Meta AdsPaid mediaBeginners
Meta Ads for Beginners: Your First Campaign, Step by Step
The structure, the objective, the budget and the first seven days. A beginner's guide to launching a Meta Ads campaign without burning money on the learning curve.
By Dukun Ads9 min read
Meta Ads is not complicated, but it is unforgiving of guesswork. Almost every expensive beginner mistake comes from one of three things: the wrong objective, too many ad sets, or changing everything on day two. This guide walks the first campaign end to end so you make none of them.
The three levels, and what each one controls
Everything in Meta Ads lives in a three-level tree, and each level owns exactly one kind of decision. Confusing them is the root cause of most bad accounts.
| Level | You decide | You do NOT decide here |
|---|---|---|
| Campaign | The objective — what result you are buying | Audience, creative |
| Ad set | Budget, schedule, audience, placements | The objective, the creative |
| Ad | The creative: image or video, text, headline, call to action | Budget, audience |
One campaign, one ad set, two or three ads. That is the correct shape for a first launch. More ad sets means the same budget split across more auctions, which means none of them gathers enough data to learn.
Choose the objective by what you can actually measure
The objective tells Meta which people to find. If you pick Traffic, it finds people who click a lot. Those are not the same people who buy. Pick the objective that matches the deepest event you can reliably measure.
- Sales — use this when a pixel is installed and firing a purchase event. It is the only objective that optimises toward revenue.
- Leads — use this when your conversion is a form fill or a WhatsApp message rather than a payment.
- Traffic — use only for genuine top-of-funnel testing, or when you have nothing installed yet. Expect cheap clicks and poor quality.
- Engagement and Awareness — useful for warming an audience or testing creative angles cheaply, not for direct response.
Set a budget that can actually learn
Meta's delivery system needs roughly fifty conversion events per ad set per week to stabilise. You do not need to hit that number immediately, but your budget should make it conceivable. A rough sanity check: if you expect a cost per result of $5, a $5 per day budget produces one result a day and will never leave the learning phase.
As a starting point, set a daily budget of five to ten times your target cost per result, run it for seven days, and do not touch it. If that number is uncomfortable, choose a cheaper conversion event — optimise for a lead instead of a purchase — rather than lowering the budget below the learning threshold.
Targeting: start broader than feels safe
The instinct is to stack interests until the audience feels precise. In 2026 that instinct is usually wrong. Meta's delivery model finds buyers inside a broad audience faster than your interest guesses do, provided it has a conversion signal to learn from.
- Set the country or city you can actually serve.
- Set a wide age range — narrow it later based on the age breakdown in your reporting, not on assumption.
- Add at most one broad interest, or none at all.
- Leave placements on the automatic setting for the first campaign.
If you want a sense of scale before launching, an audience reach estimate will tell you roughly how many people the definition covers. Anything above a few hundred thousand is workable for a first campaign.
Write the ad: one idea, one proof, one instruction
Beginner ad copy fails by trying to say everything. A working ad says one thing, gives one reason to believe it, and asks for one action.
Hook : the problem, in the reader's own words
Proof : a number, a screenshot, or a specific outcome
Offer : what they get and what it costs
CTA : one instruction — "Shop now", "Send us a message"Make two or three variations that differ in the hook only, not in the whole ad. When one wins you will know why. If you want a starting draft, the AI Ad Studio inside the Meta Ads dashboard generates copy variations from a short brief with the objective you have chosen — primary text, headline, description and a suggested call to action.
The first seven days
This is the part beginners get most wrong. The urge to fix things on day two is overwhelming and almost always destructive, because every meaningful edit resets the learning phase.
| Day | What to do |
|---|---|
| 1 | Launch. Verify the pixel is receiving events. Then close the tab. |
| 2–3 | Check delivery only: is it spending? Are impressions accruing? Change nothing. |
| 4 | Look at CTR and cost per result. Still change nothing unless spend is zero. |
| 5–7 | Let it run. Note which ad is winning. |
| 8 | First real decision: pause the losing ads, keep budget and audience the same. |
After the first week, change one variable at a time and give each change another week. Ads is a slow instrument. Treating it like a fast one is the most expensive habit in the discipline.
What to look at, in order
- Cost per result — the only number that decides whether this continues.
- CTR — below roughly 1% usually means the creative or audience is wrong.
- Frequency — above 3 in a short window means your audience is too small and fatigue is starting.
- ROAS — only meaningful once purchase values are reported correctly through the pixel.
Everything else is diagnostic detail. Reach, impressions and CPM explain why your cost per result moved; they never decide anything on their own.
A safe first launch
One campaign. One ad set. Three ads. Broad audience. Budget at five to ten times your target cost per result. Pixel installed and verified. Seven days untouched. Do that once and you will have learned more than a month of tutorials.
Dukun Ads connects your Meta Ads account alongside your organic posting, so campaign performance and content performance sit in one dashboard — with automatic daily or weekly reports pushed to Telegram, WhatsApp or email. The Meta Ads guide walks through connecting it.