Founders Factory logo
Practical Advice

A Startup Guide to Facebook Ads in 2023

Practical Advice

A Startup Guide to Facebook Ads in 2023

Words Louis Butcher

April 26th 2023 / 8 min read

Over the past twelve months, we’ve seen a lot of hysteria around ‘the death of Facebook marketing’, a result of the explosion of TikTok, iOS14 tracking woes, and investors who are becoming uncomfortable with rising customer acquisition costs (CACs) through paid social advertising.

Personally, I don’t buy into the ‘death’ of Facebook marketing, or any paid direct marketing channels for that matter. It’s true, yes, that iOS14 and broader cookie tracking changes have made these channels (especially Facebook) more difficult to measure and optimise. But we’d expect these changes from any tech platform, and the reality is that, if the channel is changing, you must adapt to continue succeeding.

Tactics that worked a few years ago won’t work now. But don’t panic, because I’m going to break down the optimum startup approach for running Facebook advertising that will leave you best placed for success.

Glossary

CAC - customer acquisition cost (how much it costs to acquire a new customer)

CTR - click-through rate (clicks received divided by number of impressions)

LAL - lookalike audience (audience consisting of people similar to those in your Custom Audience)

LTV - lifetime value (customer value x average customer lifespan)

ROAS - return on ad spend (revenue attributed to your ad campaign divided by cost of the campaign)

UGC - user generated content

Interested in receiving more insights like this ? Subscribe to our newsletter and join 17k founders, investors & innovators.

Subscribe here

Account structure

As with most things, you need to get the foundations in place first before you can build. In this case, that means getting your account structure right. The good thing here is, a much more simplified account structure is now the way to go. Here is what you need:

Campaign 1: Creative Testing

This campaign is where you’ll spend most of your time. This is for testing new ads only, figuring out which ones work and are therefore worth scaling and iterating on. Within this campaign you should have an Ad set for each new ad test and set budget at the Ad set level. This is important: if you set a budget at campaign level, your ads won’t get equal spend and you won’t have statistically significant data to determine which ads are successful.

Each of your ad sets should be set to broad targeting, meaning you only tweak country, age range, and gender. Your ad should do the targeting for you, by having clear messaging throughout. Facebook’s algorithm is smart enough (much smarter than you) to figure out who to target based on the creative and copy in your ad.

For those nervous to use broad targeting only for testing new creative, you can experiment also with some interest targeting. However this will mean you’ll be spreading your budget thinner as you test multiple audiences.

The copy used in this Scooch advert along with the subtitles used in the actual creative make it very clear to Facebook’s algorithm this is a product for dogs who suffer with itching. This is enough for it to target the right types of users on Broad Targeting settings.

Campaign 2: Scaling Campaign

The second campaign you’ll need is a scaling campaign. This is only for winning ads from your creative testing campaign. There are two options for this: the manual way where you select interests you want to target; or the Advantage+ Shopping Campaign, which does everything for you.

The former gives you more control, the latter leaves it to Facebook to figure out where to spend the money. I would encourage testing both and see what works best for your account. 

Facebook, however, is moving to more and more automation so I would encourage you to embrace it and learn how to make the most of it, rather than fighting it.

Manual way:

You’ll want to start with three ad sets: Broad, LAL, Interest. 

  • Broad will just be broad targeting only tweaking country, age range and gender.

  • LAL will be your Lookalike audiences based on past purchasers of your product. Start with a 1% and go up to 3-5% if the audience is too small and struggles to perform.

  • Interest is where you will be selecting the interests related to your product and brand. This can be done by either ‘stacking’ or ‘layering’ your interests.

Stacking is where you select multiple specific interests using OR connectors. This tells Facebook to target anyone that meets any of those interests.

Layering uses AND connectors which can refine your audience further. You select an interest the user MUST match, and then select a bunch of other interests of which they must match at least one of (AND). This is useful if you have a really specific product category e.g. golf clubs. So you could select ‘golf’ as the MUST, and then layer in other AND interests based on the customer persona e.g. fitness, GQ magazine, Esquire, Whole Foods. Be sure to keep an eye on the audience size—if it is rated “red” it is too small and will likely struggle to perform.

Advantage+ Shopping Campaign:

This is a much simpler setup, with pre-configured settings and targeting that leverages Meta’s machine learning models. 

But, there is one important thing you’ll need to do. You need to define who your existing customers are. This is so you can control how much budget (if any) you want to assign to existing customers. It is easy enough, just go to Account Settings and select the audience you want to use e.g. Purchased in the last 180 days or customer email list.

The rest is simple, just create a new campaign, select ‘Sales’, and then choose the Advantage+ option. You can then set a daily budget, the percentage you want spent on existing customers and then you can just add your winning ads in.

Campaign 3: Retargeting

Finally, you may want to add a Retargeting campaign to the mix. If your traffic is low, this might not be necessary, you need at least 1,000 people hitting your website in the last 45 days just to be able to create an audience big enough.

Create two ad sets, one targeting website visitors, the other targeting those that have engaged with your social profiles. Drop in your best performing ads from your creative testing campaign and set the budget at campaign level.

Monitor frequency, it should be between 3-5: if it’s higher, reduce budget, if it is lower, increase budget (only if it's resulting in profitable sales!).

In terms of creative, customer testimonials often work well here but make sure to test as what works well for your prospecting campaign might not work as well in your retargeting campaigns.

Creative

Now you have the foundations in place, you are ready for the fun (and hardest) bit—the ad creative. Your creative is the single most important piece of the puzzle, without good creative, you won’t get Facebook working for you and you’ll waste lots of money and time.

There are five important steps you’ll need to take to create winning creative:

  1. Ideation

  2. Assets

  3. Testing

  4. Evaluation

  5. Scaling

1. Ideation

The ideation phase is where you’ll come up with concepts and ideas for your creative direction. The best place to start is customer reviews. Trawl through your customer reviews and pick out themes as to why people love your product. Use those themes as your areas to develop into creative concepts. It may also help to start with the pain points your product or service is meant to be solving, and the product benefits that address these.

2. Assets

Your ads are only as good as the creative assets you use to produce them. I don’t mean fonts, logos, product imagery, and brand colours. You need much more than that and it all starts with UGC content.

UGC (User Generated Content) is content produced by customers or content creators, and is vital for success on the platform. Static shots and video content shouldn’t look polished and beautifully branded: rather, it should feel authentic and native to the platform. The more an ad looks like an ad, the less likely it will perform. 

You can recruit existing customers or friends of the business to help here but I would advise you to recruit legitimate content creators for this as the quality of the content really makes a difference. Platforms like Insense, Billo and Mini Social can be a good starting point for finding creators that suit your brand.

These Roomix ad creatives feel really native to the Instagram platform, they don’t look and feel like an ad, thanks to using UGC creators and native fonts and captions.

3. Testing

Having a good testing framework is important for ensuring you have a logical cadence to what you are testing. You should be looking to test a variety of concepts (informed by your ideation tactics) and then a variety of different executions of that concept.

e.g. Vegan Beauty Brand

Concept 1: Reducing your impact on the environment

  • Execution 1: Testimonial from a customer explaining how much less their carbon footprint is as a result of using your product

  • Execution 2: Three ways [Brand name] is better for you and the planet

Concept 2: Only using natural ingredients on your skin

  • Execution 1: Explaining what each of the natural ingredients do

  • Execution 2: Show the benefit of product with natural ingredients vs other brands

Just because an advert doesn’t work, that doesn’t mean the concept is wrong. It could just mean the execution is not conveying the message effectively.

4. Evaluation

Knowing how to evaluate your ad performance is critical if you want to be able to iterate successfully and scale your performance. If an advert has spent 3x your target CAC and still hasn’t got a conversion, you can assume it will likely never meet it. This is one way to determine when to turn ads off.

Looking at CAC or ROAS, however, is only the beginning. It will tell you if an ad is or isn’t working, but it won’t tell you why. Metrics to consider for determining why an ad is performing include pre-click and post-click, each of which tell you different information:

Evaluating pre-click data will be the best starting point on understanding where to improve your creative. For example, if an advert has a poor CAC it might be as simple as your hook isn’t strong enough so not enough people are seeing it. Fixing a poor hook can quickly turn a terrible performing ad into a winner. 

These two Scooch ads are nearly identical other than the two hooks, one will likely perform better than the other and then be used on other variations.

If your pre-click metrics are all strong but you are still struggling to reach CAC and ROAS targets it might be less to do with your ads and more to do with the post-click experience (e.g. your landing page needs work).

5. Scaling

So you have found some winning ads in your creative testing campaign, now what? Firstly, if it’s performing, don’t turn it off—keep it exactly where it is.

The goal here is to spend more on that winning ad whilst keeping the CAC/ROAS at similar levels.

You can scale spend two ways:

  1. Add the winning ad to a Scaling Campaign as explained above

  2. Increase budget on it in the testing campaign

When increasing the budget on the existing ad in the creative testing campaign, don’t increase by any more than 10-20% every 2-3 days. Any more than this and you risk resetting the learning period and you’ll likely see fluctuations in performance.

The last step in scaling is creating new iterations of the winning ad to put into your testing campaign in order to find new winners. A few ways you can do this:

  • Add the hook to other ads that aren’t performing as strongly

  • Shoot the same ad with different creators

  • Use the same ad with a different voiceover voice

  • Edit slightly longer and shorter versions of the same ad

  • Test the same ad with different copy variations

  • Test the same ad with different offer/discount types

So, Facebook isn’t dead, it is very much alive and I hope now you now feel more informed on how you can make it work for your business. Success doesn’t come overnight, it requires a well thought out testing structure. But through adopting a process of creative testing and evaluation, you should learn what does and doesn’t work.

Then it is a matter of doing more of what does work and less of what doesn’t.

Read more insights like this on the Founders Factory blog

Read more

About Louis

Louis Butcher is a Growth Lead at Founders Factory, supporting a number of businesses across our portfolio. Prior to this, he held a number of growth and digital marketing roles at companies including WhoCanFixMyCar.com, Plum Guide, Toyota, and Vibrant Vegan.

Words by
Louis Butcher
Share article

News from the Factory Floor

trends

Protecting Our Oceans: How Dual-Use Can Defend Both The Natural World And National Sovereignty

The scale of what we do not know about our oceans and the scale of what we are losing from them are directly related. You cannot protect what you cannot see. However the technology to change that at a meaningful scale, in real time, across the full expanse of the ocean environment is now becoming available.

founder stories

The New Talent Playbook, Part 2: Set the Bar Before You Set the Culture

Culture, in practice, is the bar you hold and what you tolerate when people miss it. You set the bar in the first conversation with hire number one and it’s reinforced (or quietly eroded) every week after that.

trends

A Profile On European Healthtech

What has changed most visibly in the past two years is the appetite from investors, health systems, and policymakers to close the gap between European innovation and real-world deployment at scale.

See more of our news