Services consulting
Paid social where creative is the strategy
Paid social consulting means running Meta, LinkedIn and TikTok campaigns where the creative does the work that targeting used to do, and where success is measured by incrementality rather than by the platform's own attribution.
What actually determines paid social performance in 2026?
The common thread in platform documentation and practitioner consensus is that targeting has become a secondary lever while creative has become the primary one. Broad targeting with a genuinely good ad outperforms narrow targeting with a mediocre one, and the gap widens as targeting controls get less precise.
This inverts the structure most teams still run, where targeting is where the strategy lives — dozens of ad sets built from an audience research meeting — and creative is a slot to be filled. In practice that allocation produces a large number of mediocre creatives measured against too many segments, and you cannot tell which is working.
The second shift is measurement. Platform-reported conversions have always been optimistic, and view-through attribution in particular. As privacy protections reduce signal, the platforms have leaned harder on modelled conversions. The result is that platform numbers are increasingly a modeling output rather than an observation, and treating them as fact is the root of most disappointing quarterly reports.
Why is creative the main lever?
Broad targeting means the algorithm decides who sees your ad. What it optimizes against is response. If the creative does not stop the scroll, no amount of targeting precision saves it — and you cannot target your way out of that.
The creative is also where nearly all the controllable performance lives, which makes it the highest-leverage investment available and the easiest thing for a team to under-resourcing.
- Hook in the first two seconds. Motion, sound-on-off, and a first frame that is not your logo. The logo can come later.
- One idea per asset. Three messages in thirty seconds communicates none of them.
- Show the problem concretely. Specific beats aspirational: the actual spreadsheet, the actual moment, the actual frustration.
- Format for the placement. A 9:16 asset and a 4:5 asset are different creative, not the same video cropped.
- Write for sound-off. Most views start muted; captions are not optional.
- Build in a natural, obvious next step. A single specific call to action outperforms three options.
A useful discipline is to write the creative as text first — hook, problem, mechanism, proof, call to action — and only then produce the asset. Production is the expensive, slow part; if the idea is weak, expensive production of a weak idea is still a weak ad.
How do you test creative properly?
Most teams test creative badly in one of two ways: they test so many variants that none reaches significance, or they change the creative and the audience together and learn nothing.
- Fix the structure first. One audience, one objective, one conversion event, broad targeting. Now the only variable is the creative.
- Test ideas, not polish. The hook and the concept are worth testing. A color change is not, unless you have exhausted the ideas.
- Judge on the metric that matters. Cost per qualified outcome, not click-through rate. CTR is a proxy that rewards clickbait.
- Run to significance or do not conclude. A creative that looks like a winner after three days is usually noise. Check whether the result would survive a reasonable sample.
- Kill losers fast, keep winners long. Losing creatives rarely recover. Winning creatives often fatigue on a schedule you should anticipate.
Is LinkedIn worth it for B2B?
LinkedIn is worth it for B2B when the buyer is a professional with a business role and a specific problem your product addresses. It is expensive per click — typically several times consumer social — and that cost only works when the conversion value justifies it.
The failure mode is running LinkedIn with a consumer-style campaign: a broad audience, a short video, a generic landing page. At LinkedIn's cost per click that cannot work. What works is narrow job-title or function targeting, a specific offer, and a landing page that continues the conversation rather than restarting it.
| Situation | LinkedIn suitability | Notes |
|---|---|---|
| B2B, high contract value, defined roles | Strong | Cost per click is defensible when deal size justifies it |
| B2B, SME with a narrow market | Conditional | Test with a small budget before committing |
| B2B, unclear persona or generic offer | Weak | Targeting has nothing to target on |
| Consumer or low-margin ecommerce | Poor | Cost per acquisition rarely works |
| Recruitment, events, professional content | Strong | Often the highest-value use on the platform |
How do you measure incrementality?
Incrementality answers the question that actually matters: what would have happened without this spend? Platform attribution cannot answer it, because the platform is not a neutral observer of its own effect.
| Method | How it works | Strengths | Weaknesses |
|---|---|---|---|
| Geo holdout test | Pause ads in matched regions and compare | Relatively simple, genuinely causal | Needs volume, regional differences can confound |
| PSA control | Run ads that are deliberately uncompetitive | Cleanest causal read available | Costs money, needs scale |
| Conversion lift study | Platform-run randomised test | Rigorous, platform handles it | Platform-defined, proprietary reporting |
| Budget step test | Increase spend in steps, watch for non-linear response | Uses data you already have | Slow, confounded by external events |
| Time series with intervention | Model the series, model the intervention | No additional spend | Needs history and statistical care |
The pragmatic answer for most businesses is a geo holdout: pick matched regions, stop spend in some, keep it in others, and compare the difference against what you would have expected. It is imperfect and it is far better than trusting a number the platform has an interest in inflating.
Which platform for which business?
Broadly, and with the caveat that creative quality matters more than placement.
| Platform | Best fit | Main constraint | Watch for |
|---|---|---|---|
| Meta | Broad consumer reach, retargeting, most ecommerce | Ad blockers and modelled conversions | Creative frequency fatigue |
| TikTok | Younger audiences, native video, category discovery | Less predictable conversion economics | Requires genuinely native creative |
| B2B with defined roles, high contract value | High cost per click | Consumer-style creative fails completely | |
| Retail, planning-intent, high AOV categories | Narrower audience fit | Underrated for some ecommerce niches | |
| YouTube | Consideration, product education, B2B video | Higher production cost, longer payback | Works best with a clear next step |
Related services
Frequently asked questions
How much does paid social management cost?
That is a commercial question rather than a technical one, and the honest answer is that it depends on scope, creative volume and how many platforms are involved. What matters more is the pricing model: any arrangement where the agency takes a percentage of your media spend misaligns incentives, because their revenue then grows when you spend more regardless of whether it works. We do not work on that basis.
Is Meta advertising still effective in 2026?
Yes, with the caveat that reported performance increasingly reflects modelled conversions rather than observed ones, as privacy protections reduce signal and the platforms lean on inference. That makes platform numbers less reliable than they were, not the channel less useful. The practical response is to measure incrementality independently rather than relying on what the platform reports.
How many creatives do we need to test per month?
More than most teams produce, which is why creative production capacity is usually the real constraint on performance. A reasonable minimum for a business running paid social actively is several distinct concepts per month rather than several color variants of one concept. If you cannot produce that, the constraint to fix is production, not targeting.
Should we run paid social and paid search at the same time?
Frequently yes. They reach the same people through different mechanisms and intents, and search captures existing demand while social creates it. The complication is attribution: both will claim credit for the same conversions, and you will need a method beyond platform reporting to know what each contributed. Where budget is constrained, we generally test rather than scale both.
How do we know if our ads are actually incremental?
You need a control. The most accessible method is a geo holdout test — pause spend in matched regions and compare the result against regions where it continued. It requires volume and careful region matching, and it is imperfect, but it is a genuine causal read rather than an attribution model. Platform conversion lift studies are more rigorous and cost more.
Is creative really more important than targeting?
Broad targeting has made it so, because the algorithm decides who sees the ad and optimizes against response. If the creative does not stop the scroll, targeting precision cannot rescue it. There are situations where a genuinely narrow audience — a specific role on LinkedIn, a retargeting list — still matters a great deal. The general pattern is that creative quality has become the higher-leverage investment, not the only one.
What is creative fatigue and how do we manage it?
The normal lifecycle of an effective ad: performance holds, frequency climbs, and efficiency declines as the same people see it repeatedly. It is scheduled rather than accidental. The teams that perform well plan the refresh before the decline is visible, which means a production pipeline rather than a single ad built once and left running for a year.
Should we hire an agency or run paid social in-house?
In-house if you have someone who can produce creative at volume and diagnose performance weekly — that combination is uncommon, because it requires both creative skill and analytical patience. Agency if you need both capabilities and do not have them, particularly for production. A hybrid often works: an agency produces creative while your team handles the account. The failure mode in every case is creative production being under-resourced.
Need paid social for your business?
Get a free, no-obligation review of where you stand today on paid social — and the three changes that would move the number that matters. No pitch deck.
Book a free 30-minute audit