kickassgrowth logo

Drop us a Line

Find out how your business
can achieve massive growth too.

office@kickassgrowth.com

Facebook Inception – More Precise Facebook (Re)Targeting

The inspiration for this technique comes from the movie Inception. This technique is a part of a deeper marketing philosophy where you use ads in a more complex way. Basically, what you do is build a more concrete audience inside an existing audience you thought you couldn’t segment more. Can you see the connection with the ‘dream within a dream’ movie reference?

For every Facebook marketer, this is reality inside reality, and a possibility to achieve killer ad results. By the time our KickAssGrowth team got around to creating this blog post, this technique had already found its place and become known as Inception. There was no disagreement on it. As movie buffs, we decided to use Christopher Nolan’s movie as the inspiration for the name. Those of you who watched the movie know that we had a good reason for doing that.  

Prepare yourself to go deeper and  think much bigger as we delve into the secrets of the Facebook Inception technique. Learn all the rules of the dream world of Facebook CA (Custom Audiences) and adopt this elegant and thought-provoking solution for an audience-in-an-audience trick.

inception-facebook-technique

What will you get here?

The Facebook Inception Technique is a technique we, the KickAssGrowth team, developed and adopted to deliver great results for our clients. We used this technique in order to reduce CAC and increase CTR. This article will show you how to implement this technique into your FB campaigns, reduce CAC and increase CTR. If you have any questions, feel free to reach out to us.

The lives of many Community Managers and Ad Specialists have been saved since Facebook implemented Pixel Retargeting. It helped them reduce CPC (cost per click) and CPA (cost per acquisition) much quicker. However, people still struggle with structuring Facebook Campaigns that will target the most converting audience. That’s why we want to share the most effective strategy we implemented. Thanks to this strategy, our clients have words of praise for our team: ‘Guys, you’re geniuses!’, ‘I don’t know what you are doing, but we’re very satisfied with the results.’, ‘You guys kick ass!’.

This Facebook Ads Technique is named Facebook Inception (kudos to our awesome Social Media team leader for being the godfather) because there are many campaigns inside campaigns inside campaigns… inside campaigns…

The first thing you’ll need to implement is Facebook Pixel. Standard Facebook Pixel has standard event codes for different AARRR funnel steps:

– View Content
– Search
– Add to cart
– Add to wishlist
– Initiate checkout
– Add payment info
– Make purchase
– Lead
– Complete Registration

Facebook Pixel

*you can check the Facebook Pixel Implementation Guide here
**and you can learn more about the AARRR funnel here

It really depends on what type of a product you have, but standard Facebook Pixel events cover most of the steps one e-commerce site has. If, for example, you’re a social network, then you will probably need just a few events such as View Content and Complete Registration. We suggest you cover most of the events in order to get more detailed tracking. After you prioritize which part of the funnel you want to track, you should paste the Facebook Pixel script into your website (a little bit technical, but it’s very easy).

The next step is to create Custom Audience in Facebook Ads Manager or, like we did, – in AdEspresso (a very popular tool for running Facebook Campaigns and tracking Facebook ads analytics and data).

The Facebook Inception Technique has 4 steps for optimizing your ads to get the most converting audience:

  1. Custom Audience (Re)Targeting
  2. Lookalike Audience based on Interests or Job Titles
  3. Lookalike Audience
  4. Interest Data

1. Custom Audience (re)Targeting (Current Users Vs. Email List)

How to Create Custom Audience (AdEspresso): Go to Tools → Custom Audiences → Create Custom Audience → Data File Custom Audience or Custom Audience From Your Website

Facebook Targeting Options

Facebook Custom Audiences

 

Custom Audience Data

There are two main types of Custom Audiences:

  • Your Current Website Visitors/Users
  • Potential Users from Email Lists

The visitors who didn’t convert in the past ( buy, subscribe or give you an email address ) have the most potential to convert in the future. They have already shown an interest in your product/service and are planning to buy it. So, when you’re targeting these audiences, you can expect very low CPC and CPA.

If you are looking for potential new users and already have an email list, or you have bought one,  you should create your Custom Audience from that email list. Not only will you achieve low CPC and CPA (if you have the right list), but you will also find an audience that converts well.

2. Lookalike Audience based on Interests or Job Titles

How to Create Lookalike Audience (AdEspresso): Go to Tools → Custom Audiences → Create Custom Audience → Lookalike Audience from Conversion Pixel

Facebook Lookalike Audience

Lookalike Audience (LAL) based on Interests or Job Titles (depends on personas) is the second most converting audience, because we target those users who are the most similar to people that we get from Custom Audience. We suggest that you create LAL audiences 1%, 2%, 3%, 4% and 5% (similarity optimized) in order to test the most converting ones. Look at this as some kind of A/B testing for the most converting audience. When you create these audiences, you should specify the audiences by using interests and job titles based on your personas (useful personas template).

3. Lookalike Audience

This is similar to the aforementioned LAL audience based on Interests or Job Titles, just without the interests and job titles. This is useful if you’re targeting a B2C audience for e-commerce shops, and because LAL audience based on Interests or Job Titles is small and narrow. However, for some products, it can be better to target LAL audience without interests, job titles or any specification. You might be skeptical about this, but this audience will convert far better than the one where you just targeted random people based on Interest data.

4. Interest Data

We create campaigns based on the personas we have and their interests, job titles and other detailed specifications Facebook provides. To get the most converting audience from these campaigns, it’s very important to have very detailed information about your personas. It’s best if you can create personas from your existing data, but if you’re just launching your product, then focus on Customer research and Competitors’ audiences. Also, when you target audiences based on the Interest Data only, and assuming you have already tried to target Custom Audiences and LAL Audiences, make sure you exclude Custom and LAL audiences, because there is a possibility you will target the same Facebook users you have already targeted with those campaigns.

Exclude Facebook Custom Audience

Ad Creatives

At KickAssGrowth, we have a starting point for Ad Creatives where we test 3 different images vs. 3 different Ad texts vs. 3 different headlines (3 X 3 X 3= 27 different ad sets).

Usually we A/B test:

– Product Images VS. People Images
– Long VS. Short text
– Direct VS. Indirect Call to Action in Headlines
– Symbols VS. No Symbols in Ad Texts

Because of these A/B tests, we know exactly what types of campaigns convert in which country after a few weeks of testing. For example, we tested a cartoonized image of Joey from Friends in the UK and Sweden. CTR in the United Kingdom was more than 5%, but Sweden was very bad, so we had to turn off the ‘Joey’ ad set in that country.

Optimizing Results

AdEspresso shows the most relevant data for all the campaigns and ads you are running. In order to compare your results, you can use AdEspresso’s standard Ad comparing feature that gives you a quick insight into which metrics are good and which are bad (you can data pause bad metrics directly from that feature). We compare Ads data from ‘All Ads’, where we can compare more and quickly optimize our campaigns for the most converting images, texts, ad sets or placement (depending on what you choose to track when you’re at the final step before starting your campaign). We always make sure we have enough data (enough impressions and clicks) before we pause the campaign and optimize it for the best results. It really depends on the campaign, but it often takes around 5 days until we have some relevant data for optimization.

Quick Facebook Ads Data Tips

  • In AdEspresso, you can split campaign tracking and see the best results by different categories (Interests data, Gender, Placement, Device etc.)
  • Always check your frequency stats. It’s very important to keep the frequency below 5, otherwise your ads will be shown too often and your clicks will be more expensive
  • Test different visuals, texts and headlines in order to find the most converting one. We always create 27 different ads in AdEspresso (3 texts X 3 images X 3 headlines)
  • Track your best periods and create Dayparting if you see huge differences in conversion and click time
  • Compare Ads data – some important differences can surprise you
  • Track and document your changes and learnings from Facebook campaigns

inception-facebook-technique-2

Don’t forget, this blog post does not represent a dream within a dream. This is just an online version of our team’s real-life brainstorm.

What do you think, is Inception a fitting name?
Is it all just a dream? Is the top still spinning for you? 🙂

Please leave us your thoughts in the comment section below.

 

  • August 18, 2016

    Hi, great stuff. Can you please let us know how many impressions and how many clicks are enough to be able to make conclusions which ads are performing better than others and start optimizing? Below is the part of text I am refering to 🙂

    ‘We always make sure we have enough data (enough impressions and clicks) before we pause the campaign and optimize it for the best results. It really depends on the campaign, but it often takes around 5 days until we have some relevant data for optimization.’

    • Aleksandar Tomić

      Leave reply
      August 18, 2016

      Hi Thomas, we are glad that you like our Facebook Inception article. 🙂
      As for conclusions and decisions on when to stop one campaign, our answer is: it’s pretty relative and it could vary. But, after a thousand impressions, you should be able to conclude something about the ad performance (CTR, CPC). So don’t waste money on never-ending waiting and hoping that one ad could become the best one ever in the long run. You just need to decide for yourself what time-lapse is tolerable and when you will stop it if it performs bad.

  • August 19, 2016

    Great article. So each different ad got its own ad set? I thought ad sets were best used for separating geo regions/demographics/interests etc.

Post a Reply to Aleksandar Tomić Cancel Reply