There's no one-size-fits-all approach to advertising for your business. A car ad for busy moms might look totally different than an ad for the same car being marketed to people interested in the outdoors. Consider adjusting your message for different audiences, even if you're selling them the same product.

Jasper's Market wants to grow both new customers and their grocery delivery service. They also want to persuade previous customers to return. While their overall goal is the same - get people to make a purchase - they want to create two different campaigns for these audiences.

Example 1: Create an audience to build awareness of your business

Goal: Build awareness with people who've never shopped with Jasper's Market Delivery Service. Audience: Reach people in the service area, age 25+, men and women. Possible objectives: Local Awareness, Clicks to Website. What's important to this audience: Your audience values fresh, organic and locally sourced produce. Jasper's Market is now delivering in your area.

Example 2: Create an audience to reach people who've spent with you before

Goal: Drive repeat purchases from previous customers of Jasper's Market Delivery Service. Audience: Custom audience based on loyalty program or website conversions. Possible objectives: Website Conversions. What's important to this audience: You're a great customer; promote specific products.

To reach their different audiences, Jasper's Market creates two different ad campaigns. For new customers, they create an ad that shows what current customers care about: local, organic produce. They also educate potential new customers about their new delivery service. For existing customers, Jasper's Market knows new products from local farms resonate well and keep those customers coming back.

Example 1: Create an ad for people unfamiliar with your business

Advertising strategy: Create an ad that drives awareness of your store's values. Your current customers value local and organic produce and also value convenience. Use what you know about them to drive new customers to your store.

Quick Tips:

• Choose creative that showcases your message. Will people understand what you're trying to say even if they don't read the text?
• Add a call-to-action, like Learn More, to get people to check out your website and learn more about your new, local delivery service.

Example 2: Create an ad to attract people who've purchased with you before

Advertising strategy: Create an ad that brings back repeat customers (people who've already made at least one purchase). Your customers value local, fresh foods from farmers and local merchants. Not only share new products with them, but share stories about where the food comes from. Create a carousel ad that shows new products and link off to each product story.

Quick Tips:

• Choose vibrant, compelling images that showcase the value of your store's products.
• Add a call-to-action button, like Shop Now, to get more people to click through to your website and take the action you care about.
• Offer existing customers a deal to persuade them to take action. For example, Jasper's Market offers new shoppers 20% off when they order online.

Create and run an ad paying close attention to the audience you define and the creative you use. Once your ad has run for a while, here are a few things you can do to measure success:

1) Go to your Ads Manager and customize your columns to view things like age, gender, region and more of the people who viewed your content. Did you reach the people you wanted to?

2) Measure your results in Ads Manager. Here are some key metrics you should view in order to see whether your ad was successful or not:

• Reach
• Frequency
• Actions
• Relevance Score

When used properly, ad sets optimized for conversions are one of the most powerful tools we offer to get you the results you care about. However, they can be challenging and often require patience to figure out.

What follows is an in-depth explanation of best practices and common points of confusion. With this information, you'll be well-equipped to set up successful conversion optimization campaigns. This article assumes you've already implemented a pixel on your website and are able to create a website conversion campaign.

Important: Remember that just selecting a "Conversion" campaign objective doesn't mean your ad set's delivery will be optimized for conversions. You have to choose the conversion you care about in the dropdown of the "Conversions" section of ad set creation, and then select Conversions in the dropdown of the "Optimization for Ad Delivery" section.

Understand The Importance Of Data

If you take nothing else away from this article, remember that the most important thing you can do to set yourself up for success is provide our system with as much data for each of your ad sets as possible. All our other advice stems from that, so focus on getting it right first.

What kind of data?

Conversions. At a minimum, your ad set needs to be getting 15-25 of the conversions it's optimized for per week to have a chance at success.

If your ad set can't get at least that many conversions, optimize for a more commonly-occurring conversion instead. For example, if the Purchase event doesn't happen 15-25 times per week for your ad set, try optimizing for AddToCart or another, higher-funnel action that does.

If none of the conversions you care about happen that often, try optimizing your ad set for something besides conversions:

• If you're highly confident in the quality of your target audience, try optimizing for impressions (if you don't care if a given person in your audience sees your ad multiple times a day) or reach (if you only want to reach a given person in your audience with one ad per day)
• If you're targeting a broad audience, try optimizing for link clicks

Neither of these strategies are ideal, but could still lead to conversions. If even they don't work for you, website conversion campaigns may not be a good option for you right now.

Why do you need it?

For conversion-optimized ad sets, our system has to figure out who is most likely to convert, then deliver your ads to those people. Conversions are rarer than results you may have optimized for with other ad sets like impressions or clicks. Because of this, getting enough conversion data is crucial, and the more we have, the better. Our system can't optimize without being able to learn from a statistically-significant number of conversions. This is due to the laws of uncertainty and statistical learning, not our conversion optimization system in particular. We're always working to improve our system, but there's nothing we can do to get around the need for this data.

Understand What You're Bidding On

Before you set a bid for an ad set optimized for conversions, make sure you understand what it is you're actually bidding on.

OK, what am I actually bidding on?

You're setting a bid not just for a conversion, but a conversion that happens within the window you choose. There are various conversion windows available, but each one includes a period of time and an action (ex: a 7-day post-click window).

Even if you primarily care about conversions that happen outside one of the windows we offer, you should not bid based on that. If, for example, you value a conversion that happens outside our conversion windows at $2 and one that happens within the one you selected at $5, you should bid $5, even if you ultimately care more about the former than the latter.

Got it. So what should I bid?

Whether you're using average-cost or maximum-cost bidding, we recommend using what a conversion within your conversion window is actually worth to you to guide your bid setting. We sometimes call this bidding your "true value." We can't tell you how much that is, but keep in mind lifetime value when determining it. If you're running a prospecting conversion optimization campaign, for example, a new customer can lead to repeat business. You should factor this into your bid amount as well. Also remember that what you bid is different from you actually get charged. In fact, you may end up paying less than your bid in some cases.

I still don't know what to bid

We'd recommend using automatic bidding, in this case. Automatic bidding sets your bid at a level that strives to maximize conversions while spending your budget evenly throughout the duration your campaign.

Choose Your Targeting Strategy

If there's only one piece of targeting information you take away from this article, make it this: Only separate a target audience into its own ad set if you'd bid differently to reach the people in that audience compared to those in your other audiences. Otherwise, combine your audiences into one ad set with a larger budget.

If you keep this in mind, you can target broadly or specifically with equal success.

Targeting Broadly

For many advertisers, this is the best way to approach conversion optimization. Create an ad set with a large audience, and let us find the people in it most likely to convert.

Advertisers often want to split their campaigns into many very specific ad sets. Here are 2 examples that show why that doesn't usually lead to success:

• Each ad set accumulates data. As stated above, we need a certain amount of data for an ad set (not a campaign) to be able to deliver it optimally. If you have 10 ad sets that get 5 conversions each in a week, you have 50 conversions, but also 10 ad sets without adequate data that may start under-delivering. If, instead, you have 1 ad set with 50 conversions, you have the same amount of conversions as the other set up, but also an ad set with enough data to deliver optimally.

• Say that it takes $10 to gather enough data to start delivering an ad set optimally. If you have 10 conversion-optimized ad sets with $10 budgets, then, in your best-case scenario, you have $100 spent on learning and $0 spent on optimal delivery. If you have 1 conversion-optimized ad set with a $100 budget, $10 is spent on learning, and $90 is spent on optimal delivery.

In summary: The broad targeting approach trusts our system to find the right people from a large pool. We can find the right people as long as our system can gather enough conversion data.

Targeting Specifically

Targeting specifically essentially means that you're providing our system with more of the data it needs up front, rather than allowing it to figure who is best to target on its own. The best tools to target specifically while optimizing for conversions are Lookalike Audiences sourced from your best customers.

Here are some tips for using Lookalike Audiences for conversion-optimized ad sets:

• Avoid broad Lookalike Audience sources. Sources like all app installers or all purchasers are not ideal for conversion-optimized ad sets. We often suggest avoiding these because they don't distinguish between great customers and average customers. (Our recommendation to only split ad sets when you'd bid differently for each audience should be a part of your thinking here.) Everyone is just grouped together. Use any proprietary data you have to help you make the distinction. If you don't have much data of your own, try dividing customers by pixel fires. Create a Lookalike Audience from a source made up of the customers who've taken the action you care about.
• Audience quality is more important than audience size. Generally speaking, a Lookalike Audience source of between 1,000 and 50,000 people is ideal. However, having a high quality source (meaning a source made up of your best customers) is more important than any specific number. For example, if the source is good, it doesn't matter if it's 20,000 people or 30,000. You may even be able to succeed with 100 people.
• Expand on a good source rather than using multiple similar ones. If, for example, you have a good source and have created a Lookalike Audience of the people who are most similar to it (say, the top 1%), we recommend expanding on that source (say, by incorporating the top 2% and 3%) rather than trying to create another similar source and targeting the people most similar to it.

  Learn more about advanced Lookalike Audience options.
• Avoid using additional targeting on top of a Lookalike Audience. Lookalike Audiences already incorporate information like age, gender and interests from your source. Layering more parameters on top of that is likely to muddle, rather than refine, your targeting.
• Use expanded targeting as a backup. Make sure the "Expand interests when it may increase conversions at a lower cost per conversion" box is checked during ad set creation. It allows us to expand your targeting if we can't get enough conversions from your specifically-defined target audience. It's a way of automating a switch from specific to broader targeting if the former is unsuccessful.

Read the stories of businesses that have succeeded with Lookalike Audiences.

Frequently Asked Questions

Below, you'll find answers to common questions advertisers have when running conversion-optimized campaigns.

Introduction

Ads that run on Facebook, Instagram and Audience Network are subject to a review process that looks at the amount of text used in your ad. Based on this review, ads with higher amounts of text will receive less or no delivery at all (unless they qualify for an exception). People are more likely to engage with ads that have less text.

Click here to use our tool to determine if your ad's reach may be reduced for having too much text on the ad image before you run your ad. While the accuracy of the tool may not be perfect, you should use the tool as a guide to help you run ads that more likely to be approved.

• Carousel ads - We review each image within the ad to see if they all comply with our text guidelines. Keep in mind that even if one image in the carousel has too much text, it'll affect your entire ad.
• Video ads - We review the thumbnail to make sure it complies with our text guidelines.

Tips

• Click here to check your images, before you try running your ad.
• If you need to include text in your image, try reducing the font size of your text. Keep in mind that if your text is too small, it may be difficult to read.
• Make sure most of the text you use is in the text box instead of directly on the ad's image.
• Avoid spreading text all over the image.

Examples

Facebook ads that contain images with little to no text tend to cost less and have better delivery than ads with image text.

Image text can fall into one of 4 categories:

More visual examples

Preferred Text

These ads are considered ideal because all the copy used for these ads is in the text box, instead of directly in the image.

Low Text

These ads are considered to have moderate text. Each of these ads has most of their copy in the text box, but there's still some copy directly on the images in each of them.

Medium Text

These ads are considered to have heavy text. Each of these ads has about half of their copy in the text box and about half of their copy on the images used in their ads. The delivery for these ads will likely be severely limited.

High Text

These ads are considered to have too much text. These ads will likely not deliver.

Exceptions

We understand that sometimes, you'll need to use an image of a text-based product for your ad (calligraphy, infographics, movie posters, etc.). For this reason, your image may qualify for an exception.

Legal text, like "terms and conditions" also count as an exception (as it applies to the product in your ad).

Examples of text that won’t limit delivery

• Book Covers

  
• Product Image

  

  Note: For product images, we don't allow close up or zoomed in images of logos. Your image must show the entire product.

Games

Entertainment

Examples of text that will likely limit delivery

• Text-based logos - Any logo that is primarily text is counted as text regardless of its size or alignment. Grapic-based logos are exceptions.
• Watermark - Watermarks are considered as text, even if they're mandatory or as per their brand guidelines
• Numbers - All numbers are considered as text

Learn how to troubleshoot other factors that impact under-delivery.

More Resources:

• What does it mean when my ad is pending review?
• My ad wasn't approved, but I think it was a mistake. What's next?

Your advertising objective is what you want people to do when they see your ads. For example, if you want to show your website to people interested in your business, you can create ads that encourage people to visit your website.

When you create an ad, you first choose your objective. The objective you choose aligns with your overall business goals:

• Awareness: Objectives that generate interest in your product or service.
• Consideration: Objectives that get people to start thinking about your business and look for more information about it.
• Conversions: Objectives that encourage people interested in your business to purchase or use your product or service.

An Engagement Custom Audience is a Custom Audience made up of people who have engaged with your Facebook content.

"Engagement" refers to actions like spending time viewing your videos or opening your lead form or Canvas. Using Engagement Custom Audiences, you can retarget ads to people who've taken these actions. You can also use it as a source for a Lookalike Audience, which will let you find people who are similar to those who've engaged with your Facebook content.

Here are the engagement types available, broken down Engagement Custom Audience type:

• Video: 3-second video views, 10-second video views, 30-second video views, video watches at 25%, video watches at 50%, video watches at 75%, video watches at 95%, video watches at 100%
• Lead ads: opened form, opened form but didn't submit (otherwise known as a "drop-off"), opened and submitted form
• Canvas ads: opened, opened and clicked on link

Keep in mind:

• Engagement Custom Audiences are different than Custom Audiences from your website. While both do take actions into account, Engagement Custom Audiences use actions taken on Facebook, whereas Custom Audiences from your website use actions taken on your website tracked by a pixel (otherwise known as "events").
• When you create an Engagement Custom Audience, you tell us how many days you want us to go back when collecting engagement. This means that if you tell us to look back 30 days and someone has engaged 29 days ago, that person will be in your audience. However, if they fail to engage in the next day, they will then be removed from it. Anyone new who engages within the time period you choose will be added to the audience. This means that the audience is constantly being refreshed, so you don't need to edit or create a new Engagement Custom Audience unless you want to change the time period or the type of engagement.

Learn how to create Engagement Custom Audiences from video views, lead ad forms, Canvases and Page engagement.