To optimise ads on Meta and TikTok, work in layers. Judge each ad set against your KPIs after about three days, fix or switch off the high-spending ads that underperform inside it, then give the remaining ad sets a weekly review before you move budget. When the primary KPI has too little data to judge, use secondary metrics such as click-through rate, cost per landing page view and conversion rate.
Key takeaways
- Optimisation happens on three levels: ad sets judged over longer and shorter intervals, a KPI-based assessment, and in-depth work on ad sets and individual ads.
- An initial three-day check against predefined KPIs decides whether an ad set continues or stops.
- A weekly review gives promising ad sets time to recover before you write them off.
- Budget decisions depend on the scenario: increasing spend and saving costs call for different moves.
- When primary KPIs lack data, CTR, cost per landing page view and conversion rate show where the problem is.
Mastering ad optimisation is essential on platforms like Meta and TikTok, where advertisers grapple with complex algorithms and changing user behaviour. This guide sets out a structured approach.
Why do Meta and TikTok ads need a structured optimisation process?
Both platforms sell ads through auctions and use machine learning to decide who sees each ad. Both also share the same basic structure: a campaign holds the objective, the level below it holds budget, audience and optimisation event (Meta calls it an ad set, TikTok an ad group), and the ads hold the creative. Every decision in this guide happens at one of those levels.
The delivery systems need time and data to learn, so changing things too often makes results noisier, yet leaving an underperformer running wastes budget. A fixed process with set checkpoints resolves that tension.
What are the three levels of ad optimisation?
Ad optimisation on Meta and TikTok unfolds across three levels:
Longer and shorter interval ad sets: optimisation is split into ad sets with different time frames, so each gets a strategy that suits it.
Initial KPI-based assessment: performance over three days against predefined key performance indicators (KPIs) decides whether an ad set continues or stops.
In-depth ad set and individual ad optimisation: a focused analysis of each ad set identifies the ads that underperform so you can adjust them.
The reason for different intervals is data volume. An ad set that optimises for a frequent event, such as a landing page view, produces enough results to judge within days. One that optimises for a rare, high-value event, such as a purchase of an expensive product, needs a longer window before its numbers mean anything.
How do you run the initial KPI-based assessment?
Set the KPIs before launch: the cost per result, cost per lead or return on ad spend that would make the campaign worthwhile. After three days, compare each ad set against them and decide.
| Result after three days | What to do |
|---|---|
| Comfortably within the KPI | Keep it running and consider it for more budget at the weekly review |
| Close to the KPI | Keep it running and look inside it at ad level |
| Well outside the KPI after meaningful spend | Turn it off |
| Too few results to judge | Decide on secondary metrics instead |
How do you optimise inside an ad set?
Within an ad set, delivery systems tend to concentrate spend on a few ads. Check whether the ads taking most of the budget are also the ones producing results. The priority is the high-spending ad that underperforms: it does the most damage, so pause it or replace it with fresh creative. Low-spending ads matter less, because they are not costing much yet.
Why review performance weekly?
Regular weekly reviews are crucial. A week gives potentially successful ad sets more time to recover and show what they can do.
Daily numbers are noisy. Performance varies by day of the week, and some conversions are reported days after the click, depending on the attribution setting. A weekly view smooths out both, so you compare ad sets on a fair basis and catch genuine trends instead of random swings.
When should you increase spend, and when should you cut costs?
Different scenarios call for either increasing spend or optimising costs, depending on factors such as how many ad sets you run.
- Increasing spend: move budget to the ad sets that beat their KPIs, and raise budgets in steps rather than all at once, because large sudden changes can unsettle delivery. Adding new ad sets with fresh audiences or creative is another way to grow.
- Saving costs: switch off ad sets that miss their KPIs and consolidate the rest. When budget is spread across many small ad sets, none gathers enough data to optimise well, so fewer, better-funded ad sets often perform more steadily.
Which secondary metrics matter when KPIs lack data?
When primary KPIs lack data, secondary metrics such as click-through rate (CTR), cost per landing page view and conversion rate come into play. Each points to a different part of the funnel.
| Metric | What it shows | What a drop suggests |
|---|---|---|
| Click-through rate | Whether the ad and audience fit | Creative fatigue or a mismatch between creative and audience |
| Cost per landing page view | What it costs to get a person onto your page | Rising costs, a slow page or tracking issues |
| Conversion rate | Whether visitors take the action you want | A weak offer, a confusing page or the wrong audience |
What should a daily dashboard for Meta and TikTok include?
We recommend a unified metrics dashboard for Meta and TikTok Ads Managers that covers the key metrics for thorough campaign analysis. Useful columns include spend, impressions, cost per thousand impressions (CPM), clicks and CTR, cost per click, landing page views and their cost, results and cost per result, return on ad spend and frequency. The two platforms name some metrics differently, so agree one set of names and definitions and use them for both.
What are the detailed optimisation steps?
Three levels of ad optimisation: work through ad set and individual ad optimisation, considering both longer and shorter intervals.
Initial KPI-based assessment: assess each ad set's performance against KPIs and turn off the ones that underperform.
In-depth ad set and ad-level optimisation: address high-spending, underperforming ads within each ad set.
Weekly performance review: give ad sets a chance to recover based on their weekly performance.
Scenarios for budget management: choose a strategy based on whether the goal is to increase spend or save costs.
Secondary metrics for insight: use secondary metrics for a deeper understanding of ad performance.
Optimising based on secondary metrics: adjust your optimisation based on what those secondary metrics show.
Effective ad optimisation on Meta and TikTok demands a balance between primary and secondary metrics. This guide offers a roadmap for the complex landscape of digital advertising, and as that landscape changes, the same discipline still applies. For those managing large-scale campaigns, consider automation tools like SOMIN for efficiency and better results; for example, you can score ads before you spend so weak creative never reaches the auction.
Related reading
- Which social media advertising metrics matter, and how do you optimise them?
- How do Google and Meta smart bidding algorithms differ?
- Why do ad costs fluctuate, and how can you stabilise them?
- What is creative clickability, and can you predict it before spend?
- What do ad optimisation platforms promise in their own ads?
FAQ
How long should I wait before judging a Meta or TikTok ad set?
Run an initial assessment after three days against your predefined KPIs, then review performance weekly. The weekly review gives promising ad sets time to recover before you switch them off.
What should I do when my main KPI has too little data?
Use secondary metrics such as click-through rate, cost per landing page view and conversion rate. They show whether the ad, the landing page or the offer is the weak point.
Which ads should I fix first inside an ad set?
Start with the high-spending ads that underperform, because they waste the most budget. Pause or replace them before you worry about low-spending ads.
Why use one dashboard for Meta and TikTok?
A unified dashboard puts the key metrics for both Ads Managers side by side, so you can compare campaigns on the same definitions and spot problems faster.
