Facebook Ads: 9 Factors That Impact Cost Per Conversion

One of the primary reasons marketers run Facebook ads is for conversions. They are trying to sell a product or build an email list.

Facebook ads, of course, aren’t magical. Many advertisers are disappointed when they run campaigns that don’t yield the expected results. They assume it’s easy. They assume that once they start the campaign, the money will roll in.

While not magical, Facebook ads are powerful. They are an amazing tool to help brands reach their ideal audience with a message. Facebook doesn’t force users to convert.

Many advertisers ask me, “Why didn’t this campaign lead to conversions?” My answer is typically a shrug. There is a long list of reasons why that campaign didn’t work out.

I wrote this post to help answer such a question. Here are nine major factors that need to be understood and split-tested for optimal results.

And by “results,” note that I’m focusing on Cost Per Conversion. That is the one KPI that matters in this case — not CPC or CTR, for example. Your focus is on the cost per desired action.

1. Product

Let’s face it, the product needs to be first. You could create the perfect campaign. It could have eye-catching imagery, convincing copy and a slick landing page. But none of that matters if the product sucks.

The perfect campaign can’t fix a flawed product. This is the most often missed explanation for disappointing results. Marketers often refuse to consider that maybe — just maybe — their product isn’t all that interesting.

2. Audience

The product may be most important, but audience is close behind. You can run the perfect campaign for the perfect product, but it won’t convert unless it reaches the right people.

Understand that there is likely a large audience of people at various stages of willingness to convert. You can’t hit them all with the hard sell.

See your potential audience like this:

Those Who Don’t Know You But Have Relevant Interests

When targeting these people, most will ignore your ad. They don’t know enough about you or your product yet to trust you. Some will have interest. Of those who have interest, a small percentage will be willing to buy.

You need to understand that those who have interest but aren’t ready to buy have value. Reach them with the right message.

Those Who Know You But Haven’t Bought From You

There is an important group of people who have visited your website, like your Facebook page or are on your email list. They already know enough about you to form an opinion — hopefully positive. While you have a reputation, these people haven’t yet used your product.

As a result, a larger percentage of this group will have interest. A larger percentage will be willing to buy.

Still, you need to understand that many simply aren’t ready yet. They need to be nurtured.

Those Who Have Bought From You Before

The smallest, though most valuable, group of people. They know who you are. They’ve bought from you. As a result, they are likely to have a strong opinion about your brand and your product.

A very large percentage of this group — if they’ve had a positive experience with your product — is likely to have some interest. Message them with the understanding that they know your product. Leverage that.

It’s important to message these groups differently — they are not created equal.

3. Copy

Sometimes long copy works. Sometimes short copy works. There is no universal right and wrong here, but know that it impacts the performance of your campaign.

Your audience should also help determine your copy. Should it be conversational or professional? Slang or no slang?

If your ad copy includes bad grammar and misspellings, don’t expect it to perform. Of course, your audience matters here, but this will almost always result in a bad impression.

Little things like ALL CAPS or Capitalizing Each Word Of Your Copy can impact performance. I’d say using either technique will typically result in negative performance, but that’s not a guarantee.

And finally, your copy should work to convince the user to perform the action you want them to perform. That could be a hard sell or a soft sell. It could mean using calls to action. But ultimately your job is to entice the user to click and convert.

When it comes to copy, you need to split test to find what works best for you and your audience.

4. Imagery

There are some things that are easy. Use my dimensions infographic to make sure that the image sizes you use are appropriate. Link thumbnails should use a 1.91:1 aspect ratio to stretch across the width of news feed.

Quality will be important. Is it fuzzy? Is it easy to read? Does it make any sense? Does it grab the user’s attention?

Beyond that, there are many questions to ask regarding imagery. Should you feature one face or multiple faces? Should you feature your product or logo? Should you use a professional photo, candid or stock imagery (usually no stock imagery!)? Should you use bright and bold colors or something softer or darker?

I have my personal preferences, but these are things that must be split tested.

5. Post Type

If you run a campaign with the objective of a conversion, I will almost always recommend that your ad be a link share. Any click on the thumbnail or description will lead to your landing page, which is where you want them.

Of course, some marketers will swear by the photo share post type with a link in the description. While some often follow the wrong metrics when boasting success (focusing on engagement), this isn’t always the case. Particularly for very visual products (like fashion), photo shares should be tested.

And you also should consider the video share. Particularly given the availability of the call-to-action button, videos can also be very successful.

6. Landing Page

You’ve done everything perfectly so far. Great product, great ad. But if your landing page sucks, forget about it!

Your landing page could be the problem for any number of reasons…

It doesn’t follow through on your ad’s promise
The copy is too long or short
It sells too hard or not enough
It is confusing
It doesn’t look professional
It doesn’t have a clear call-to-action
It isn’t optimized for mobile devices
It asks for too much information
A conversion is too difficult and requires too many steps
The design and layout aren’t optimal

Advertisers who create nearly perfect campaigns will often blame Facebook for poor results when the problem can often be traced to the landing page.

7. Social Proof

If other people — particularly those you know — like or endorse something, you’re more likely to buy it. On the flip side, you’re less likely to buy if that product is receiving a negative response.

Social proof comes in many forms, but one that is visible within your ad is the comments.

No Comments: Neutral, but lots of unanswered questions.

Positive Comments: Could lead to more conversions.

Negative Comments: Could lead to fewer conversions.

The comments are easy for advertisers to miss because they are difficult to manage. But running ad sets to the same ad (published or unpublished post) will help consolidate comments. This not only makes moderation of those comments easier, but can improve perception if your ad gets positive feedback.

8. Optimization

If you are running a campaign for conversions, I’ll almost always recommend that you optimize for conversions. But that is a starting point, and as is the case with everything on Facebook it isn’t a universal rule.

When you optimize for an action, Facebook needs a decent sample size to work with. This makes sense as they need to see as many people as possible who have converted so that they can find others like them.

As a rule of thumb, I’ve often heard that it’s best to have at least 50 conversions per day in order to get an acceptable sample size and optimize for conversions. Again, it’s not a universal rule — I’ve had success optimizing for conversions without hitting this number. But something to consider.

You can also choose to optimize for website clicks or engagement. In either case, Facebook will show your ad to people most likely to either click your link or click anywhere on the ad. That may not be your desired action (the conversion is), but just getting these people to your landing page can help lead to a conversion.

Note that optimization doesn’t stop at the action. You can also choose to optimize for impressions (CPM) or limit impressions to once per day per user (Daily Unique Reach). These two should be used almost exclusively for highly relevant audiences.

Optimizing for conversions will typically lead to the most conversions — since Facebook is showing your ad to people most likely to convert. But feel free to split test with other optimization approaches as well.

9. Placement

Many advertisers are scared of the sidebar. They assume it doesn’t work based on low CTR. Maybe the sidebar doesn’t work for you or your ad. But just make sure you’re looking at the right numbers when making this determination.

The truth is that this is fluid. Placement performance is influenced by many factors, but perhaps most important is competition. As competition goes up, the cost to reach people in that placement increases. As competition goes down, so does the cost.

During the past couple of years, universal costs have shifted repeatedly. It wasn’t long ago that mobile — while getting a high engagement rate — was the most expensive placement. It cost more to reach people there due to lower distribution and higher competition.

But many advertisers have now seen this switch. As users move away from desktop and to mobile devices, the inventory on desktop is decreasing while the competition is increasing. Meanwhile, more users on mobile combined with the addition of Audience Network results in more inventory and less competition.

The sidebar has seen major structural changes. Facebook previously showed up to seven sidebar ads at a time. This increased inventory while lowering cost and engagement. The number of sidebar ads was then cut to two, increasing engagement and costs while decreasing inventory. And now that number has been upped to three.

Don’t ever assume that one will always work or the other won’t work. Keep experimenting!