Case Study: This Simple Strategy Improved My Facebook Advertising By 400% In Just 4 Weeks

Let me start by giving you the secret.

A successful ad campaign on Facebook is all about finding the winning ad. That means you’re trying to find the ad that generates the most clicks as well as the most conversions. (We’re defining a conversion as the action you want your prospect to take, whatever that might be.)

So how exactly do you find the winning ad?

Deliberately and methodically, and today I’m going to show you how to do it using one of my own Facebook ad campaigns as an example.

This brief case study has 3 parts:

• A description of what my Facebook ad campaign is designed to do.
• Actual statistics from my campaign so you know I’m not pulling these numbers out of thin air … or other places.
• The steps I took to improve my results by 400% in just 4 weeks.

Let’s dive in …

Part 1: A Description Of My Facebook Campaign

CAMPAIGN OBJECTIVE: To grow my email list of prospective clients. FYI, I’m a custom web developer in the Detroit area.

CAMPAIGN STRATEGY: I am doing the traditional “bribe to subscribe” — offering something useful and valuable in exchange for a prospect’s email address. What I’m offering is a free copy of my book, “The Modern Website Makeover”.

After they become an email subscriber, I then stay in touch with them via periodic emails so they get to know who I am and what I do. The goal is to build trust, demonstrate expertise and stay top-of-mind so I’m the guy they call when they need assistance with their website or digital marketing efforts.

TARGET AUDIENCE:  A Facebook Lookalike Audience based on people that previously visited my website.

AD BUDGET: $17.50/day

Part 2: My Campaign Statistics

Here are the results of the first 4 weeks of my campaign:

WEEK 1:

• New Email Subscribers: 10
• Ad Spend: $114.34
• Cost Per Subscriber: $11.43
• Proof: Subscribers | Ad Spend

WEEK 2:

• New Email Subscribers: 32
• Ad Spend: $122.92
• Cost Per Subscriber: $3.84
• Proof: Subscribers | Ad Spend

WEEK 3:

• New Email Subscribers: 60
• Ad Spend: $122.33
• Cost Per Subscriber: $2.04
• Proof: Subscribers | Ad Spend

WEEK 4:

• New Email Subscribers: 51
• Ad Spend: $122.61
• Cost Per Subscriber: $2.40
• Proof: Subscribers | Ad Spend

From week 1 to week 4 the cost per subscriber dropped from $11.43 to $2.40. That’s a 79% reduction. The number of weekly subscribers went from 10 to 51. That’s a 400% increase.

Part 3: How I Improved My Campaign By 400% In Just 4 Weeks

Now the fun stuff.

The key to finding the winning ad is to treat your Facebook ad campaign like a well-engineered lab experiment.

Specifically, you want to create an “ad matrix” and test a bunch of different ad variations to see which ad performs the best. Kind of like a digital survival of the fittest.

And the best part is Facebook is going to do most of the heavy lifting for us.

Creating The Ad Matrix

The two most important parts of each Facebook ad are the headline and the image. So to create our ad matrix we’ll need numerous headlines and numerous images to generate all the ad variations for our experiment.

For this ad campaign I’m going to test 6 headlines and 3 ad images, which will give us 18 different ad variations.

Here are my 6 headlines to test:

1. How To Redesign Your Website Without Spending A Fortune

2. How To Modernize Your Website Without Spending A Fortune

3. What To Do About A Stale, Outdated Website

4. What To Do About An Ugly Website

5. How To Give Your Website A Facelift

6. How To Redesign Your Website So More Prospects Say Yes

Here are my 3 ad images to test:

Image 1:

Image 2:

Image 3:

When it’s all assembled, here’s what the ad actually looks like:

• NOTE 1: In the ad above, the headline we’re testing is outlined in red, the image we’re testing is outlined in green. These are the only things that will change in each ad. Everything else stays the same in all 18 ad variations.
• NOTE 2: In the interest of not having you scroll a mile (and thus, giving you carpal tunnel syndrome), I’m only showing one of the 18 ad variations. The other 17 ads look just like this one, except for the headline and the image.
• NOTE 3: Although we’re testing 18 ad variations in this case study, I’ve seen Facebook ad campaigns test as many as 40-80 different ad variations to find a winner. You can go to town with this if you really want to. It all just depends on how many variables you want to test and how much money you have to work with.

Time To Fire Up The Facebook Ad Machine

Now we need to load our 18 ad variations into Facebook and begin our experiment to find the winning ad. And the best part, we just sit back and let Facebook do the work.

As soon as you turn your campaign on, Facebook is going to start randomly displaying all of your ad variations to your target audience and tracking the results.

If you set up your campaign correctly, the top performing ad is going to gradually bubble up to the top.

It’s true that some technical know-how is required to set this all up. If you want to do this yourself but you’re new to Facebook advertising, here’s a good resource to get you started. There is a little bit of a learning curve, but it’s totally do-able.

If you want to use Facebook advertising for your business but you’re just not a techie (and have no desire to be), then just hire someone to set up and manage your campaign(s) for you. Feel free to contact my office if you’d like, as this is a service we offer.

Eliminate The Losers, Run With The Winners

During week 1 of my campaign, I let all 18 ad variations run the entire time because I wanted to give the Facebook algorithm time to work. The more data you gather, the more confident you can be in the results. However, even in just the first week you can already start to see that some ads are definitely performing better than others, as shown below:

During the middle of week 2 I felt that I had collected enough data to begin eliminating some of the losing ads. So I paused 9 ads that weren’t doing well and let the 9 remaining ads continue to run, as shown below:

During week 2 I learned that the following 3 headlines were not resonating with my targeting audience:

• What To Do About A Stale, Outdated Website
• What To Do About An Ugly Website
• How To Give Your Website A Facelift

Buh-bye.

In week 3 the data told me I could eliminate a few more ads that weren’t resonating with my audience, as shown below:

During week 3 I learned that the following 2 variables weren’t resonating with my target audience:

• How To Redesign Your Website So More Prospects Say Yes
• Image 2

We’re now closing in on the winner. It’s down to 2 headlines and 2 images:

• How To Redesign Your Website Without Spending A Fortune
• How To Modernize Your Website Without Spending A Fortune
• Image 1
• Image 3

Even though I think I know what the winning ad will be, I’m still going to let the remaining ads run a few more days just to be sure.

By the end of week 4 it’s pretty clear what the winning ad is, as shown below:

The winning combination is:

• How To Redesign Your Website Without Spending A Fortune
• Image 1

Whether or not you guessed this would be the winning combo, you now have data that validates this result.

Over time it’s recommended that you create new ads to challenge your winner to see if you can outperform it. But if you wanted to just stick with only your winning ad for a while that’s perfectly fine because you know it’s been thoroughly tested.

4 Key Takeaways From This Case Study

Don’t prejudge. Even though you think you know which ad will win, trust the system and let the results play out. I’ve had many ads that I “knew” would win — but they bombed — and vice-versa. You just never know until you run the test.

Be patient. The Facebook algorithm is really good and you can get results relatively quickly, but it’s not instantaneous. You need to give it a few days to gather data and self-correct so it can zero in on the winner.

Failure is good. Test as many variables as you can. It’s just as valuable to know what *doesn’t work* as well as what does work.

Repurpose and recycle. Use the results of your ad campaign on your website and throughout all your marketing. Remember, you invested time and money to learn what works so use it wherever you can instead of just guessing.

[Update] Week 5 Results