CSV before warming is to split it into recency buckets. The labels we use in WarmList (and the same labels show up in the glossary) are:

Fresh — connected in the last 30 days, OR posting actively in the last 30 days. These you can DM directly without warming, because they're either still in active mutual-recall mode or because they're showing up in your feed regularly. Skip the 3-touch sequence; go straight to a contextual DM.

Dormant — connected 30-180 days ago, no recent posts. The right warming sequence here is light: 1-2 comments on whatever they've posted recently (you may have to scroll a bit), then DM. The relationship isn't dead, it's faded — 1-2 touches is enough to reactivate.

Long-cold — connected 180+ days ago, no recent activity. These need the full 3-touch warming sequence or, honestly, archival. If a connection has zero LinkedIn activity in 6 months, the chance they reply is low regardless of the warming work — 60-70% of long-cold connections are ghost accounts at this point.

The split matters because the work-per-conversation differs by 3-5× across buckets. Spending 20 minutes warming a long-cold ghost account is wasted; spending 20 minutes warming an active dormant connection produces a real conversation.

The warming workflow

Once classified, the actual workflow has three loops running in parallel:

Daily — Fresh DM loop. 5-10 minutes a day. Check your Fresh list (the auto-classifier surfaces this). Pick 2-3 to DM directly with a contextual reference to their recent activity. No warming sequence required — they're already warm.

Daily — Dormant comment loop. 10-15 minutes a day. Comment on 5-8 posts from the Dormant list, prioritising posts under 7 days old. The freshness gate matters: commenting on a 60-day-old post produces zero engagement signal because nobody (including the post author) gets notified anymore.

Weekly — Long-cold review. 30-60 minutes once a week. Look at your Long-cold list and either (a) pick 2-3 to start the full 3-touch warming sequence on, or (b) archive the rest. Most go to archive. The 2-3 you pick should be candidates whose recent post (yes, look for one) suggests they're still active enough to warm.

The total time budget is roughly 15-20 minutes a day across the three loops. From a 1,000-connection CSV that's been classified, this typically produces 15-30 active conversations per month.

What the WarmList import does

WarmList's onboarding wizard runs the classification automatically on a Connections.csv upload:

1. Upload the CSV (drag-drop into the wizard, takes 2-3 seconds).
2. The auto-classifier reads the Connected On column and runs a per-row check on latest_post_at (via the Chrome extension reading your LinkedIn session — no cloud scraping; see Browser vs cloud LinkedIn automation for why this matters).
3. Each connection is labelled Fresh, Dormant, or Long-cold and lands in the Pool with a recency badge.
4. The daily queue ranking automatically prioritises Fresh and Dormant rows over Long-cold; you don't have to manage the bucket logic manually.

The Pool view supports bulk actions: archive every Long-cold older than 12 months, promote a batch of Dormants to the active warming queue, etc. See the manual for the full workflow.

The "what about Hiretual / SeekOut / Apollo exports?" question comes up often. Same import flow — WarmList's importer auto-detects column headers regardless of source. Hiretual\'s standard export lands cleanly; same for SeekOut and Apollo. The classification step ignores source-specific metadata and works off URL (LinkedIn handle) plus the post-recency check.

Why this is the highest-ROI move right now

The compounding logic: every connection on your CSV represents a moment in the past where someone agreed to be in your network. That signal persists, but it decays. Reactivating dormant connections is 5-10× cheaper per conversation than sourcing strangers, because the trust groundwork was done in the past.

The recruiters and sales reps who figure this out in 2026 are the ones who get to outbound numbers without burning their accounts on cloud automation. The ones who don't end up running cold-only motions on accounts that are already throttled because they never engaged.

For pricing see Pricing. For the safety-limits side of the equation see LinkedIn connection request limits in 2026. For why InMail-only motions stopped working, see InMail reply rates collapsed in 2026.