$445 of thought-leader ads, 9 reactions, zero conversions, and €13,450 of work after
I tested the thought-leader ads playbook on my own LinkedIn profile, with my own money, before recommending it to anyone else. The ad account says it failed. What happened outside the ad account says otherwise, with some honest caveats.
Original format Read the full case study as a Google Doc the long version: 16 sections, every table and the outreach data
Key takeaways
- Paid distribution changes who sees your posts, not just how many. 77.5% of paid impressions reached owners, CXOs, partners, VPs and directors. Organic reach on the same posts was 41% decision-makers and 16% entry level.
- Both scoreboards are blind to this channel. The ad account logged zero conversions, and of 18 people asked how they found me, not one mentioned an ad.
- The network was the fuel, the ads the match. A warm audience produced almost everything. A cold US list produced nothing.
- Nothing compounded. Five weeks after spend stopped, reach was back to the pre-campaign floor.
Why I ran it on myself
Most of my case studies are client accounts. This one is my own profile: no company page, no landing page, no lead form, no Insight Tag, no UTMs. The only path was see the post, look at the profile, connect, and I write to you. It's the most honest data I have and the most limited, a sample size of one.
The starting point matters. The profile was quiet, not dead: one post in the 51 days before, 848 impressions. But it sat on about 4,200 connections built over a decade of real working relationships. The ads didn't create that network. They made it visible again.
What ran
24 March to 24 May 2026: 62 days, 34 with delivery, 33 ad units promoting 14 of my posts to three audiences.
| Audience | Spend | Share | Result |
|---|---|---|---|
| A. Founders and CMOs in Romania (a hand-built proxy for my network) | $269.95 | 61% | Produced almost everything |
| B. Romanian diaspora in founder and marketing roles | $85.35 | 19% | Two clients and a strong inbound signal |
| C. Cold ABM: US companies from my incentivized-ads research | $78.95 | 18% | Zero |
| Format test (document ads) | $10.60 | 2% | Inconclusive |
I meant to upload my connections as a matched audience, but the match rate didn't clear LinkedIn's 300-member minimum. So audience A was a lookalike of my network built from targeting facets, not a literal retarget.
What the ad account saw
| Metric | Value |
|---|---|
| Spend | $444.85 |
| Impressions | 18,581 |
| Clicks | 1,412 (7.60% CTR) |
| CPC | $0.32 |
| Landing page clicks | 86 |
| Reactions | 9 |
| Comments | 5 |
| Conversions | 0 |
A 7.6% CTR looks impossible until you see what the clicks were: 1,326 of 1,412 were profile clicks and "see more" expands. That's exactly what you want when the offer is "look at who I am". By any normal dashboard, this campaign should have been killed in week two.
What actually happened
| Period | Days | Profile impressions | Per day | New followers |
|---|---|---|---|---|
| Before (1 Feb - 23 Mar) | 51 | 1,201 | 24 | 118 |
| Campaign (24 Mar - 24 May) | 62 | 32,864 | 530 | 258 |
| After (25 May - 30 Jun) | 37 | 2,319 | 63 | 74 |
A 22x lift in daily impressions, of which 43% were earned on top of the paid ones. The honest counterweight: I also posted 14 times instead of once. Per post, the organic lift was somewhere between 1.2x and 4x. Both are real, they're different claims.
Inbound connection requests went from 0.42 a day to 1.47 during the flight and back to 0.56 after. The people connecting shifted too: founders, owners and C-level went from 27% to 45% of new connections. Romanian posts beat English ones by 3.5x in engagement per post.
Nobody said they saw an ad
My opening message asked new connections one question: how did you find me? 18 answered. Not one used the word ad, sponsored or promoted. Thirteen named my content, three named mutual connections, two named my headline. To the person on the other end, a thought-leader ad is indistinguishable from a post that showed up in the feed. Self-reported attribution can't see this channel, and neither can the ad account.
The outcome
In the four months after the campaign ended, the experiment turned into €13,450 of invoiced work from four client relationships, with about €4,000 a month still recurring. The single biggest piece came through a referral from someone in my network who saw the posts and never became a client. No attribution model catches that path.
What didn't work
- Cold produced nothing. Being clearly relevant to a stranger was worth less than being vaguely familiar to someone in my orbit.
- Video and articles. Video cost 16x more per click than image posts, articles had half the CTR.
- Follow-up. I never sent a second message to anyone who didn't reply, and I didn't answer 17 of the 40 people who did. That's the most expensive mistake here and it costs nothing to fix.
- No tracking. Fine at $445. It falls apart at $5,000.
What it needs to work
- An existing network with real relationships in it.
- A personal profile you'll post from, in your buyers' language.
- Someone who writes messages by hand and answers every reply.
- A high-value, low-volume sale, and reporting that can survive looking bad for a few weeks.
Spend, delivery and profile metrics from LinkedIn Campaign Manager and creator analytics. Attribution of clients was done by hand, and roughly half of the network growth came from my own outreach rather than the ads.