A Facebook account can report a healthy return on ad spend while the business is quietly losing margin, attracting low-quality leads or exhausting its best customers. That is why knowing how to audit Facebook campaigns means looking beyond the dashboard’s headline ROAS. A useful audit connects platform performance to conversion quality, contribution margin and the capacity to scale without breaking efficiency.

For established eCommerce and lead generation teams, the goal is not to find a dozen minor tweaks. It is to identify the few constraints holding back profitable growth, then build a clear testing and optimisation plan around them.

Start a Facebook campaign audit with the business model

Before reviewing campaigns, define what a good result actually means for the business. Meta can optimise towards purchases, leads or landing page views, but none of those outcomes automatically equals profit.

For eCommerce, establish the allowable acquisition cost by product category, average order value, gross margin, repeat purchase rate, fulfilment costs and any discounting. A £50 first-purchase CPA may be excellent for a high-retention subscription product and unacceptable for a low-margin, one-off purchase.

For lead generation, check what happens after the form is submitted. Review lead-to-qualified-lead rate, booked-call rate, show rate, close rate and customer value. A campaign producing cheap leads can look efficient in Ads Manager while wasting significant sales-team time.

This step also prevents a common audit error: judging every campaign against one blended CPA or ROAS target. Prospecting, retargeting and retention activity play different roles. The question is whether the combined system is producing incremental, profitable demand.

Confirm tracking before trusting any result

If tracking is incomplete, campaign-level optimisation becomes guesswork. Start with Meta Pixel and Conversions API coverage, then compare Meta-reported purchases or leads against the source of truth: the eCommerce platform, CRM or back-end sales data.

Small discrepancies are normal. Attribution windows, consent settings, cross-device behaviour and reporting delays mean numbers will not match perfectly. Large or inconsistent gaps deserve immediate attention, especially when Meta reports more conversions than the business can verify.

Check that key events fire once, contain accurate values and are prioritised appropriately. For eCommerce, that usually includes ViewContent, AddToCart, InitiateCheckout and Purchase. For lead generation, inspect Lead, CompleteRegistration, Schedule and any qualified-lead or closed-won events that can be returned through offline conversion tracking.

Also review event match quality, deduplication between browser and server events, domain verification and UTM consistency. Good attribution does not make a weak offer work, but it gives the team a reliable basis for deciding where to spend the next pound.

Review account structure and budget allocation

A clean account structure makes performance easier to interpret. Look for campaigns that have a distinct job, a sensible optimisation event and enough budget to generate meaningful learning. If 15 ad sets each receive a small daily budget, the account may be creating fragmentation rather than control.

Assess how budget is divided between prospecting, retargeting and existing customers. Retargeting often delivers the strongest platform ROAS because it reaches people already close to buying. That does not mean it should absorb most of the budget. Its available audience is limited, and overfunding it can increase frequency while claiming credit for conversions that would have happened anyway.

Look for unnecessary audience overlap, duplicated ad sets and campaigns competing for the same people. Consolidation can improve delivery and simplify decision-making, particularly where conversion volume is modest. However, do not consolidate simply because it looks tidier. Separate campaigns can be justified when products, markets, offers or optimisation goals genuinely differ.

Budget changes deserve scrutiny too. Frequent, aggressive edits can reset learning and obscure what drove a result. Strong accounts have a clear process for scaling proven campaigns, reducing spend on deteriorating activity and protecting experiments from premature decisions.

Audit Facebook campaign performance by funnel stage

A campaign audit should diagnose where performance breaks down rather than treating CPA as the only metric. Review the journey from impression to revenue.

At the top of the funnel, examine CPM, reach, frequency and outbound click-through rate. A high CPM is not automatically bad. Premium audiences and competitive periods cost more. But if CPM rises while click-through rate and conversion rate fall, creative fatigue, poor audience fit or weak market demand may be the real issue.

At the click stage, focus on outbound CTR and cost per landing page view rather than link clicks alone. A material gap between link clicks and landing page views can indicate slow page load times, broken redirects or accidental clicks from poorly placed creative.

Then assess landing page view to conversion rate. If ads earn attention but purchases or qualified leads remain weak, the bottleneck may sit on the site: unclear value proposition, weak proof, unsuitable mobile experience, delivery surprises or a form that asks for too much too soon. Paid media and conversion rate optimisation need to work together here.

Finally, segment results by new versus returning customers where possible. A high overall ROAS built mainly on repeat buyers gives a distorted picture of acquisition performance. For scaling decisions, customer acquisition cost and new-customer revenue matter far more.

Evaluate creative as a performance system

Most Facebook accounts do not fail because there are too few targeting options. They stall because the creative programme is too narrow, too slow or too dependent on one winning advert.

Review creative at the concept level, not only by individual ad. Group adverts by angle: product demonstration, problem-solution, testimonial, founder story, price comparison, objection handling or offer-led messaging. This reveals which messages create demand and which only produce cheap engagement.

Assess the first seconds of video, the opening image, copy-to-creative alignment and the landing page promise. The advert should make a clear claim, earn attention quickly and set honest expectations for the click. A dramatic hook that attracts the wrong customer can lower CPA while reducing conversion quality.

Frequency is useful context, not a fatigue verdict on its own. A frequency of five can be completely acceptable for a small, high-intent retargeting pool, while a frequency of two may be excessive for a broad prospecting audience if click-through rate is already declining. Compare frequency with spend, reach, CTR, CPA and the age of each creative.

The best remedy is a repeatable creative testing cadence. Build new variations around proven messages, while regularly testing genuinely new concepts. Changing a headline or background can help, but it rarely creates the next major growth lever on its own.

Check audiences, but avoid targeting theatre

Review audience performance across broad targeting, customer lists, lookalikes, interest groups and retargeting pools. The aim is not to declare a permanent winner after a short reporting period. Meta delivery changes, audience sizes differ and results can shift with creative or offer changes.

Broad targeting often works well when conversion tracking is strong and creative clearly signals who the product is for. Interest targeting can still be useful when there is a distinct niche, a new account needs directional control or an offer has a very specific buyer profile. It depends on conversion volume, geography, product awareness and the quality of the first-party data available.

Check exclusions carefully. Existing customers, recent purchasers, employees and converted leads should usually be handled deliberately so prospecting reports do not become inflated. For lead generation, suppressing poor-fit leads or completed applications can improve both efficiency and customer experience.

Turn findings into a prioritised action plan

An audit only creates value when it leads to decisions. Rank findings by expected commercial impact, confidence and effort. Fix broken purchase-value tracking before debating small audience tests. Address a landing page that converts at half the expected rate before increasing a campaign budget.

A practical plan normally includes three workstreams: measurement fixes, efficiency improvements and growth experiments. Give each action an owner, a deadline, a success metric and a decision rule. For example, a creative test should state what result would justify further spend, not merely that new adverts will be launched.

Review results on a consistent rhythm, with enough data to avoid reacting to normal daily variation. At the same time, do not wait for perfect certainty when a serious tracking issue or sharp decline in lead quality appears. Good operators combine disciplined analysis with decisive action.

Profitable scale comes from an account where tracking is credible, creative keeps improving and every pound has a defined role in the funnel. When the team can see the real constraint clearly, Facebook campaigns stop being a collection of adverts and become a measurable growth system.