Lessons from £250 million in FB ad spend

Last year I invested £7k to get coached by one of the best Facebook marketers in America.

He's spent £250 million (yes million!) on Facebook. Working for billion dollar brands. He knows his shit. 

Why did I make this investment?

Truth is, I've always been good at Facebook advertising. I've spent hundreds of thousands and got decent results. Yet, I knew that if I wanted to take Merwave to the next level (7 figure brand) I had to level up my game. And that meant getting someone who's been there and done it to show me what to do. 

Below I'm going to share some of my takeaways...

What Mistakes Was I Making?

When I launched in April 2021 my ads were crushing it. My biggest issue was not having enough stock. 

Then as soon as we invested tens of thousands into stock, my ads started to nose dive. Brilliant. Switching up the creatives and copy didn't really help things. 

I had multiple campaigns running. Using lookalikes and interests. Was moving budgets around daily. Using ABO. No testing strategy. Pretty much chucking everything at a wall and hoping one of them stuck. 

Frustrated and worn out, I dropped £7k on my high level Facebook consultant.

Within our first session I quickly learnt everything I'd been doing was actually harming my performance. Facebook is a machine learning platform. By having multiple moving parts you are preventing its algorithm from doing what it does best - learning. 

So we simplified everything.

One Campaign To Rule Them All

If you really want to generate the greatest results from Facebook advertising then you have to keep things super simple.

Which is why my account now looks like this...

One campaign. 3-4 ad sets. 3-5 ads per ad set. 

The campaign is the manager. The ad set is the salesman. The ads are the sales pitches.

The Campaign

The more campaigns we have, the more diversified your ad spend. It restricts our ability to spend our budget efficiently. We're forcing it to spend in areas that on that giving day may be completely useless. 

Which is why I now have one CBO (campaign budget optimisation) campaign. Our entire budget is consolidated into this one campaign. Instead of us guessing where to spend our budget, we're allowing Facebooks algorithm to do it for us.

Optimising machine learning via one campaign leads to greater stability.  

The Ad Sets

Remember, the ad set is the salesman. So each ad set must target a different angle/concept.

So for Merwave, angles could be women who've never used wavy hair products before, or women who don't even know they have wavy hair, or women who don't like their current hair products. 

2-3 ad sets will be testing grounds. All built using Dynamic Creative. I'll throw in 3 creatives, 2 headlines and 2 body copy and let Facebook find the winner. 

Once I find a winner I'll take that out and move it into a new ad set called the 'Proving Ground'. This ad set will contain only the best performing ads from our testing ad sets. The purpose of this ad set is to prove that the ads within it are scalable. If they eat up all of our daily budget and generate stable results, we know it has longevity. We can then create a completely new campaign, and move it into there. 

I will then have a re-targeting ad set. This will contain testimonials, unboxing, us vs them style ads. I'm targeting website visitors and FB/IG engagers here. 

Because you're using CBO, Facebook will now allocate your budget to the ad sets that it believes will deliver the best business objective (purchases) on that given day. 

So let's say you have a £500 daily budget...

On Monday most of your budget is getting spent on two of your testing ground ad sets which generate good results. 

Yet on Tuesday, Facebook believes it can get better results by spending more budget on your re-targeting ad set, so budget shifts over there.

On Wednesday, you've found a solid ad within your testing ad set so move it to the 'Proving Ground' ad set. Facebook LOVES this ad.

So on Thursday and Friday it spends most of your budget on the Proving Ground ad set because it can get you the most stable results. 

You cannot do this with ABO campaigns. They restrict Facebook from distributing your budget like this. Which is why we always go with CBO. 

The Ads

Big revelation...the targeting is done at the ad level.

Facebook will scan your ad then position it in front of people who are most likely going to take our desired action. 

Remember, your ads are sales pitches. A pitch for one customer won't work on another. 

Which is why each ad is a unique sales pitch. No two ads are the same. We have a wide blend ranging from demos, testimonials, before and after images, videos, GIFS, images. Every angle and style is covered. 

Give the machine (Facebook) the right sales pitches and you've got a much greatest chance of hunting down your perfect customers. 

Targeting

Broad. Every single time. It's the lowest cost audience to market to. And allows greatest scale. 

Facebook is smarter than you will ever be. So stop restricting it by only allowing it to target interest groups (which are being phased out). You'll just spend more money. Lookalikes are also pretty useless now due to IOS14. 

So we always go broad. 

Solving The Attribution Problems

 I don't even bother looking at my Facebook dashboard anymore. The data is so far wide of the mark it's useless. 

I decide whether to increase or reduce budgets by making data driven decisions after looking at my EcoSystem ROAS document (will cover this in another newsletter) and Triple Whale (an analytics platform). 

I then expand and contract my budgets based on the last 3-4 days data. Some days are incredible, others are shocking. As long as I'm seeing incremental new customer revenue lift across my store, I'm happy. 

Bonus Bits...

These changed my outlook on how to run Facebook ads...

- Stop making decisions based off your Facebook ROAS. It's bullshit. The data does not represent your source of truth (what's happening on your actual store). 

- Less is more. I now spend 2-hours tops inside my Facebook account per week. I make big changes once a week. Budget changes every 3-4 days. Only have one campaign. The more tweaks you make, the worse your performance will be. 

- Creative is king. Targeting is done at the ad level. You need a wide blend of concepts, angles and ads to maximise the success of your campaign. I'm testing 5-10 creatives per week so we have a constant flow of new ads. 

- Click through rate and cost per click are not metrics to gauge performance on. Spend and CPA are the only metric you need to focus on. 

What's Coming Up?

Over the next few months I'm going to breakdown how we launch new products, why we refuse to sell samples, testing ads in Facebook, building high converting landing pages, how we drive 30% of revenue from email, plus heaps more. 

This won't be theory. It's stuff I've done myself with Merwave. 

Hope you got some value out of this. If you think someone would benefit from it, I'd be grateful if you could forward it onto them. 

Tom Reid

Founder @ Merwave