Set up your first Meta ads campaign

An end-to-end walkthrough of a first Meta campaign, from the accounts you need through campaign, ad set and ad, to a modest test and what to watch in week one.

Adam Murray4 September 202610 min read
On this page

The first Meta campaign is where the abstract advice about paid meets a tool with a hundred settings, and it is easy to freeze. This walkthrough takes you end to end: the accounts you need before you start, then the campaign, ad set and ad in the order Meta builds them, a modest first test, and what to actually look at in the week after launch.

One thing to establish up front, because it shapes everything below. Meta Ads Manager changes often: the labels move, the settings get renamed, and any specific number can be different by the time you read this. So this guide is deliberately concept-led. It tells you what each step is for and what decision you are making, and it defers every exact button path and current specification to Meta's own documentation, which you should treat as the authority. Where you see a link to Meta, follow it for the current specifics rather than trusting a walkthrough (including this one) to be current on the details.

Before you build: the pre-flight

Three things need to exist before you can run an ad, and getting them in place is genuinely the hardest part of a first campaign. Set them up when you are not in a hurry to launch.

A Facebook Page. Ads run from a Page, not from a personal profile. If your brand does not have one, create it first. Meta explains what a Page is and how to create one in its Business Help Center.

A Meta business account and an ad account. The business account (Meta's organising layer for your Page, ad account, people and payment method) is what keeps your advertising assets in one governed place rather than tangled up with a personal login. Inside it sits the ad account you will actually spend from, with a payment method attached. Meta walks through setting up a business account and an ad account in its Business Help Center; do this properly now, because untangling assets set up the wrong way later is tedious.

The pixel, or dataset. The pixel is the piece of tracking that lets Meta see what happens on your website after a click: whether someone viewed a product, added to a basket, or bought. You do not strictly need it for an awareness campaign, but you need it for anything optimising towards a site action, and you need it to build site-based audiences later, so install it early even if your first campaign does not use it. Modern setups pair the pixel with server-side sending through the Conversions API to recover signal lost to browser and operating-system privacy changes. Meta documents the pixel, the dataset and the Conversions API in its Business Help Center; the deep Meta Ads Manager series covers what the pixel and events do and why the Conversions API exists, and it is worth reading before you rely on conversion data.

With those three in place, you are ready to build. Meta's structure has three nested levels (campaign, then ad set, then ad), and you build them in that order. The dedicated series article on Meta Ads Manager structure explains the hierarchy in full; this is the applied version.

The campaign: choose the objective

The campaign is the top level, and its one big decision is the objective: what you are asking Meta to optimise towards. This choice matters more than any other setting in the whole build, because the delivery system will faithfully pursue whatever you pick.

Choose the objective that matches the rung of the funnel you are actually on, as laid out in how to run ads. For a first campaign with a warm-ish audience and something to click through to, a traffic or engagement objective is a sensible, forgiving place to start; a pure sales objective demands more volume and a working pixel before it delivers well. Do not pick sales because it sounds like the point. Pick the objective whose metric you can actually read at your budget. In Ads Manager you create a campaign and choose your objective from the current list; because Meta revises how objectives are named and grouped, confirm the present options in its Ads Help Center rather than looking for a label from an old screenshot.

At campaign level you may also see a budget option: setting the budget once here and letting Meta distribute it across ad sets, rather than budgeting each ad set separately. For a first campaign with a single ad set it makes little difference; the deep series article on budgets and bidding explains when campaign-level budgeting helps and when it does not.

The ad set: audience, placement, budget

The ad set is the middle level, and it is where most of the real decisions live: who sees the ad, where it appears, and how much you spend.

Audience. This is the targeting choice from how to pick your audience for ads. For a first campaign, resist the urge to over-specify. A broad audience, or a lightly-defined core audience, gives the delivery system room to find responders and avoids the trap of paying more to reach a tiny, over-stacked pool. If you have a warm custom audience from site or engagement data, that is an easy, high-intent place to start. The deep series article on building audiences covers setting each type up inside the tool, including the consent and matching caveats for custom audiences.

Placement. Placements are where your ad can appear: feeds, Stories, Reels and the rest, across Facebook and Instagram. The default is to let Meta decide, which is usually the right call for a first campaign because it lets the system find the cheapest effective placement. What this means for you is a creative decision: to run everywhere, you need your creative in the right shapes, because a landscape video crammed into a vertical placement is a letterboxed compromise. Meta publishes its current placements and their requirements in its Business Help Center, and the series article on creative and placements goes into the trade-offs.

Budget and schedule. Here you set a daily or lifetime budget and, optionally, a start and end. For a first test, a modest daily budget you can sustain for long enough to learn is right, and "long enough" is defined by Meta's learning phase, the period during which a new ad set delivers erratically while the system explores. Meta documents the learning phase and the volume of results it looks for before an ad set settles in its Ads Help Center; the working figure it has published is around 50 optimisation events within seven days per ad set, and because that number can change, confirm it there. The practical consequence is arithmetic you should do before launch:

daily budget × 7  ÷  estimated cost per result  =  results per week

Keep your first campaign to one ad set. Splitting a modest budget across several guarantees none of them settles, which is the most common way a sound first test produces nothing but noise.

The ad: the creative and the destination

The ad is the bottom level: the thing people actually see. Three parts matter.

The creative. Your image or video. For a first campaign, use something you have reason to believe works: ideally a piece that already over-performed organically, since that answers the hardest question, whether the creative holds attention, before you spend. Structuring a Meta campaign from a proven organic winner is the method for that hand-off, including the caveat that an organic win proves the creative and nothing else. Make sure the first seconds work without any of the goodwill your followers extend, because a cold viewer gives you none.

The copy and the format. The text around the creative and the format you have chosen (single image, video, carousel). Keep the ask clear and singular. The series article on creative and placements covers the formats; whatever you choose, check Meta's current creative specifications in its Business Help Center rather than guessing at dimensions, because these change.

The destination. Where a click goes. If there is a click-through, tag the URL with campaign parameters before you launch, not after. Untagged traffic is unattributable and you cannot retrofit it. Use a naming convention you apply the same way every time, so campaign-over-campaign comparison is possible later.

Before you publish, Meta reviews ads against its advertising policies, and paid creative is held to a stricter standard than organic: a claim that passed unremarked on an organic post can get an ad rejected. Read Meta's current advertising standards in its Business Help Center before you build, so a rejection does not surprise you on launch day.

Launch, then watch the right things in week one

Once it is live, the discipline is to leave it alone long enough to mean something, then read it honestly. Both halves are hard.

Leave it alone through learning. New ad sets deliver erratically at first, and numbers from the first several days are a read on the system's exploration, not on your creative. Judging the campaign before it clears learning is reading noise; restarting it because of that noise resets learning and guarantees more of it. The series article on budgets and bidding explains why frequent edits reset the learning phase.

Watch delivery diagnostics, which are trustworthy. Whether the ad set is spending its budget, its reach and frequency, and where delivery is going are first-party operational facts and they are accurate. If it is not spending, your audience may be too small or your budget too thin for the objective.

Read results as evidence, not verdict. Platform-reported conversions are claims about influence under Meta's own attribution model, and they tend to run ahead of what your own systems recorded. Note the attribution window, lead with your own numbers, and do not blend prospecting and retargeting into one average. Reading ad results without fooling yourself is the full method, and the series article on reporting covers which columns matter by objective.

Decide, before you look, what number would make you scale, what would make you stop, and what you would conclude if it fails: bad creative, wrong audience, or wrong offer. Without that last one, a failed first campaign becomes "paid does not work for us", which is almost never the actual finding.

Where Acumin fits

Acumin does the work either side of Ads Manager, and is honest about the seam. Studio makes the creative you put behind the campaign. Snapshot reads your public organic numbers, which is how you find the proven piece worth funding. The Content Strategy cockpit holds the plan the campaign serves.

The honest limit, stated the same way it is stated in the product: reading Meta ad results back automatically is pending Meta's app review. Until that lands, you read your results in Meta's own reporting and enter what you want tracked, while Acumin prepares and drafts. Acumin drafts; you send.

Do this today

Do not build a campaign today. Do the pre-flight instead: confirm your Facebook Page exists, your business and ad accounts are set up with a payment method, and your pixel is installed and firing. Then do the budget arithmetic above with a daily figure you can sustain for a month, and confirm one ad set clears the volume Meta's Ads Help Center currently documents. When those two things are true, you can build a first campaign that has a real chance of teaching you something, which is the entire point of a first campaign.

Written by
Adam Murray
Founder, Acumin

Adam builds Acumin. He spends his days on the same two problems this library is about: working out what a piece of content is actually worth, and getting a brief through production without it turning into something else.

Want this done on your own channel?

Acumin reads your public content and your category and hands back what to make next. Every call comes with its confidence and the evidence it rests on. The first Snapshot is free.