Most LinkedIn accounts we audit have the same architecture: one campaign, cold audience, lead gen form, straight to “book a demo.”
That’s not a funnel. That’s a cold call with a $32 CPM.
And the results look exactly like you’d expect — CPLs at the painful top of the range, leads that ghost the sales team, and a slow-building conviction that “LinkedIn doesn’t work for us.” The platform gets blamed for a structural problem. Roughly 5% of your market is in-market to buy at any given moment; a demo-ask aimed at cold traffic is a pitch to the other 95% to ignore you.
Demand generation on LinkedIn is the fix — we’ve compared lead generation vs. demand generation head-to-head before — and it’s less mysterious than the term suggests. It’s a three-stage structure where each stage has one job, and each stage feeds the next automatically. Get the structure right and average creative will still produce pipeline. Get it wrong and brilliant creative won’t save you.
The Three-Stage Structure
| Stage | Audience | Job | Ask |
|---|---|---|---|
| 1. Awareness | Cold ICP | Build recognition + fill retargeting pools | None — just attention |
| 2. Engagement | Warm (engaged with Stage 1) | Build trust and preference | Light: read, watch, follow |
| 3. Conversion | Hot (multi-touch engaged + site visitors) | Harvest demand | Demo, call, trial |
The stages aren’t three separate ideas — they’re one system. Stage 1 exists mostly to fill the audiences Stage 2 runs against. Stage 2 exists to qualify who’s warm enough for Stage 3. Delete any stage and the others stop working.
Stage 1: Awareness — Buy the Audience, Not the Lead
Cold campaigns have one job, and it isn’t leads. It’s reach into your ICP at healthy frequency (6–10x per month) with content that earns attention: a strong point of view, a contrarian take on a problem they have, a data point they haven’t seen. Video is the workhorse here, for a specific mechanical reason — LinkedIn lets you build retargeting audiences from video viewers, so every 25%+ view is a person quietly moving into your Stage 2 pool without ever clicking anything.
What doesn’t belong in Stage 1: lead gen forms, demo asks, “book a call” CTAs. Anyone who converts on those from a cold impression was in the in-market 5% and would have converted anyway — and you’ll pay $500+ cold CPL prices for the privilege. Cold CPL runs $250–$700+ across B2B verticals; the entire economic logic of the funnel is that you don’t pay those prices for most of your leads.
Stage 2: Engagement — Where Trust Compounds
Stage 2 targets the people Stage 1 warmed up: video viewers, ad engagers, company page visitors, website traffic. These people know who you are. Now you give them depth — case studies, teardown-style content, webinar invites, founder commentary. The ask stays light.
This is the stage impatient advertisers skip, and skipping it is expensive. The trust that makes someone accept a demo invitation gets built here, across weeks of low-pressure touches. It’s also where frequency does its quiet work: a buyer who’s seen eight useful pieces from you over six weeks doesn’t need to be convinced when they finally hit a conversion ad. They’ve already decided; the ad is just the doorway.
Stage 3: Conversion — Harvest, Don’t Hunt
Stage 3 runs your demo, trial, and call CTAs — but only against audiences that have earned them: multi-touch engagers, repeat site visitors, high-intent page viewers (pricing, case studies), and CRM lists. This is where warm CPL economics live: $75–$250, a fraction of cold prices, with dramatically better lead-to-opportunity rates because these buyers arrive pre-sold.
The counterintuitive rule: keep Stage 3 small. It’s the cheapest stage per lead, so the instinct is to pour budget into it — but conversion campaigns can only harvest the demand the first two stages created. Overfund Stage 3 and you exhaust the warm pool in a week, frequency spikes, and CPL climbs back toward cold prices. The pool refills at the rate Stages 1 and 2 fill it. That’s the whole system.
The Budget Split — and How It Shifts
There’s no single correct allocation, but there is a correct shape: heavy on awareness early, gradually rebalancing as retargeting pools mature.
| Program Age | Stage 1 (Cold) | Stage 2 (Warm) | Stage 3 (Conversion) |
|---|---|---|---|
| Months 1–3 | 70% | 20% | 10% |
| Months 4–6 | 55% | 25% | 20% |
| Months 7+ | 45% | 30% | 25% |
Months one through three are deliberately awareness-heavy because the warm pools don’t exist yet — there’s nothing for Stages 2 and 3 to run against. This is also why the first quarter of a demand gen program produces modest lead volume at high CPLs, and why that’s not a malfunction. We’ve shown how demand generation produces better LinkedIn leads over time — the funnel version of that story is simply: you’re watching the pools fill.
One budget note before any of this: the three-stage structure needs room to run. Below roughly $3K/month in ad spend, you can’t fund three stages at meaningful frequency — the minimum budget for LinkedIn ads still applies, and at the true minimum you’re running a compressed two-stage version (cold video + site retargeting) rather than the full architecture. For the full cold-to-retargeting ratio math at different spend levels, see how much you should spend on LinkedIn ads.
How the Pools Actually Connect
The retargeting audiences are the plumbing of the whole funnel, so build them on day one — even the ones you won’t advertise to for months:
- Video viewers (25%, 50%, 75% completion) — your largest, cheapest pool
- Ad engagers — clicked, reacted, commented on any campaign
- Website visitors — via the Insight Tag, segmented by page depth (all visitors vs. pricing/case-study visitors)
- Company page engagers — visited or followed
- CRM lists — open opportunities, closed-lost, target accounts
Audiences need 300 members before LinkedIn will serve against them, and they take weeks to fill at normal spend levels. Every week a pool doesn’t exist is a week of engagement you paid for and didn’t capture. Layer the pools by intent, too — a 75% video viewer who also hit your pricing page belongs in Stage 3 today, not in Stage 2’s nurture rotation. And if you’re running target-account lists, our demand gen playbook for ABM with LinkedIn ads covers how account lists slot into this same pool structure.
How to Tell the Funnel Is Working
Because most of the funnel’s work happens before anyone converts, you need stage-level indicators — not one blended CPL:
Stage 1: reach against ICP, frequency in the 6–10x band, pool growth rate, cost per video view. Stage 2: pool sizes crossing serving thresholds, engagement rates from warm audiences (should be multiples of cold), branded search and direct traffic starting to lift. Stage 3: warm CPL in the $75–$250 range, lead-to-opportunity rate versus your cold baseline, pipeline created.
A blended CPL across all three stages is nearly useless — Stage 1 will always look expensive per lead and Stage 3 will always look cheap, by design. Judge each stage on its own job. And when you report upward, translate all of it into pipeline terms; measuring LinkedIn ads performance beyond conversions is its own discipline, but the short version is that nobody funding this funnel cares about your cost per video view.
Expect the crossover — where warm-sourced pipeline overtakes cold-sourced — somewhere in months four through six. From there the economics improve every month the structure stays intact.
Structure First, Then Everything Else
Creative testing, bid strategies, format selection, copy — all of it matters, and all of it is downstream of structure. The three-stage funnel is what turns LinkedIn from an expensive lead vending machine into a system that gets cheaper and better as it runs: pools compound, trust accumulates, and an ever-growing share of your leads arrive at warm prices already knowing who you are.
If you want someone to look at your current account architecture and map the gap between what’s running and the structure above, book a strategy call — a funnel audit is usually the fastest way to find money you’re leaving on the table.
