Service
Paid social advertising
Creative is the targeting. We build the system that finds out which creative works.
In short
On paid social, the algorithm does most of the targeting and the creative does the rest. That makes creative volume and clean measurement the two things worth investing in, and audience micro-segmentation mostly a way to fragment your data. We manage Meta, LinkedIn, and TikTok as one system with shared measurement and platform-specific creative.
The problem
Why paid social results decay
Paid social decays faster than search because the audience is not asking for anything. Performance is a function of creative freshness, and a set of ads that worked for two months is usually not being beaten by a competitor, it is being ignored by people who have already seen it.
The measurement problem is worse than on search. Browser-side tracking loses a meaningful share of events to blockers and privacy settings, so the platform under-reports, then optimises toward the fraction of conversions it can still see.
The third issue is structural. Splitting a small budget across nine ad sets means nine ad sets that each stay in the learning phase forever and never gather enough signal to optimise.
Scope
What the work covers
Three platforms, one measurement layer, creative built per platform rather than resized.
Meta advertising
Campaign consolidation so budgets exit the learning phase, Advantage+ used where it earns its place and avoided where it hides useful structure, and creative tested in batches rather than one ad at a time.
LinkedIn advertising
The channel where targeting still does real work, so job function, seniority, and company list targeting are built deliberately. Costs per click are high, which makes offer and landing page quality more decisive here than anywhere else.
TikTok advertising
Native-first creative, because polished brand film reliably underperforms against material that fits the feed. Hook testing in the first two seconds is the main lever.
Server-side tracking
Conversions API on Meta and equivalents elsewhere, with event deduplication so the same conversion is not counted twice. This recovers signal the browser loses and it improves what the algorithm has to learn from.
Creative testing
A test structure that produces a decision rather than a chart: enough spend per variant to be readable, one variable at a time where it matters, and a documented record of what has already been tried.
Remarketing
Segments by depth of engagement, sequenced messaging rather than the same ad repeated, and frequency caps set so retargeting stops being an annoyance you paid for.
Process
How the work runs
The order matters more than any individual step.
Measurement first
Pixel and server-side events audited and deduplicated before spend decisions are made, for the same reason as on search: you cannot optimise toward a number you cannot trust.
Consolidate
Most accounts we take on are running too many ad sets on too little budget. Consolidation is usually the first performance change, and it costs nothing.
Build a creative pipeline
A steady supply of new concepts matters more than any single winning ad, because every winning ad has a shelf life.
Test to a decision
Tests run to a pre-agreed spend threshold and end in a written call: keep, kill, or iterate.
Report on contribution
Platform-reported numbers alongside what your own analytics and CRM saw, with the gap explained rather than hidden.
Deliverables
What you receive
Artefacts you keep, whether or not we keep working together.
- Pixel and Conversions API audit with deduplication verified
- Consolidated campaign and ad set structure
- A creative testing roadmap with results logged
- Audience and exclusion architecture per platform
- Frequency-capped remarketing sequences
- Monthly reporting comparing platform-reported and analytics-observed results
Related reading
Where AI actually belongs in a marketing stack
The long-form version of the thinking behind this service.
Related services
- conversion rate optimization services
Find the friction, form a hypothesis, run it to a decision.
- content marketing services
Fewer pieces, each aimed at one question, each measured.
Questions
Paid social advertising, answered
The questions we are actually asked, answered without hedging.
Which platform should we start with?
Whichever one your buyers already use, which is usually obvious from your existing traffic. If you sell to businesses with a considered purchase, LinkedIn tends to justify its cost per click. If you sell to consumers, Meta is normally the volume channel and TikTok the discovery one.
How much creative do we need?
Enough to keep replacing what fatigues, which is a rate rather than a number. A useful planning assumption is a fresh batch of concepts every four to six weeks, with more required as spend rises because the audience saturates faster.
Why does the platform report more conversions than our analytics?
Because they measure different things. Platforms count view-through and click-through conversions within their own attribution window, and analytics tools use last non-direct click. Neither is lying. We report both and reconcile them rather than picking whichever looks better.
Do we need the Conversions API?
If you are spending meaningfully, yes. Browser-only tracking loses events to blockers and privacy settings, and those losses are not random, so the optimiser learns from a skewed sample. Server-side events recover a large part of that signal.
Can you produce the creative as well?
We handle concept, copy, and direction, and we brief and manage production. We are direct about where an outside specialist will do better than we would, particularly on video.
Start with an audit
Send us what you are running. We will tell you what we would look at first in paid social advertising, and whether it is the right place to start at all.