Back to Blog
10 min readE-commerce

Meta Ads 2026 for an Online Store: Where to Learn and How to Structure Campaigns

Last updated: Sep 26, 2026

Short answer: you can learn Meta ads for an online store for free from Meta's own sources: Meta Blueprint courses, the Meta Business Help Center and the Conversions API documentation. The only paid part is the optional certification: the exam costs the equivalent of $99-150 (as of September 2026). After Andromeda, the campaign structure is simple: one sales campaign with Advantage+ on, one budget at campaign level, automatic placements, a broad audience and a few clearly different creatives in 4:5, 9:16 and 1:1. Underneath it, the pixel together with the Conversions API. My own layout (avatar × awareness level × destination) is in a separate post, Meta Ads 2026: My Full Strategy Step by Step.

Starting from zero? Join my free community and ask your questions there. Join the community →

This post is a map: where to learn, and which settings Meta itself recommends in its documentation. Every recommendation links to the Meta page I read on 26 September 2026. Where I give my own opinion or quote a guide from outside Meta, I say so.

What people ask me most

I collected the Meta ads questions people sent me: 22 questions from call booking forms (9), YouTube comments (7), Discord (3) and 1:1 sessions (3). The same topics keep coming back: campaign structure, Advantage+ and Andromeda. A few examples, translated from Polish, with no author details:

  • "I'd mainly like to talk about ad structure." (call booking form)
  • "I have a problem with my campaign strategy. It doesn't work, even though I started doing it the way your course shows." (call booking form)
  • "How does targeting work? Always broad, only changing the creatives? If so, Adv+ or the old method?" (YouTube comment)
  • "Are you using Andromeda yet?" (YouTube comment)
  • "Once a product passes the testing stage, meaning 50 euros a day, do you keep the campaign budget or change it?" (YouTube comment)

Each of these questions is answered by one of the sections below.

Where can you learn Meta ads for free in 2026?

Start with the free Meta Blueprint courses and the Meta Business Help Center. Both are official, free and updated by Meta itself. Pay for someone else's course only after you've worked through these and know exactly what you're missing.

Source What you get Cost
Meta Blueprint self-paced online courses, live and recorded webinars, learning paths free
Meta Certification an exam that validates your knowledge, free study guides and a practice test exam $99-150 (as of September 2026)
Meta Business Help Center how every feature works: Advantage+, budgets, learning phase, creative fatigue free
Meta Ads Guide sizes, aspect ratios and text limits for every placement free
Conversions API documentation technical CAPI setup and event deduplication free

Meta describes the Blueprint courses as free and self-paced (facebook.com/business/learn). That's enough to understand campaign objectives, the auction, budgets and measurement.

My free material

How much does Meta Blueprint certification cost, and do you need it? (as of September 2026)

A Meta Certification exam costs the equivalent of $99-150, and you see the exact price for your country at checkout (Meta FAQ: exam cost). As a store owner you don't need it. None of the Meta pages I read makes certification a condition for running ads, and Meta aims its exams at people building a marketing career.

Certification Level Exam length Valid for
Digital Marketing Associate beginner 90 minutes 24 months
Media Buying Professional intermediate 105 minutes 12 months

A few more facts from Meta's pages (as stated on Meta's site, checked 26 Sep 2026; some pages require login):

  • You take the proctored exam online on your own computer or at a Pearson VUE test center. You book it at least one day ahead (Meta FAQ: where to take the exam).
  • Each exam comes with a free 30-question practice test.
  • Neither exam page lists Polish. Both are offered in English and German, among others.
  • Meta aims the Digital Marketing Associate exam at people early in a marketing career: students, job seekers, career changers.

My opinion: certification makes sense if you want to run ads for clients and put it on your CV. If you run your own store, that $99-150 is better spent on your first days of ad testing. The practice test is free, so use it to find the gaps in what you know.

What did Andromeda change in Meta ads?

Andromeda is the Meta system that narrows tens of millions of ads down to a few thousand candidates for a given person. Meta described it in December 2024 on its engineering blog (engineering.fb.com, 2 Dec 2024). Three points from that post matter for a store:

  • Meta built Andromeda for the fast growth in creative volume coming from Advantage+ creative and generative AI tools.
  • Meta reports +6% retrieval recall and +8% ads quality on selected segments. Those are Meta's figures about its system, not about your account.
  • According to Meta, advertisers who hadn't used Advantage+ creative before and turned on its AI-driven targeting features saw 22% higher ROAS. That's Meta's number too, and you can't verify it from outside.

The engineering post prescribes no account structure. The idea that "creative is the new targeting" comes from industry guides and advertiser practice. I quote Meta only on what it wrote itself: its best practices for Advantage+ campaigns ask for "a wide variety of diverse creative assets" to increase relevance (Meta: Advantage+ sales campaigns).

How should you structure Meta ads campaigns for an online store after Andromeda?

Run one sales campaign with Advantage+ on, one budget at campaign level, automatic placements and a broad audience. Move your control from targeting to creative and measurement. Every element below has a source in Meta's documentation.

Campaign: Advantage+ sales campaign

Meta now calls fully automated sales campaigns "Advantage+ sales campaigns". Older guides call them Advantage+ Shopping. Meta says you get its most advanced optimizations when a sales campaign has Advantage+ on for audience, placements and budget. Meta also says advertisers using these campaigns saw a 9% lower cost per conversion on average (Meta: Advantage+ sales campaigns). Manual settings are still available. Even with Advantage+ off, Meta recommends choosing the broadest settings you can (Meta: Advantage+ campaign experience).

Budget: one campaign budget instead of many small ones

Advantage+ campaign budget ("campaign budget" in Ads Manager, CBO in my other posts) is one budget for the whole campaign that Meta shifts in real time to the best ad sets. Meta says it works best with at least two ad sets and tells you to judge results at campaign level, not per ad set (Meta: Advantage+ campaign budget). Meta's best practices also advise against manually pausing and unpausing ad sets, and against pausing ad sets with low delivery (Meta: campaign budget best practices).

Placements: automatic

Meta recommends Advantage+ placements because the system looks for the cheapest results across all placements at once. It doesn't chase the lowest cost for each placement separately. In Meta's own example, turning off Instagram, which at first glance seemed more expensive, produced fewer conversions (8 instead of 9) at a higher average cost (Meta: how Advantage+ placements work). If you pick placements manually, Meta suggests keeping at least 6 (Meta: ads delivery best practices).

Audience: broad

Meta recommends Advantage+ detailed targeting, which lets the system go beyond the interests you enter (Meta: ads delivery best practices). In practice you set the country you sell to and, if needed, an age range that follows from the product. You don't stack interests.

Creative: diversity, not variants of the same ad

Creative is the lever you have left. In its creative fatigue guidance, Meta tells you to make new ads with an image or video that is "materially different" from the original (Meta: creative fatigue). Ten versions of the same video with a different caption colour give the system no new signal. Different selling angles, different people on camera and different formats do. How I match those angles to awareness levels is in my Meta Ads 2026 strategy and in the creative testing glossary entry.

Formats: 4:5, 9:16 and 1:1

Prepare each creative in three aspect ratios. The Meta Ads Guide recommends:

Placement Aspect ratio Resolution from the Ads Guide
Facebook and Instagram Feed (image, video) 4:5 1440 × 1800 px
Reels and Stories 9:16 1440 × 2560 px
Carousel 1:1 at least 1080 × 1080 px

Sources: image, feed video, Reels, carousel. If you don't have every version, Advantage+ creative can crop or expand video to 4:5 and 9:16 and fit images to other aspect ratios (Meta: Advantage+ creative). Check the result before you publish, because the product can end up out of frame.

The first seconds: show the value right away

Meta's mobile-first guide says to get your main message across in the first 3 seconds and to design video that works without sound, because people usually scroll the feed with sound off (Meta for Business: mobile-first creative). In practice, the product and the problem it solves should be on screen from the first frame, with captions.

Measurement: pixel together with the Conversions API

Meta recommends running the Conversions API alongside the pixel and calls this a "redundant setup" (Meta for Developers: deduplication). According to Meta, CAPI data is less affected by browser loading errors, connectivity issues and ad blockers (Meta: about the Conversions API). Two things to check:

  • On Shopify, WooCommerce and Wix you turn CAPI on through the platform integration in a few clicks, at no extra cost (Meta: CAPI setup options).
  • When the same event comes from both the pixel and your server, Meta requires deduplication. The recommended method is the same event name and the same event_id in both sources. Without deduplication, the system can't tell one purchase sent twice from two purchases.

More on the pixel itself in the pixel glossary entry.

A starter structure for a new store with a small daily budget

On a small daily budget, run one campaign with one ad set and a few clearly different creatives. Meta gives the reason itself: an ad set usually exits the learning phase after about 50 results in the week after its last significant edit (Meta: learning phase). Every extra ad set splits those results into smaller pieces.

Run the numbers on your own example. If a purchase costs you 60 PLN in ads, 50 purchases a week means 3,000 PLN a week, about 430 PLN a day for one ad set. The 60 PLN is an example, not a market average. Plug in your own cost per purchase. Formula: cost per purchase × 50 ÷ 7 = daily budget per ad set. If your budget is far smaller, a second ad set only makes things worse. In my strategy I use several ad sets, one per awareness level. That layout is for a budget that can feed them; on a small budget I start there with two or three levels and add more once the data starts to say something.

Layer Starting setting Why (source)
Campaign 1 campaign, objective: sales, Advantage+ on Meta: enables its most advanced optimizations (1362234537597370)
Budget 1 daily budget at campaign level Meta: campaign budget moves money to where the results are (153514848493595)
Ad sets 1; add a second only when the budget can feed it Meta: ~50 results a week per ad set, combine similar ad sets (112167992830700, 950694752295474)
Audience country + broad, no interest stacking Meta: Advantage+ detailed targeting (950694752295474)
Placements Advantage+ placements Meta: lowest average cost overall (965529646866485)
Ads a few clearly different creatives (angle, person, format); I use 5 per ad set Meta: "wide variety of diverse creative" (1362234537597370); the number 5 is my own layout
Formats 4:5, 9:16, 1:1 Meta Ads Guide
Measurement pixel + CAPI with event_id deduplication Meta for Developers
Optimization event purchase; on a very small budget Meta allows an easier event Meta: harder events need more budget (203183363050448)
Changes no edits during learning, judge the whole campaign Meta: edits reset learning (112167992830700)

A note on optimizing for an easier event, such as landing page views: Meta says this lowers the budget you need. My opinion: the system will then find people who click and browse, not people who buy. Treat it as a temporary fix. How much to set aside for testing at all is in Facebook ads cost for dropshipping and minimum dropshipping budget.

This table is a starting point. This is not a promise of results. Outcomes depend on product, budget, execution, market and experience. Individual results vary.

How often should you refresh Meta ads creative?

Refresh creative when Meta flags fatigue (in a single-creative ad set) or when your cost per result clearly climbs, not on a fixed calendar. Meta shows "Creative limited" in the Delivery column when cost per result is higher than your past ads but less than double, and "Creative fatigue" when it's double or more (Meta: creative fatigue). Note: Meta shows these statuses only for ad sets with a single creative (not for dynamic creative or Advantage+ catalog ads). With several creatives per ad set, as in the structure above, you watch each ad's cost per result against its own earlier results. At that point Meta suggests adding a new, materially different creative and keeping the old one running, since it may still deliver results.

A guide from an ecommerce tracking company (not Meta, source at the end) proposes a fixed rhythm: new hooks weekly, new angles every two weeks, new formats monthly and a full refresh every quarter. Use it as a starting point for planning production. It's not a Meta rule. On a small budget, weekly hooks will quickly outrun what you can actually test.

How I add new creatives to existing ad sets instead of opening new ones is in my Meta Ads 2026 strategy.

The most common campaign structure mistakes for stores

Most mistakes I see come down to splitting the budget too thin and editing campaigns too early. Every point below has a Meta source.

  1. Five ad sets at 20 PLN a day each, optimizing for purchases. None of them gets near 50 results a week, so they all sit in "Learning limited" (Meta: learning phase).
  2. Editing during the learning phase. Any significant edit can reset learning and delays optimization (same source).
  3. Turning off an "expensive" placement. In Meta's example, turning off Instagram produced fewer conversions at a higher cost (Meta: Advantage+ placements).
  4. Pausing ad sets inside a campaign budget. Meta says pausing a low-delivery ad set can raise the cost per result of the others (Meta: campaign budget).
  5. Judging a single ad set instead of the campaign. With a campaign budget, Meta tells you to look at the whole.
  6. Pixel only, or CAPI without deduplication. In the first case you lose some events. In the second you count purchases twice (Meta for Developers).
  7. Ten variants of one ad. Meta wants creatives that are materially different, not small tweaks (Meta: creative fatigue).
  8. Video without captions, with the product showing up in second 5. Meta says to deliver the main message in the first 3 seconds (Meta for Business).

Who this post is for, and who it isn't for

This post makes sense if you have a store or are launching one, want to run Meta ads yourself and are looking for settings you can check against the documentation. It doesn't make sense if you want a guaranteed result or "the setting that always works". You won't find that here or in Meta's documentation. You can ask about your own account in the free community.

FAQ

Do I need Meta certification to run ads for my own store?

No. None of the Meta pages I read makes certification a condition for running ads. The free Meta Blueprint courses give you the same baseline knowledge, and the $99-150 exam is mainly useful if you want to run ads for clients (as of September 2026).

Advantage+ or a manual campaign to start?

Meta recommends a sales campaign with Advantage+ for audience, placements and budget, and keeps manual settings as an option. With Advantage+ off, it recommends the broadest settings possible. For a new store I'd pick Advantage+, because on a small budget you don't have the data to narrow audiences by hand anyway.

Did targeting stop mattering after Andromeda?

Manual audience narrowing matters less, because Meta itself recommends broad settings and Advantage+ detailed targeting. You still set the country and any age range that follows from the product or the law. Different creatives now do most of the work of reaching different groups of people.

How many creatives do I need to start?

Meta doesn't give one number. It asks for "a wide variety of diverse creative" and also warns against too many ads, because each ad then delivers less often and fewer of them exit the learning phase (Meta: managing ad volume). I start with five clearly different ads per ad set. On a small budget, three is fine.

Where to go next

Pick the row that matches where you are:

  • You want a human answer about your own store, free: the Discord — about 1,350 people (September 2026), and I am there every day answering questions.
  • You want the structured course, and you read Polish: GQ Start, 7 days free (PL) — one store, one task a day. It is a Polish-language course; the link is marked so you know what you are clicking into.
  • You already know what you cannot do and want feedback: 1:1 mentoring — it starts with a 30-minute qualification call, and I give the price after that call. I do not guarantee results. Individual participants' results may vary.
  • You want the material every week: Operator on Whop — 200 PLN per 30 days (read from the Whop API, 27 September 2026), weekly calls and material.
  • Free AI agent templates: dawidgac.com/pl/skills.

Sources (accessed 26 September 2026)

Meta, official:

Outside Meta (background only, flagged in the text):

  • An ecommerce tracking company's Andromeda guide, creative rotation rhythm: https://www.wetracked.io/post/andromeda-update-ecommerce

First-party data:

  • Meta ads questions: call booking forms (Calendly, May 2025 - September 2026), YouTube comments, Discord and 1:1 sessions, anonymised.

Tools I use on this page

The links in this section are affiliate links — if you sign up or buy through one of them I earn a commission. Clicking costs you nothing extra; the price is whatever it is on their page.

Want more practical material?

Tick the consent box in the form and I will send you more playbooks on e-commerce, AI, and scaling a store.

I only send emails if you tick the consent box above and confirm it in the email that follows. The data controller is Dawid Gac. Privacy policy

About the Author

DG

Dawid Gac

E-commerce Educator & Entrepreneur

Dawid Gac is a Polish entrepreneur, e-commerce educator, and co-founder of EcomBrain. His public work includes YouTube education, community resources, and a documented 1:1 mentoring method where feedback is attributed to the named coach.

YouTubeInstagramDiscord
Browse all articles