You do not have to begin with a traffic campaign simply because it is your first ad. If your aim is to increase purchases on your own store and you can collect purchase events correctly, first consider using the Sales objective with purchase-based optimization. Traffic campaigns have a role when visits themselves are the goal, or when you send people to an external page where the final purchase is difficult to measure.
When preparing a first campaign, advice such as “your pixel is new, so you need visitors first” can make the sequence of campaigns seem like the main decision. Rather than ask how old the account is, I first ask what should actually increase if this campaign succeeds. A campaign that needs more visits and one that needs more orders require different settings and different criteria for reading the results.
Clicks and purchases are different requests
Depending on the selected settings, a traffic campaign focuses on increasing actions such as link clicks or landing page views. The Sales objective can focus on conversion actions tied to revenue, such as purchases. Meta's official 2026 study guide, page 69 Choosing the campaign objective's name is not the end of the setup. The ad set's conversion location, performance goal, and conversion event must all support the same purpose. Meta's training also treats these as settings to align together. Meta guidance on campaign objectives and conversion settings
Even after selecting Sales, for example, check what the subsequent settings ask the system to maximize. If the aim is more purchases on your own store, examine whether using the website as the conversion location and optimizing conversions around a purchase event fits that aim. Follow the interface guidance for the options available to your account, but make sure you can explain the final setup in one sentence: “Increase purchases on the website.”
This does not mean traffic campaigns never produce purchases. A person who intends to buy may click the ad. But expecting orders from a setup designed to increase clicks is different from configuring for more purchases and evaluating the campaign by purchases. Cheap clicks alone do not establish that purchases are cost-effective.
If the goal is sales, check measurement and checkout first
Before starting a sales campaign, follow the journey from the product detail page to order completion yourself. To ask the ad system to find buyers, you need to know which action you are reporting as a purchase. The Meta Pixel and Conversions API are tools for sending website customer actions to Meta and optimizing advertising. How they are installed depends on your store platform and its integrations. Meta Pixel and Conversions API guidance
Do not stop at the statement “the pixel is installed.” Ask the site operator or implementation specialist to check the following with you.
- Are product views, add-to-cart actions, checkout starts, and completed purchases recorded as distinct actions?
- Can any route register a purchase before payment is completed?
- Are any real orders counted repeatedly or missed?
- Are the purchase value and currency sent correctly against the order record?
- Do the data source and event selected in the ad set belong to this store?
These are checks to use before starting a campaign, not findings from an inspection of a particular account for this article. Compare test orders with collected events. If data is sent from both the browser and server, also ask the implementation specialist to check deduplication for the same order.
Separately from measurement, you also need to try checkout on a phone. Check whether the advertised price matches the actual price, the ad leads to an in-stock option, shipping costs appear at a clear point, and registration or payment does not block progress. Changing the ad objective will not fix a checkout error. In that case, the order journey needs attention before the ad settings do.
Some situations do suit a traffic campaign
Recommending that you consider Sales does not mean Traffic is useless. Sending people to a particular page may itself be the campaign's role, as in the following situations.
| Situation | Approach to consider first | Results to examine alongside it |
|---|---|---|
| You want people to read a public guide or an update | The Traffic objective with a performance goal suited to page visits | Landing page views, reading, and subsequent actions |
| You operate a store where purchases can be measured | The Sales objective with conversion settings aligned to purchases | Purchase count, cost per purchase, and order records |
| You send people to an external marketplace and cannot connect purchase events | Define the campaign's scope around visits you can measure | Clicks and visits, plus separate performance data supplied by the marketplace |
| You collect contact details for follow-up consultations | Settings that match the actual inquiry process, such as the Leads objective | Inquiry count and the proportion that lead to qualified consultations |
The table illustrates choices for different situations. Some external marketplaces support purchase measurement through an integration, so check the available capabilities first. If you cannot connect purchase data, a gap remains between “we sent visits” and “those visits increased orders.” It is better not to report the two as the same achievement.
After choosing Traffic, distinguish link clicks from landing page views as well. Someone may click a link but close it before the page opens. If reaching the page is the objective, there is a reason to consider landing page views among the available performance goals. Meta introduced this metric to help advertisers understand visits that actually arrive after a click. Meta's explanation of landing page views
Arriving does not mean the visitor has read the content properly. Check subsequent actions that fit the content's purpose, such as viewing a section midway through a guide or clicking the next instruction.
When purchase data is scarce, review the scope before changing the objective
It is harder to make decisions when a new store has little purchase data. Still, avoid treating a fixed period of traffic advertising as an automatic preparation phase that completes purchase optimization. More visitors and sufficient information about actual buying behavior are different things. Meta's ad systems make predictions for multiple objectives, including clicks, video views, and conversions. Visit counts alone cannot tell you that purchase learning is complete. Meta's explanation of objectives in ad prediction models
I would first narrow down what to learn about one product. Splitting a small budget across many products, audiences, and ad sets can leave too few results in each to tell you what to improve. This is an operating principle for obtaining useful evidence from an attempt, rather than a fixed formula for account structure.
For example, a weekly budget of KRW 300,000 and an assumed cost per purchase of KRW 20,000 imply 15 purchases arithmetically. This is an illustration for budget planning, not a forecast of actual results. Dividing that budget among five ad sets does not mean each will generate enough purchases. You also do not yet know whether the assumed cost per purchase is achievable. That is why it makes sense to check fragmentation and measurement gaps before increasing the budget.
You could also test a more frequent action, such as adding to cart, as an interim optimization target. That is a decision to optimize for something different from a purchase. If you do so, do not stop at “continue if cost per add to cart falls.” Track actual purchases alongside it and decide when you will reassess on a purchase basis. Many carts do not, by themselves, fulfill a sales objective.
Compare the cost of the same action when assessing results
Placing the cost per click of a traffic campaign beside that of a sales campaign can obscure the decision. If purchases are the goal, compare cost per purchase using the same counting criteria. The numbers below are a fictional example to explain the principle.
| Metric | Ad A | Ad B |
|---|---|---|
| Spend | KRW 100,000 | KRW 100,000 |
| Link clicks | 1,000 | 250 |
| Purchases counted on the same basis | 2 | 5 |
| Cost per click | KRW 100 | KRW 400 |
| Cost per purchase | KRW 50,000 | KRW 20,000 |
Ad A has cheaper clicks, but in this example Ad B has a lower cost per purchase. The table does not prove that a particular objective is superior. It illustrates which calculation to prioritize when purchases are the goal.
For real campaigns, align the period, products, offer, and attribution criteria, and examine Ads Manager figures alongside order records. If customers tend to click today and buy several days later, do not decide based only on the first day's results. For products with many returns or cancellations, purchase value in Ads Manager is also insufficient to establish the profit retained.
If clicks are plentiful but purchases are absent, check the destination page, purchase-event collection, and the product's price and conditions before immediately changing the objective. First separate missing platform reporting from a genuine absence of orders; the next action will then be clearer.
Write four lines before building your first campaign
Before opening the settings screen, write down “the action to increase,” “where that action happens,” “the data to check,” and “the budget and review timing for this attempt.” For sales on your own store, for example, this becomes “completed purchase / our store / purchase events and order records / review within the agreed budget, allowing for purchase delay.” For promotion of a public guide, the first line changes to “visits to the guide page.”
If the action you want is a purchase and measurement and checkout are ready, consider the Sales objective first. If this campaign's role is to generate visits, Traffic is a natural fit. If purchases cannot yet be measured, first establish what data you can connect. Choosing the first settings around the action you need to assess is clearer than choosing a sequence of campaigns simply because you are new to advertising.