Home About
Case Studies Community Pricing Blog
Book the teardown
Channels · Paid Social

Targeting stopped being the lever. Creative became it.

Broad targeting and automated placement now beat hand-built audiences on most platforms. The algorithm decides who sees the ad; the creative decides who it goes looking for.

Does paid social work for B2B? Depends on the deal shape, not the industry. Self-serve motions, prosumer overlap, and founder-led categories work well on Meta. Six-figure enterprise deals with named buying committees belong on LinkedIn, where you can buy the title. Running the wrong platform for your deal shape is the expensive error, not running social at all.

Fit

Where paid social works, and where it doesn't.

Where it works

  • Self-serve or product-led motions where an individual can start without procurement
  • Categories with a prosumer edge, which now includes most AI products
  • Anywhere you can produce creative at volume

Where it doesn't

  • Small TAM with named accounts, which you buy on LinkedIn, precisely
  • Teams that can't ship more than a few concepts a quarter
  • Products that can't be evaluated without a sales conversation
Mistakes

Three expensive mistakes.

Never exiting the learning phase.

Meta ad sets need roughly 50 conversions a week to stabilize. Six ad sets on a budget that supports one means nothing ever stabilizes, and every result is noise being read as signal.

Confusing variants with concepts.

Forty assets that are one argument in four colorways is one test. Concepts are different arguments: a different objection, a different proof, a different person on camera. Twenty real concepts teach more in a month than two hundred variants.

Running LinkedIn as a volume channel.

LinkedIn's cost structure punishes volume goals. Lead gen forms convert several times better than landing pages and produce leads who don't remember filling anything in. Right call sometimes, wrong call when quality is the constraint.

How we run it

What actually gets built.

Creative as the Input

20–40 concepts live, 5–10 hooks each, weekly review with a published win rate.

Consolidated Ad Sets

Enough budget per ad set to clear the learning threshold, rather than a structure that looks tidy and never stabilizes.

CAPI With Deduplication

Server-side events with proper event IDs and match-quality monitoring. Misconfigured CAPI silently drops or double-counts 20–40% of signal.

Platform Fit by Deal Shape

Meta for self-serve, LinkedIn for named committees, Reddit for technical categories that punish anything that reads like an ad, TikTok only with a demo-able moment.

When not to run it

The honest part.

If your ACV is under roughly $15–20K, LinkedIn's CPMs rarely pencil. If your buyer list is small enough to name and your deal is six figures, Meta is the wrong instrument regardless of how good the creative is.

Questions

What buyers ask first.

For self-serve and prosumer-adjacent products, consistently. For enterprise deals with long committee cycles, rarely, and no creative budget fixes that.
20–40 live for accounts above $100K a month. Below that, 10–15, with the discipline that they're distinct arguments rather than variants.
Above roughly $20K ACV with a defined committee, yes. Below it, the precision is real and you can't afford enough frequency to use it.

Related: Creative · Paid Search · Measurement

Start here

Find out what's actually broken.

Thirty minutes, direct with Jer. He shows up with findings, not discovery questions.

Not ready to talk? Run the Leak Report →