I Spent (Other People's) $14,000 on LinkedIn Ads. Here's What Actually Worked.

Fair warning: this is not a post full of best practices I found on a marketing blog. This is data from accounts I actually manage — campaigns I built, tested, obsessed over, and sometimes watched flop in real time.

Over the past year I’ve been running LinkedIn Campaign Manager across three very different clients: a B2B SaaS platform, a personal brand investor newsletter, and an organizational consulting firm. Combined, that’s roughly $14,000 in spend, three completely different audiences, and enough A/B test data to have some opinions worth sharing.

Here’s what the numbers actually said — broken down by format.

Thought Leadership ads are doing something the other formats simply can’t.

I’ll lead with the most jaw-dropping number in my data: 20.91% click-through rate. On LinkedIn. In B2B. The platform benchmark sits around 0.5-1%, so yes — we were running at roughly 40x that. Average CTR across the campaign was 15.36%, at $0.06 per engagement, which LinkedIn’s own system flagged as 86% below benchmark cost. This was the Thought Leadership format: first-person, founder-voice, story-first sponsored content.

The ad that did it opened with: “$15 million. Gone. Just like that. All because of 50,000 missed calls.” No product name. No feature callout. Just a specific number that made a dental practice owner do instant math on their own business and think: wait, is that us?

That’s the job of a great ad. Not to explain your product. To make your reader feel briefly seen and slightly panicked.

Single image conversion ads: strong, but only if your headline has teeth.

Across a conversion-optimized campaign (bottom of funnel, $200/day budget), we hit $1.97 per conversion at scale with over 3,200 total conversions. The best creative hit 2.28% CTR. But there was a sharp divide between the ads that worked and the ones that didn’t, and it came down entirely to the first line.

“The Silent Tax on Your EBITDA” — clicked. “Redefine Excellence in Dental Care” — clicked. “Elevate Your Practice” — clicked. Meanwhile, anything that opened with what the product does? Crickets. Your buyer is not waiting to be educated. They’re waiting to feel understood.

Newsletter signup ads are wildly cheap and almost nobody is running them.

This is the sleeper finding in my data. For a client’s personal brand newsletter, I’m running a conversion campaign — 10 ads, all single image, all driving to a newsletter signup page. The cheapest conversion so far? $0.41 per subscriber. The current campaign average? $0.96. For context: most email list-building tools charge you $2-5+ per subscriber acquired through paid channels. We’re well under $1.

What’s working? Not “Subscribe to my newsletter.” Headlines like “Subscribe now and stay ahead of the room” and “In the investor world, timing is everything” — both under 45 cents per conversion. They’re selling the feeling of being ahead. You’re not selling content. You’re selling identity. That’s a very different thing and the numbers show it.

Document ads: expensive per lead, but the leads actually mean it.

A downloadable industry report campaign generated 25 leads at $29.99 each. Expensive compared to click-based campaigns, yes. But a document ad requires someone to stop scrolling, read a preview of the content, and fill out a form. That’s three intentional actions in a row. The cost reflects real intent, and for a high-consideration B2B sale, that’s a lead worth having.

Lead gen form ads: the results depend entirely on what you’re offering.

One campaign I’m running for a leadership consulting firm — a live workshop on organizational performance — generated 17 leads at $71.89 each via LinkedIn’s native lead gen forms. The offer was genuinely compelling (a real workshop with a real method used by Fortune 100 companies). The CTR was 0.59%, which isn’t bad. But the cost per lead stings a little. Native lead gen forms remove friction, but they also remove some of the commitment. Someone filling out a form on a landing page wanted it more than someone who clicked “yes” on a pre-filled LinkedIn pop-up.

Event ads punch above their weight when the audience knows the brand.

Promoting a live workshop for the same consulting client, we hit 4.46% CTR on an event engagement campaign. Small sample, but that number is real. When the audience is warm and the event is specific and interesting (“Culture doesn’t eat strategy, unconscious models do”), people engage. Event ads work when there’s genuine reason to show up. They flop when the event is essentially a sales call in disguise.

Video ads: I’ll die on this hill.

Video underperformed in every account I manage. CTRs as low as 0.04%. High CPMs. In niche B2B direct response, a static image with a sharp first line beats video almost every time. Video has a place — retargeting, brand awareness, warm audiences — but if you’re spending your primary budget on video hoping it drives clicks and leads, the data says rethink it.

The bid strategy thing that nobody talks about enough.

Maximum Delivery outperformed Cost Cap bidding consistently across campaigns. When you’re testing, let the algorithm work. Imposing cost controls before you know what resonates starves your best ads of delivery. Find what works first, then add guardrails.

The one-line summary of everything:

The ad that sounds least like an ad wins every time. Across every account, every industry, every format — the ads that led with a human voice, a specific number, and a named fear outperformed the polished brand creative. Every. Single. Time.

Pain beats promise. Specificity beats inspiration. A story beats a feature list.

If you’re building a campaign from scratch, start with one thought leadership post that tells the most uncomfortable true story in your industry. Get it working organically first. Then put money behind it. Then build your conversion campaigns once you know what your audience actually responds to.

That’s the playbook. If you want help running it — or a second set of eyes on what you already have — that’s exactly what I do. My inbox is open →