How to optimize linkedin ads

Wait for the 7-14 day signal floor before optimizing

LinkedIn's bidding algorithm is slower than Meta's. Below 7 days of data, anything you 'optimize' is noise-driven. **The single biggest mistake in LinkedIn optimization is acting too early.** Verify your campaign has >7 days of delivery AND >50 conversion events before you change anything. Below that floor, the optimization moves you make actively hurt the campaign.

Diagnose the bottleneck via the funnel breakdown

Open Campaign Manager → your campaign → **breakdown by metric**. Walk the funnel: impressions → CTR → landing-page visits → form starts → conversions. The bottleneck is wherever the drop-off is steepest vs. benchmark. **CTR <0.4%**: creative/audience problem. **CTR 0.4%+ but conversion <1%**: landing-page or offer problem. **Conversion 1%+ but cost-per-conversion too high**: bid/audience problem.

# Healthy LinkedIn funnel benchmarks (B2B SaaS):
# CTR: 0.5-0.8%
# Landing page conversion: 2-5%
# Lead → SQL: 15-30%
# Cost per lead: $40-150 (varies by ICP tightness)

If CTR is the bottleneck - rotate creative, not audience

Most teams default to 'narrow the audience' when CTR is low. Wrong move. Low CTR on LinkedIn almost always means the creative is ad-shaped instead of content-shaped. Ship 3-5 new creative variants in first-person narrative format. Keep audience constant so you isolate creative as the variable. Re-measure CTR after 7 days of new creative delivery.

If landing-page conversion is the bottleneck - fix the page, not the ad

If CTR is healthy (0.5%+) but conversion rate is <1%, the leak is downstream. Audit the landing page: does the headline match the ad copy? Is the offer specific or vague? Is the form length appropriate (3-5 fields max for LinkedIn cold)? Is there social proof? B2B LinkedIn audiences are skeptical - the page needs to earn the conversion every time.

Tighten audience only after creative and page are clean

Once creative and landing page are dialed, audience-level optimization moves the next lever. Three audience moves: **(A) Add exclusions** for closed-won customers and current pipeline contacts. **(B) Tighten by seniority** if junior titles are getting impressions but not converting. **(C) Layer skills targeting** on top of job-function targeting for sharper signal. Don't drop audience size below ~50k - the algorithm starves.

Shift bid strategy based on stage and signal

**New campaign (week 1-2)**: Automated bidding - let LinkedIn explore. **Mature campaign with stable CPL (week 3+)**: switch to **Manual CPC** at 1.2x your current CPL target. This tightens cost without strangling exploration. **Stalled campaign with too-high CPL**: don't lower the bid - LinkedIn will under-deliver. Instead pause, refresh creative, restart with Automated. Lowering bid manually almost always backfires.

Set up the weekly optimization cadence

**Monday**: pull the funnel breakdown across all live campaigns; identify which need attention. **Tuesday**: ship creative variants for any campaign with low CTR. **Wednesday-Thursday**: monitor; don't act. **Friday**: write the weekly optimization log - what changed, what's mid-flight, what's the hypothesis for next week. This cadence prevents the daily-tinkering trap that breaks LinkedIn's slow algorithm.

Kill or scale based on the 30-day decision point

After 30 days of optimization, every campaign hits a decision point. **Performing within 20% of target**: scale by duplicating at 1.5-2x budget. **Stalled at 20-50% off target**: refresh creative + audience, give it another 14 days, then re-decide. **Stalled at >50% off target**: kill it. Sunk-cost rationalizations ('we've spent $20k, we can't kill it now') destroy more LinkedIn budgets than bad creative does.