Good day Redditors.

I have written posts about how we approach advertising to drive success for our clients and our brands.

This post will be about a summary of learnings from spending over $20M in ad spend.

I know it's not that much ad spend, there are people who manage this a month.

To be quite fair, 60% of the ad spending has come from this year.

So, let's start.

I remember the early days of advertising in Meta. Back in 2018, literally anyone who had a somewhat decent product where able to generate sales.

What was even more fun was that It was done with quite a few ads.

Then all of the advertising was about "hacking the ads manager"

Testing interests

Creating lookalikes from followers, website visitors, purchases, email lists, etc.

Back in 2018, we were able to scale a bike company brand from $20k a month to $240k a month in just 10 months.

We found one ad, and we tried to spend as much money as possible with that ad doing all sorts of "hacking ads manager" stuff.

This was the year when we decided to launch sales on Halloween and generated $100k in just a few days.

Our black friday that year didn't even come close.

Fast forward to 2020.

This was still before IOS, I noticed that it was tough to replicate the scaling with just one ad.

In 2020 we started to test new ads every month, I remember this vividly.

Created 3 ads for that month that we wanted to test, and we hit two winners.

One video, one image.

Wow... So that became our benchmark, test 2 new ads every month and still do "hacking the ads manager" trying to push these ads via interests, lookalikes, and duplicating campaigns from abo to CBO, to CBO from abo.

That time literally, our team spent at least 2 hours per day in Facebook ads manager trying to figure out the newest targeting hacks by using stacking.

We played around with interests stacks and lookalikes stacks.

Even combo stacks. We really thought that this is what generates the best results.

Then....

Armageddon happened.

IOS 14 Mother***er.

This update wiped our performance across all accounts like no other.

We thought that this was the extinction of Facebook ads.

Everyone was screaming Facebook ads are dead.

Clients were dropping us left and right.

Then after a month of horrible performance and losing clients, we still had to pay our people.

We understand that "Hacking Facebook ads manager" days have come to an end.

We started to test fewer interests but still kept the lookalikes. :D

This was the time when we started to test more ads - 4 a month.

Yes, we doubled the testing.

It's actually funny to look back, thinking that testing 4 ads a month was actual testing.

We are at the end of 2021.

Testing about 8 new ads every month for our brand and our clients.

Mainly using a stack of interests and a stack of lookalikes.

Now it's January 2022, we signed this one client with the capacity to create ads and we were also building our in-house content department in order to scale creative testing.

We analyzed:

Facebook and Instagram content - what type of content are we seeing daily?

Turns out we were seeing content that we were interested in.

Facebook and Instagram ads - again, most of the ads that were shown to us and that we reacted to were the stuff that we were interested in.

Interesting...

This is still January 2022.

What if we start researching customer psychology and customer behavior and just understand what type of content our target customer likes to see? ( I Have a post on how we approach research, so I won't describe this in detail in this post).

Let's create ads that speak to our target customers by introducing awareness stages

Unaware ads - ads that take an unaware audience to most aware audience.

Problem-aware - ads that talk to people who know about their problem.

Solution-aware ads - ads that target people who understand that they have a problem but don't know which solution to choose.

Product aware - ads that target people who understand their problem and also know how that is the way on how to solve their problem but don't know which brand/product to choose.

Most aware - people who already know everything about their problem, how to solve it, what product to choose, they are just waiting for the right time to buy from you.

So how can we test the content?

One thing is creating the ads another is testing it.

We had this complicated ad account structure, one campaign ABO campaign to test audiences, and one CBO for scaling the best audiences with the best ads.

We aimed to create 20 ads.

There was no way we could test 20 ads and interests on the ABO level, this was like lighting money on fire.

Then we decided to use:

One CBO campaign.

Use best-performing interests and lookalike combo stack and one broad ad set ( no targeting)

To test how the ads messages work.

Why no targeting?

Well, by doing content analysis, we were interested in testing the theory of is the algorithm capable of delivering our ads to the right customer without us doing the targeting side.

We launched 3 video ads that were talking about the customer problem and educating how this product will solve it.

On day one, we are getting sales with interest and the lookalike combo stack is winning.

Day two combo stack still winning by a lot.

On day three, our broad ad set was picking up more and more sales.

Day four, our broad ad set, took 80% of that day's spend, leaving the combo stack in the dust.

Day 5, our combo stack dot us one sale.

Day 6, we turned off our combo stack and started creating new broad ad sets with new ad tests.

It took off.

We immediately did this for all of our accounts.

Same result.

Imagine spending 4 years "hacking the ads manager" and seeing that no targeting is producing better results for niche markets than targeting.

That was a real eye-opener on what Facebook ads algorithm can do.

Since January 2022, we have been spending more and more time researching and developing the ads than the actual Facebook ad account.

We have measured that, on average, we spend up to 10 minutes a day optimizing the accounts, and 2-3 hours are spent on researching and analyzing the data to create a better ad.

One of our first steps is to test 50-100 ads in the first month to understand what messaging hits.

By doing this, we are able to find winning messages really fast.

What previously took 6-12 months now can be done in 1-3 months

Testing a lot of ad messages provides you with a lot of winners.

Then creating at least 10 variations on winning ads, you can come down to 1-3 messages that spend 60-80% of your daily budget, which you can comfortably scale.

Today.

If you say that ad testing does not work. You clearly haven't spent enough time researching and testing multiple ad variations.

The algorithm has become more and more content-oriented.

This is the reason we are not experiencing shit days with meta like most of brands and advertisers do.

Yes, we have a few a month, but it's nowhere near the amount that I read on this Reddit channel.

So, to sum up what it takes to generate good results with Facebook ads.

Daily research 1-3 hours.

Daily ad creation 1-2 hours.

Daily ad optimization of 5-10 minutes.

Ads is the targeting.

Understanding your customers is by far the most important thing in order to create ads that they will react to.

Content is king - that's why it's important to input hours and hours into content creation.

Facebook is not rewarding mediocre input anymore.

Top brands are still scaling.

You can do this too.

Keep in mind this is just about Facebook ads, I'm not talking about what happens after the click, your website, your business numbers, upsells, crossells, email flows.

Sometimes I like to think about Facebook ads as salespeople.

We know that big businesses have whole sales departments generating sales every minute.

If you only rely on one salesperson, it's going to get burnt our calling the phone 24/7.

The same thought process can be applied to meta.

The more salespeople you have out here calling the phone, and knocking on doors, the more sales you are going to get.

Find the right salespeople.

Thanks for reading.