The ideal campaign structure for a new Meta Ads account

The ideal campaign structure for a new Meta Ads account consists of three separate funnel layers: a cold-audience campaign for new reach, a retargeting campaign for engaged visitors, and a retention campaign for existing customers. Setting up this separation from day one gives Meta's algorithm room to learn quickly without budgets bleeding into each other.

Key takeaways

• A new Meta Ads account needs at least three campaign layers: cold, warm and retention.
• Audience separation is the foundation: cold, retarget and existing customers must never sit in the same ad set.
• Conversion tracking via Meta Pixel or Conversions API must be correctly configured before spending the first euro or dollar.
• Start with broad creatives and focused audiences, then scale based on data, not gut feeling.
• AdBrains automates audience building, budget allocation and creative rotation for new accounts through its own AI modules.

Why structure matters more in Meta Ads than in other channels

Meta Ads operates on a fundamentally different logic than Google Search. Without search intent to guide it, the algorithm has no signal about what users want at any given moment, so you decide who gets reached and the algorithm works within those boundaries. That places the campaign structure in the role of primary steering mechanism. Get it wrong and cold audiences end up seeing the same creative as people who already bought from you, budgets quietly drain toward existing customers, and individual ad sets never accumulate enough data to learn anything useful.

The Meta Business Help Center (2026) is explicit about this: ad sets require a sufficient volume of conversions before splitting or testing becomes viable. For newer accounts, the practical implication is straightforward. Start concentrated. Keep the number of ad sets low and expand only once the foundation is stable. Fragmenting the learning budget across too many variables too soon is one of the most common and most avoidable mistakes, and it is one of the main reasons new advertisers end up with Facebook ads not converting despite a reasonable budget.

In practice, accounts that go live without a structural plan tend to spend their first two or three months firefighting problems that were baked in from day one. A solid foundation saves media budget, and it shortens the algorithm's learning phase considerably.

Step 1: conversion tracking as the starting point

Conversion tracking is not a technical afterthought. It is the foundation everything else depends on. Feed Meta's algorithm the wrong signals and it will optimise toward the wrong outcomes. For a new account that means one clear rule: get the Meta Pixel and the Conversions API (server-side) in place before a single campaign goes live.

The Conversions API routes events directly from your server to Meta. That matters because browser restrictions and iOS privacy updates chip away at the data quality the Pixel can capture on its own. Meta Business Help Center (2026) confirms that running Pixel and Conversions API together produces better signal coverage than relying on the Pixel alone. For e-commerce clients like Elletens.nl or ToetsJeKennis.nl, purchases and add-to-cart events get recorded reliably even when a visitor has an ad blocker running.

Set up the following events as priorities at launch:

• Purchase or Lead: the primary conversion goal you will ultimately optimise toward.
• Add to Cart and Initiate Checkout: signals that feed the retargeting phase.
• View Content: for building a broad retarget pool from day one.
• Complete Registration: relevant for lead generation accounts such as Clima-Active.nl or LeroyBrouwer.nl.

Step 2: building the three campaign layers

Meta Ads funnel structure: throughput per phase

Example calculation: at 10,000 daily impressions

Impressions (cold audience)

Broad reach via interest targeting

Clicks to website

CTR depends on creative quality

Engaged visitors (retarget pool)

Visitors who stayed longer than 15 sec.

Conversions

Purchases or leads from warm audience

Conversion rate structurally improves with proper funnel separation

Meta Ads works with three distinct layers, and each one demands its own thinking around audience, message and budget. The table below shows how those layers relate.

The budget split above is a starting point. It is not a fixed rule, and as the account accumulates data the optimal balance will shift. A new account needs a heavier ToF budget in the opening weeks just to build a retargeting pool large enough to be useful. Once that pool reaches sufficient size, typically several thousand unique visitors, the MoF and BoF campaigns start to perform properly.

Take E-4motion.com (electric folding bikes) as a concrete example. The ToF campaign there targets people with an interest in cycling, sustainable mobility or outdoor activities. MoF retargets everyone who visited a product page without converting. The BoF campaign is narrower: it focuses specifically on visitors who started a test-ride request but never finished it. That is a strong intent signal and it earns its own dedicated campaign.

How AdBrains automates campaign structure

Manual structure vs. AI-driven structure

What changes when AdBrains sets up your Meta Ads account?

Manual management

• One campaign for cold and warm audiences
• Ad sets composed by intuition
• No systematic audience building
• Budget allocation adjusted manually
• Ad copy rarely tested
• Remarketing set up after months

AdBrains AI approachRecommended

• Separate funnel layers from day one
• Audiences based on signal data
• PROD, Incubator and RLSA audiences automated
• Budget allocation dynamic per phase
• Multiple ad variants tested from launch
• Remarketing active immediately at launch

AdBrains has built its own AI infrastructure for setting up and managing Meta Ads accounts, and it goes well beyond what the standard Meta Ads Manager offers. A manual manager sets up the funnel structure once and checks it periodically. Our AI agents, by contrast, monitor the structure continuously and step in the moment something drifts from the optimal pattern.

When a new account launches, the audience management automation system makes sure PROD audiences, Incubator audiences and RLSA-equivalent retarget groups exist from day one and are updated every week. These are not static lists. The AI tracks which users belong to which funnel layer and automatically moves them into the appropriate campaign as their behaviour generates more signals.

The multi-agent verification system is central to how this works. Every change to budget allocation, audience composition or campaign objective gets reviewed by four independent AI agents before anything is applied. In practice this is faster than manual management, because the agents run in parallel around the clock rather than one account at a time.

For creatives, AdBrains uses a system comparable to its RSA improvement setup in Google Ads. Ad variants with low engagement are automatically identified and swapped out for variants that perform better on the signals Meta returns. For a client like Clima-Active.nl, where each quote request carries significant revenue potential, this means underperforming ads are not left running for weeks while budget quietly disappears.

The automatic tCPA/tROAS optimisation system adjusts the bidding strategy per campaign layer daily. In the ToF phase, where reach is the goal, the AI optimises toward CPM. Moving into the MoF and BoF phases, it shifts to Cost per Result optimisation based on each client's actual conversion margins. Matching that frequency and precision through manual management is not realistic.

The server-side signal enrichment module enriches Conversions API events with first-party data before they are sent to Meta, which raises the match rate between Meta users and website events and directly improves how the algorithm optimises. For new accounts with limited historical data this difference is particularly pronounced: the algorithm needs stronger signals to learn at any useful speed, and the infrastructure provides them.

Step 3: ad set structure and preventing audience overlap

Within each campaign layer, the same principle applies: keep ad sets small enough to learn from, large enough to scale. Meta advises (Business Help Center, 2026) targeting audiences of at least several hundred thousand users for the ToF layer, giving the algorithm enough room to optimise effectively.

Audience overlap is one of the most common mistakes in new accounts. When the same user appears in multiple ad sets at the same time, you end up competing against yourself in the auction, which pushes up CPM without adding any reach. A few practical steps to prevent this:

• Always exclude existing customers from cold-audience campaigns using a customer list exclusion.
• Exclude website visitors from ToF campaigns once the retarget list is large enough.
• Build lookalike audiences based on your best customers, not all website visitors.
• Use Meta Ads Manager's Audience Overlap Tool to identify overlapping ad sets before they go live.

ToetsJeKennis.nl sells online exams and courses at an average order value of around €50. Excluding existing students from the ToF campaign means the entire budget goes toward people who haven't bought yet. That sounds obvious. In practice, it gets skipped surprisingly often when building an account from scratch.

Step 4: creative strategy per funnel layer

The message that works for a cold audience differs fundamentally from the message for a warm audience. In the ToF phase, the goal is awareness and appeal: someone does not yet know your brand and has no purchase intent. In the BoF phase, someone has already viewed multiple pages and needs a concrete reason to convert now. This is exactly why creative targeting in Meta Ads has become as important as audience targeting itself: the ad determines who responds, not only who gets shown it.

A practical rule of thumb per layer:

• ToF: brand introduction, storytelling, videos showing the product in context. No direct sales message.
• MoF: product benefits, social proof (reviews, ratings), comparisons with alternatives.
• BoF: urgency, the specific product page the visitor viewed, a unique offer or guarantee.

For lead generation clients like LeroyBrouwer.nl, the BoF phase typically performs best with a direct call-to-action that lowers the barrier: a free intake, a no-obligation quote or an online calculator. Those are the creatives that convert in the warm layer, not the brand introduction video that performs well in the cold layer.

Importance of funnel layers at Meta Ads launch

Priority score per campaign type for a new account (scale 1-10)

Remarketing (warm audience)

9

Lookalike audiences

8

Interest targeting (cold)

7

Broad targeting (Advantage+)

6

Retention / existing customers

5

Budget management in the early phase

A new account needs time to collect data. Meta's algorithm generally needs at least 50 conversions per ad set per week to complete the learning phase, according to Meta Business Help Center (2026). With a low purchase frequency or limited budget, that is not always achievable quickly. Two strategies help accelerate the learning phase.

The first is temporarily optimising toward a higher-volume event, such as Add to Cart instead of Purchase. This gives the algorithm data faster, after which you shift the campaign objective to your ultimate conversion goal. The second is budget consolidation: fewer ad sets, more budget per ad set. Never spread a budget across five ad sets when three achieve the same result.

Example calculation: with a daily budget of €50 for a new e-commerce account like Elletens.nl, it makes sense to start with one ToF campaign and one combined MoF/BoF campaign. Only once both consistently generate 20 or more results per week should you start splitting and testing further.

FAQ: campaign structure for a new Meta Ads account

How many campaigns do you need at launch?

Two or three campaigns are enough. One for cold audiences (ToF), one for warm audiences (MoF/BoF combined), and if you already have a customer base, an optional retention campaign. Launching with more than that fragments your budget and drags out the algorithm's learning phase considerably. Add campaigns later, once you have enough conversion data to justify them.

What is the minimum budget to start a Meta Ads account effectively?

Meta has no official minimum, but in practice you need at least €20-30 per day per campaign to get through the learning phase in a reasonable timeframe. At €30 a day with an average cost per lead of €15, a single campaign produces roughly two leads daily, which is enough to get the algorithm optimising. Accounts with a higher order value, like E-4motion.com undefined, may find it worth pushing the daily budget higher and optimising toward a higher-volume event in those first weeks. Mapping out these start up advertising costs per campaign and per growth phase before launch prevents budget surprises once the account goes live.

When should you add Advantage+ Shopping for e-commerce?

Advantage+ Shopping Campaigns are Meta's automated campaign type: the system picks its own audiences and placements. They tend to perform well once an account has built up some conversion history and the product catalogue has enough signal behind it. For a brand new account with no historical data, the better approach is to build a manually structured funnel first and bring in Advantage+ as a supplementary campaign once measurement is solid, generally somewhere between four and eight weeks in.

How do you handle a small or non-existent retarget audience at launch?

A new account starts with no retarget pool, so MoF and BoF campaigns will have limited reach for the first few weeks and conversions will be sparse. Put more budget behind the ToF campaign early on to fill that pool faster. If you have an email database, upload it as a customer list straight away and build lookalike audiences from it. That combination speeds up warm audience growth noticeably and gives MoF campaigns something meaningful to work with sooner.