Ads Manager: how to use Meta Ads Manager effectively

Same product, same 10 million VND budget, and two people end up with results that differ by 3 to 4 times in profit. The gap is rarely a prettier ad. It comes from the dashboard the winner understands and the loser clicks through at random: Ads Manager. This is Meta's tool for building campaigns, setting budgets, targeting audiences and reading results for ads on Facebook, Instagram and Messenger. Our ads team works in it every day, and this article collects the parts worth your time: the 3-level structure, a step-by-step campaign build, how to read the numbers, and the optimization rules that don't break learning data. It takes about 12 minutes to read, and you can follow along with the tool open.

What is Ads Manager?

Ads Manager, officially Meta Ads Manager, is the cockpit for all paid advertising across Meta's platforms. On one screen you build a campaign, choose who sees the ad, set a daily spend, and then track how each unit of ad money turns into clicks and orders over the last 7 days. You open it by going to adsmanager.facebook.com, or by clicking Ads Manager in the menu of your personal Facebook account.

Don't confuse it with the blue Boost Post button on a fanpage. Boosting is a cut-down version for non-specialists: few controls, rough audience targeting, and most of the money spent buys engagement, not orders. Anyone who treats advertising as a real sales channel works in the full Ads Manager, which has the 3-level structure and the measurement tools this article walks through. Boosting is a toy. This is the real toolkit.

How is it different from Business Suite and the Ads Library?

Meta's tools often get mixed up, so let's set out each one's role once:

In one sentence, the standard routine for someone running ads: once a month, look through the Ads Library to see what the market is saying; twice a week, build and read results in Ads Manager; and leave daily page care to Business Suite. Three tools, three jobs. Don't mix them up.

4 things to prepare before you spend your first dollar

Without a solid foundation, ad money disappears without a trace. Here is a 4-point checklist, and you should finish all of it before you click create campaign:

1. A properly set up ad account: set the right time zone and currency from the start, because neither can be changed later, and attach a payment method in your own name. We cover how to organize accounts for a company in managing ad accounts.
2. A fanpage with some life in it: running heavy spend on a blank page created 3 days ago just invites the system to scan it. Build up content for 1 to 2 weeks first.
3. Meta Pixel installed on your website: the eye that records who viewed a product, who abandoned a cart, who bought. Without it, Pixel every sales ad you run is running blind.
4. A landing page that holds up and is tracked: link with UTM so you can cross-check the numbers from both sides, and make sure the page opens in under 3 seconds on a phone.

These four items take 2 to 3 days to finish. The longest is about 1 hour to install the Pixel, if you already have edit access to the website. This is also the first gap between the two advertisers at the top of this article. Foundation first, money second.

The 3-level structure: campaign, ad set, ad

Everything in Ads Manager is arranged in 3 levels. Once you know which level decides what, the confusion goes away:

• The campaign level answers 'what do you want': awareness, traffic, leads or sales. Pick the wrong one here and the whole chain is off, because the system optimizes for exactly what you say you want.
• The ad set level answers 'for whom, how much money, and where': the audience, budget, schedule and placements.
• The ad level answers 'what do you say': the image, video, headline and call-to-action button.

We make every account follow a naming rule from day one. It costs an extra 30 seconds each time you create something: the campaign name states the objective and the month, the ad set name states the audience, and the ad name states the content idea. It sounds like paperwork, but three months later you will be looking at a report of 40 to 50 rows, each representing 2 to 5 million VND already spent, and a clear name is what keeps you from flying blind. Naming is building the report in advance.

Building your first campaign: 6 steps

The standard flow for a sales objective, checked by us on the current interface on 02/08/2026:

1. Click the To create blue button, then choose an objective from the 6 groups. For sales, choose Sales, and to collect leads, choose Leads. Meta explains each group in its official objectives documentation.
2. At the ad set level, choose the conversion location: on your website through the Pixel, in messages, or right on Facebook. If you sell through chat, also read our article on message ads.
3. Set a daily budget and a schedule. Beginners should start at 200,000 to 500,000 VND a day for one ad set, which is enough for the system to learn without hurting your wallet.
4. Define the audience: location and age, and keep the interests section short. Newer systems with broad targeting usually beat narrow, detailed targeting.
5. Placements: beginners should leave it on shutter manufacturing, so you don't switch off a cheap placement by mistake. Splitting placements by hand is only worth doing once you have numbers to back it up.
6. Build the ad: image or video, copy, call-to-action button, add a link with UTM tags, then click Publish. Ads that break policy are sent back, and the list of common errors is in our article on ads that are not approved.

Review usually finishes within 24 hours. While you wait, don't keep editing back and forth: every big change you save sends the ad back to the start of the review queue, sometimes costing another whole day. Once it's published, leave it alone.

Budget and machine learning: where most people go wrong

Meta's system runs on machine learning. Each new ad set needs about 50 conversions within 7 days to get out of the learning phase and run steadily. That number leads to a few practical rules:

• The budget has to be enough to feed the learning system: if the objective is orders and each order costs 100,000 VND in ads, a budget of 100,000 a day will almost never reach 50 results in 7 days, and the ad set will stay stuck in the learning phase.
• Don't slice the audience too thin: 5 small ad sets sharing a small budget means 5 hungry ad sets. Use fewer ad sets, each with enough to eat.
• Leave bidding on automatic when you start: Meta sells ads through an auction, and how it works is explained on its official pricing page, where beginners who set bids by hand usually lose to those who let the system optimize.

You can set the budget at the campaign level and let the system split it across ad sets, or set it on each ad set to control it yourself. We usually let the system split it when the ad sets target the same kind of audience, and split by hand when we want to push money into one strategic audience. There is no absolute right answer, only the answer you can measure after 7 days of running. Measure first, argue later.

Measurement: set up tracking before you run

This is the driest part of the article and the most valuable. All three layers of tracking have to be switched on:

Layer one is the Pixel on your website. Install it once and it works for years, recording events such as page views, add to cart and purchase. Since iOS tightened privacy, also add the server-side Conversions API so fewer numbers go missing. Layer two is the conversion event, which must be the final goal: optimize for 'purchase' and not for 'page view', and then wonder why plenty of people view but nobody buys. The system finds exactly the kind of person you tell it to look for. Layer three is UTM tags on every link, so you can cross-check against your website's own analytics. Meta's numbers and the real numbers always differ by some margin, and knowing your own margin makes you read the numbers far more reliably than the average advertiser. A gap of 10 to 20% is normal. Three layers of tracking. Skip any one and that part is blind.

Reading reports: 5 columns are enough to decide

Opening the reports tab and seeing 100 columns of numbers is overwhelming, but your daily decisions only need 5 columns. Pin them as your own column set and reuse it every day:

• CPM: the price of 1,000 impressions. It tells you whether the audience is expensive or cheap and whether the system dislikes the ad.
• CTR: the share of people who see the ad and click. Under 1% for a sales ad means the ad is dull.
• CPC: cost per click. Read it together with CTR to tell whether the problem is the ad or the audience.
• Cost per result: the key column. It shows how much ad money one order or one lead is costing you.
• Frequency: how many times, on average, one person sees the ad. Above 3 to 4 times with falling results means the audience is tired of this ad.

Read them in sequence to reach a diagnosis. A high CPM is about the audience and the industry. A low CTR is about the ad. Lots of clicks with no purchases is about the landing page. A high frequency means you've run out of new people to reach. Each number points to exactly one thing to fix. Going through all 5 columns takes about 10 minutes, and the standard rhythm is twice a week. Fix the right thing, and you waste less money on random tests.

Optimizing without breaking the learning data

The most common mistake for newcomers with money to spend: half a day passes with no orders, so they jump in and edit, but every big edit sends the ad set back to the start of learning, and the data it just collected is thrown away. The craft rule our team applies for clients running Facebook ads:

1. Give each ad set at least 3 to 4 days of data before judging it, unless it is spending money and showing a flat zero.
2. Change one variable at a time: change the ad and keep the audience, or change the audience and keep the ad. Change both at once and, win or lose, you won't know why.
3. Raise the budget in steps of 20 to 30% every 2 days, instead of doubling it overnight. A sudden jump sends the system back to learn from scratch.
4. Duplicate instead of editing in place when you want to test a new direction: keep the old ad set running, duplicate it and edit the copy.
5. Switch off by numbers, not by feeling: set a threshold in advance. For example, if an ad set has spent 2 times the target cost with no result, switch it off. If it hasn't reached the threshold, leave it alone and let the system work.

A simple weekly loop: check the numbers twice a week, on Monday and Thursday, 30 minutes each time. Change exactly one variable if you need to, and write down what you changed so next week you know why the numbers moved. Good ad management looks a lot like slow and steady work, and two 30-minute sessions a week is the right pace. The impatient make Meta rich.

FAQ

How much money do you need to start?

Technically, a few tens of thousands of VND a day will run. But to get numbers you can decide on, 200,000 to 500,000 VND a day for the first 2 weeks is a realistic starting level for a sales objective.

Why is my ad not spending?

The most common causes: it hasn't been approved, the budget is too low for the audience, a manual bid is set too low, or the account has hit a spending limit. Check the billing tab and the delivery status first, because most problems are there, and an unpaid ad balance can also stall delivery. Easy problem, cheap fix.

What do I do if my account is disabled?

Go to Account Quality and file an appeal, following the steps in disabled accounts. Prevention is better: pay with a payment method in your own name, don't run ads in banned industries, and don't borrow accounts of unknown origin. Accounts last when you look after them.

Should I hire someone or run ads myself?

With a budget under 10 to 15 million VND a month, teaching yourself and running it by following this article is reasonable, since the cost of trial and error is still lower than an agency fee. With a bigger budget, or when ad spend is the lifeblood of your revenue, hiring a team with a process for reading the numbers every day costs less than the money you would burn by figuring it out alone.

Bottom line: the tool does not win for you, a system does

Ads Manager is the cockpit, and whether you fly high or crash comes down to process: a base of 4 preparation items, a 3-level structure with proper names, a budget big enough to feed the learning system, 5 columns of numbers read twice a week, and one variable changed at a time. All of this comes down to discipline, and discipline can be learned. It is not a talent.

And if you want a team to take on that discipline for you, the digital marketing team at MONA takes on everything from building the tracking foundation and account structure to running campaigns and reporting in real profit-and-loss numbers. We have done this for clients since 2016, as part of MONA's 14,000+ projects. Call 1900 636 648, and let us look at your current ad account for 15 minutes. Using the framework in this article, we will point out 3 places where it is leaking money: whether the learning system gets enough to eat, whether your tracking has enough eyes, and whether your last edit broke the data. After that you are welcome to fix them yourself. Plug the leaks first.