Solved: Advertising Pattern Impact on Facebook Learning Algorithm?

Hi there,

Thanks for getting in touch. I saw your question about your advertising pattern and how it affects the Facebook algorithm, and I'm happy to give you some initial thoughts and guidance based on my experience. It's a really common question, especially for small business owners who are juggling everything themselves, so you're not alone in wondering about this.

First off, props to you for managing the design, production, and marketing all by yourself. That's a massive undertaking. The short answer to your question is yes, your current pattern of running ads for a week and then stopping for a week is almost certainly disturbing the learning algorithm and holding back your results. Let's get into why that is and what you can do about it.

We'll need to look at your advertising pattern...

The core of the issue is how Meta's (Facebook's) advertising algorithm works. You've probably heard of the "learning phase." When you launch a new ad campaign or ad set, it enters this phase. During this time, the system is actively trying to figure out who the best people are to show your ads to. It's testing different pockets of your target audience to see who is most likely to take the action you want them to (in your case, making a purchase).

Think of it like training a new employee. On their first day, you don't expect them to be perfect. You give them tasks, see how they do, provide feedback, and over time they learn how to do their job efficiently. Meta's algorithm is the same. It needs a consistent flow of data (impressions, clicks, sales) to get better. To exit the learning phase and become truly optimised, it typically needs about 50 conversions (sales, for instance) within a 7-day period.

By running your ads for a week and then stopping, you're essentially firing that new employee just as they're starting to get the hang of things. When you switch the campaign back on a week later, the algorithm has lost its momentum. It often has to re-enter the learning phase from scratch, or at least a significant part of it. This means you're constantly paying for the algorithm to learn, but never really reaping the rewards of it being fully trained. This almost always leads to more unstable results, a higher cost per sale, and you'll find it really difficult to scale things up becuase the performance is so unpredictable.

I'd say you should aim for an 'always-on' approach...

So, what's the solution? The best practice is to move to an 'always-on' campaign structure. This doesn't mean you have to spend a fortune every single day. It just means keeping your core campaigns active continuously, so the algorithm never has to go back to square one.

During your production weeks, instead of switching the ads off entirely, you could simply lower the daily budget to a minimum level. Maybe just a few quid a day. This keeps the data flowing and the algorithm ticking over. It's like keeping the engine warm instead of letting it go completely cold. When you're ready to start taking on more orders, you just crank the budget back up, and the algorithm can ramp up performance much more quickly and efficiently because it never stopped learning.

I remember one instance where we worked with a client who had a similar issue. They were in the subscription box eCommerce niche, and after implementing an 'always on' strategy on Meta Ads, they saw a 1000% return on ad spend.

You probably should look beyond just the schedule...

Now, while fixing the stop-start pattern is a big step, there's another elephant in the room. You mentioned you advertise "via the page," which I'm guessing means you're using the "Boost Post" button. While that's a simple way to get started, you're leaving a massive amount of power and control on the table by not using the proper Meta Ads Manager.

Boosting posts is designed for simplicity, but it's very limited. It typically optimises for engagement (likes, comments, shares) or maybe link clicks, not for what you really want, which is sales. In Ads Manager, you can set up a proper "Conversions" campaign and tell Facebook to specifically find people who are most likely to actually buy your products. This requires installing the Meta Pixel on your shop's website, which is a small piece of code that tracks actions like page views, add to carts, and purchases. Without it, you're flying blind and so is the algorithm.

Using Ads Manager opens up a whole new world of targeting, creative testing, and reporting that is simply not available through the Boost button. It lets you build a proper advertising funnel. For an eCommerce store like yours, this is absolutly fundamental. You'd typically have campaigns aimed at different stages:

• Top of Funnel (ToFu): Reaching new people who have never heard of you before, using interest-based or broad targeting.
• Middle of Funnel (MoFu): Re-engaging people who have shown some interest, like visiting your website or watching a video ad.
• Bottom of Funnel (BoFu): Retargeting people who are close to buying, like those who have added a product to their cart but not completed the purchase.

This kind of structure, which you can only build in Ads Manager, is how you turn casual browsers into loyal customers, and its something you're missing out on right now.

You'll need the right targeting...

Once you're in Ads Manager, you can get much more specific with who sees your ads. For a custom garment shop, the possibilities are huge. Instead of just general interests, you could get really granular.

For your "ToFu" (cold) audiences, you could test ad sets targeting people interested in:

• -> Hobbies that often involve custom gear (e.g., local sports teams, gaming clans, bands).
• -> Life events (e.g., people with an upcoming birthday, newly engaged people for hen/stag dos).
• -> Specific design aesthetics (e.g., "graphic design," "streetwear," "minimalist fashion").
• -> Competing or complementary platforms (e.g., people interested in Etsy, Redbubble, or Printful).

The real magic, however, comes from retargeting. With the Meta Pixel installed, you can create "Custom Audiences" of people who have already interacted with your business. For instance, you can create an audience of everyone who visited your website in the last 30 days but didn't buy anything. Then you can run a specific ad to them, maybe showing off your best-selling designs or offering a small discount to entice them back. This is your "BoFu" and its almost always your most profitable audience.

Once you've made some sales, you can then create "Lookalike Audiences." This is where you tell Meta, "find me more people in the UK who look just like my existing customers." The system analyses the thousands of data points of your past buyers and builds a brand new, highly-qualified audience for you to target. It's an incredibly powerful tool for scaling your business. I remember working on a campaign where we helped a client in the eCommerce niche generate 1500 leads at $0.29/lead by using lookalike audiences on Meta Ads.

This is the main advice I have for you:

I know this is a lot to take in, especially when you're already so busy. To make it a bit clearer, I've put together a table with my core recommendations. This is the path I would take to move your advertising from a basic, inefficient pattern to a professional, scalable system.

Making these changes will fundamentally transform your advertising efforts. It's a shift from just "running some ads" to building a proper customer acquisition machine for your business.

I realise this can feel a bit overwhelming. You're an expert in design and production, and it's a full-time job in itself to become an expert in paid advertising. Getting this stuff wrong can mean a lot of wasted time and money. Working with a specialist can help you get this structure built correctly from the start, avoiding costly mistakes and getting you better results, faster. It lets you focus on what you do best – creating amazing custom garments.

If you'd like to go through this in more detail and see how it could apply specifically to your shop, we could book in a free consultation to have a look at your situation together. It's often really helpful just to have a second pair of expert eyes on things.

Regards,
Team @ Lukas Holschuh