Facebook Advertising Handbook:

MINDSET

Success Mindset

Patience, realism & results

Typical Costs

It’s normal to spend full offer value during testing. Eg: $5,000
Later target CPA should be 20-40% of your offer value

Long-Term Thinking

Get rich quick, or get rich for sure?
$1 in $2 out “instant cash” funnel is a fallacy
The most successful companies make 80% of their revenue sometime after 90 days into the relationship. This means that most companies who are looking for short-term results are missing out on the lion's share of the available revenue.
Lead nurturing is incredibly important and is where the majority of your results will come from if you execute well.

How Facebook Works

The difference between Facebook and Google (intent)
What people “like” on Facebook become targetable audiences
We get charged on a CPM basis when someone sees our ads
We are participating in an auction with other competitors
The structure of a campaign inside Facebook

Advertising Strategy

We’re going to test different ads & audiences to find what works
We will increase our spending on audiences that work
And eliminate our spending on audiences that don’t work
We will also retarget leads to remind them to book a call
At a later point, we will combine winning audiences to scale
When your account is more mature you can use lookalike audiences
If you already have a big email list you can upload this as a custom audience, to build a lookalike from (geography dependent)
Data vs Emotions. We verify all of our decisions using data and logic, we don’t make quick decisions based on impulses or feelings.

SETUP

Setup Part 1

Introduction to business manager
Create an ad account
Add yourself as an admin
Add a pixel, assign it to your ad account
Add your Facebook page/create one, add page image, info and details
Go to page, click settings, turn profanity filter on
Add a backup administrator
Security center >> Turn on TFA
Add an email address to your business manager
Brand safety >> domains >> verify
Facebook profile verification (optional)
Facebook business verification (optional)
Add your Instagram account (optional)
Add billing info

Setup Part 2

Pixel usage - advertising and analytics
Pixel usage - advanced matching
Pixel usage - Cookie usage
Aggregate event measurement
Setup conversion API
Integrate Interface pixel and conversion API key
Add pixel to Interface outcome pages
Manage events/prioritize events

TESTING

Audience Research

We are initially looking to run ads against 4 bullseye audiences, but we'll brainstorm more than we need because you’ll be able to come back to this list and refer to your research as you run new tests.
Tutorial: Let’s take a look at the targeting settings inside the Ads manager.
Demographic filters
Geographic filters (focus, one country per ad set)
Different types of audience angles (influencers, publications, behaviours, education)
For audience research, add an audience to “detailed targeting” and then click “suggestions”
The audience size should be 800k to 3m
Download the audience research template from Google Sheets to start planning out your audience (https://bit.ly/3ClsrCp)
Let’s conduct audience research now for one of our clients (Nasreen).

Setup Your Tests

The difference between campaigns, ad sets and ads
Choosing a campaign type/conversion objective
Always use the Leads/Conversion objective (not lead forms) always set for “website” as we want to provide a URL to your website where Interface is embedded.
We’re going to create 5 ad sets. 4 will have different audiences applied and one will have no audiences applied because that will be our control group
I’m going to start by creating one ad set which we will duplicate later
First, let’s overview the ad set settings
The next thing I want to mention is that during the testing phase we keep the budget at the ad set level, not the campaign level
We start testing new ad sets/audiences at USD $20/day
Next let’s fill in our targeting settings, budget and other details
Now that we have the first ad set created before I duplicate it for each audience - I’m going to create our two ads first so that saves us some time later.
Let’s configure the ad settings, upload the ad creative and paste it into our ad text
When it comes to ad URLs you have two choices. The first is to link to your website where you have embedded Interface on a simplified page structure (this is the best practice due to domain verification) or you can link directly to Interface if you don’t have an existing website to embed on (temporary solution).
How to embed Interface on a stand-alone landing page on your website
It’s important that we append a UTM template to your destination link so that Facebook passes details of which audience triggered the lead.

UTM Template:

?utm_source=FB&utm_medium={{adset.name}}&utm_campaign={{campaign.name}}&utm_content={{ad.name}}

Ok, next I’m going to duplicate this ad and just swap out the creative as most of the other details will stay the same.
Now that we have our two ads inside an ad set we are ready to create 4 duplicates of the ad set so that we can add each of our chosen audiences and launch the campaign.

Launch Tests

We’re going to run this test with 4 different audiences and one control, so I’m going to take the existing ad set and duplicate it 4 times so that we end up with five.
Let’s go into each duplicate, rename the ad set and apply the new audience to each.
We’re now at the point where we can put our first ad set into the review process with Facebook. As these contain new ads not previously seen by Facebook I only want to send them one instance of the two ads to review, and when they get approved I will put the other instances into review as well. If we accidentally violate any policies, I would prefer us to have 1-2 rejected ads instead of 8 as that would impact the health of our account.
When it comes to staying compliant on Facebook I highly recommend taking Facebook’s Blueprint training: https://bit.ly/3pDMZi3
Once you have the first two ads approved, publish the rest of the ad sets and then do nothing for a week.
When all of your ad sets are live, take a 7-day holiday. It’s important that you give them enough time to perform and optimize. If you’re changing settings, budget, changing images or anything else you will ruin the testing phase.
After a week, let’s apply an exclusion audience to the ad sets of existing leads from the last 180 days. Don’t worry this exclusion won’t reset the learning phase.

Scale & Trim Tests

After 7 days, start making cuts to the ad sets that are delivering either no results or results that are much higher than average cost.
Facebook is missing a lot of data since the iOS 14 changes. For that reason, it’s important to export your leads from Interface and perform a manual analysis to verify that Facebook’s data is correct.
Tutorial: exporting data from Interface and into Google sheets for analysis
Reallocate your budget to the ad sets that are working.
You do this by doubling the spend from $20 to $40 on the ad sets that work
After doubling the budget, wait 7 days and if performance is consistent start scaling the budget of the ad set by 15-20% every 4 days.
After each budget increment, wait and then perform a look back over the previous 4-7 days. If performance is consistent you can continue scaling by 15-20%. If performance drops (CPL or volume) then scale back to the previous budget that it was stable at. Wait another 3-6 days and continue to scale back the budget if performance still hasn’t stabilized (until you find a budget that performs). Occasionally ad sets stop performing completely, in this case, turn the ad set off. Optionally wait 30 days, duplicate the ad set and restart the process from the beginning at $20, or just move on to the next audience.
After you have spent a statistically relevant amount on each ad version (Facebook does not rotate budget evenly) eliminate the ad version with the worst performance which should make a material impact on lowering your cost per lead. Make sure you don’t do this too early. Generally speaking each ad version you need to get around 10-20 conversion events (and then analyze CPL) before you can determine if it’s working or not.
Testing fresh creatives is an important part of being a successful advertiser, however, we never drop new creatives into proven ad sets or campaigns as it’s likely to tank the performance of your core campaigns. We’ll cover how you properly test new creatives in a later part of this training.
When you have scaled an at set to $80-$100 and it has stable and consistent performance in terms of delivering the right volume of leads at an acceptable cost it’s time to consider moving this audience into a scaling campaign (see scaling training).
When you cut an ad set, pick a new audience and launch a new ad set (which you will also test for 7 days).
You always want to be running at least 5 ad sets at any given time, the open control plus 4 audience-based ad sets. That way you are running and starting to scale audiences that are working, trimming ones that don’t and replacing them with new tests so that you are always scaling both vertically (increasing budget) and horizontally. (testing new audiences and ad sets).
Pro tip: sometimes certain audiences fail because they don’t respond well to your ad creative. When you eventually create new ads you can retest old audiences that did not perform well with the new ad creative and often you can be surprised by the results. Note: Don’t do this until you have executed the creative testing strategy first.

SCALING

Scaling audiences

To start scaling vertically we create a new campaign called “scaling” with one ad set which should combine the initial audience winners from the “testing” campaign. For example, you might have horizontally scaled two audiences to $80 per day consistently with proven performance. We then combine these audiences in a “Combo” ad set that we launch with a much higher budget for example $150 to start with.
If you are satisfied with performance after 72 hours you can begin to increase the budget on this scaling ad set by 15-20% every 4-7 days.
Keep performing lookbacks and determine performance before increasing the budget.
Decrease the budget if performance drops until you hit you find a level that works.
If the scaling ad set does not perform at all, turn it off, duplicate and reattempt it again in a few weeks. Facebook is a dynamic marketplace that constantly changes and shifts (it might do well at a future point).
Never edit an existing scaling ad set that is running for the purposes of adding/removing audiences, settings or ad creative as it will reset the learning phase. The only exception to this is when we are adding new (but proven) creatives. Please consult the creative testing training before you do this.
When you have new audience tests graduating from the “testing” campaign it is better to wait until you have several graduates to cluster together in a new scaling ad set. Eventually, you will end up with multiple scaling ad sets inside the scaling campaign.
When you have more than one scaling ad set you should move the budget to the campaign level (known as CBO or Advantage+ campaign budget) to let Facebook’s AI decide how to distribute the budget across the scaling ad sets (never do this inside the testing campaign - budget should always be kept at the ad set level there to ensure even budget splits across ad sets while testing.
Targeting used to be the most important success factor in running Facebook Ads but since iOS 14 the focus has shifted to bigger broader audiences and then using great “ad creative” to bring in your ideal prospects.
Therefore it is important every few months to inspect the frequency in the delivery column against each ad set to see how many times on average people have seen your ad in that audience. When you hit a frequency of 1.5-2.0 it means that on average most people in the audience have been exposed to your ad one or more times. When this happens you hit a threshold of exposure/saturation. This can happen even though your reach is only 100,000 inside an audience size of 1 million. Sometimes Facebook decides to show your ad to a subset of an audience (for unknown reasons) or that audience isn’t fully addressable (not active on FB) so they keep showing your ad to a smaller subset of the overall audience.
As you scale your campaign and increase your budget, your ad will increasingly become exposed to more and more people. Essentially people get tired of the same ads and start to ignore them. That’s why big brands like Coca-Cola or Nike run lots of different campaigns and are constantly cycling through different ad creatives.
So every quarter (at least) be prepared to completely refresh your ad creative.
Don’t change your ad text and image at the same time. Only test one variable at a time when testing new creatives.
Experiment with different ad creative formats. With the Interface program, we have gotten you started with image-based ads. These are great because they instantly convey the value proposition of the ad within seconds, for example, “Discover Your Something - take the quiz”, or “Get this free something to help with XYZ”. Videos are more tricky because someone has to usually unmute you and watch enough of the video to understand what it’s about. Eventually, you should experiment with video ads at the top of the funnel when you feel ready. We will also be talking more about this in the retargeting section. A great idea for a video ad is to take your Aha! And teach it to your audience with a call to action at the end for people to download your lead magnet or take your quiz.

Scaling with lookalikes

When it comes to scaling, lookalike audiences are the holy grail. Instead of using audience targeting, you are instead telling Facebook to find more people that “look like” your existing leads.
In order for this to work correctly, lookalike audiences need a lot of “seed data” against the pixel. For example somewhere between 300-1000 existing leads from one country before you can create a lookalike audience that will actually work.
When you do have this volume of data (check your Interface leads) you are ready to create a lookalike audience.
As the Facebook pixel has likely missed quite a few conversion events, it makes sense to export your leads from Interface and upload these emails to Facebook as a custom audience. You can then combine the emails with the “lead pixel event” to build a better source audience for your lookalike.
When you create the 1% lookalike you can then duplicate one of the ad sets inside your scaling campaign. You will add the lookalike audience as its targeting and remove any detailed targeting (audiences, interests & behaviors). As you gain more data you can also create 2% and 3% lookalike audiences that give you more room for scaling.
Evaluate the success of (and scale) a lookalike campaign in the same way you would normally scale an audience-based campaign.

RETARGETING

Retargeting is a great way to thank people for coming into your world, tell stories, add value and of course, remind them to book a call.
Once you have completed the initial round of audience testing and are starting to scale it’s time to set up your retargeting campaign.
You can build retargeting audiences inside the audience manager. Simply create a new custom audience, select the website and then change the event from website visitors to leads. You can then select the timeline of when people should fall in and out of this retargeting audience. For example, if I create an audience with only 1-day retention and then run an ad against this audience - people will only see this for 1 day after becoming a lead, they then leave this audience and don’t see the ad again. This could be very effective if I want to introduce myself and thank them for joining my world.
I could then create another audience and exclude leads from the previous 1 day, but include leads for the next 7 days (so it doesn’t overlap with the welcome video). With this custom audience, I could create another ad that teaches them an interesting concept or technique from your course, with a soft CTA to “message you” on Facebook messenger for more info (this is a specific campaign objective type that opens their messenger app). The goal here is to strike up a conversation and then get them booked on a call.
Finally, I could create another custom audience, which excludes leads from the last 7 days but includes leads from days 8-30. Inside this time window, we could run an ad that talks more about the benefits of getting on a call with you and links directly to your booking page.
Beyond day 30 there are endless opportunities to further subdivide the time range out to day 90 and beyond, by showing customer testimonials, more teaching videos and other calls to action that keep you fresh in the audience's mind.
A general rule of thumb is to spend $1 per day for every 1,000 people inside the custom retargeting audience (you can view its size inside the audience manager)
Most of our retargeting audiences run at just a few bucks a day. A very cost-effective way of staying top of mind and leading people to the next step.

TESTING NEW CREATIVES

Creative is one of the power levers you can pull to dramatically increase your results
Always have a separate campaign to run new creative tests. We do this so that new creative tests that flop (of which there will be many) won’t bring down the performance of the core campaign.
Test each new creative inside its own ad set. You can also add 2-5 additional sub-variants of the new creative inside that same ad set. Give it a descriptive name and maybe use an ID number, that way you can reference it to a Google sheet where you track all of your changes and creative tests.
Change only ONE variable per sub-variant so that you can isolate the learnings
Inside this campaign and ad set we are only testing the creative, we don’t use audiences, behaviours, interests or education. We run these with open targeting (the same as our control group). The only targeting is basic age, gender, and country.
When it comes to testing creatives, there are two types of testing. We have “net new” a never seen before image or video and “iterative”, which is fine-tuning existing ads to improve performance. You should always be taking your best creatives and performing

iterative testing on them to improve their performance.

Scale your creative tests (doubling them 15-20% every 4-6 days). Make these new creatives prove themselves at a higher budget, over a longer time (7 days) before duplicating the ads into your scaling ad sets.
Leave the creative ad set on if it continues to receive conversions. (It helps you scale).