Hi there,
Thanks for reaching out! It sounds like a really frustrating situation when a campaign that's been a solid performer for over a year just suddenly stops working. I've seen this happen quite a few times, especially with recent changes on Meta's platform. It's almost always down to a combination of the algorithm changing how it works and the audience you're targeting simply running out of steam.
I'm happy to give you some initial thoughts and a bit of guidance on how you might be able to get things moving again. The good news is this is almost certainly fixable with a bit of a structured approach to testing. It's not about your audience being "too small" in absolute terms, but more about how Meta now sees that audience and how exhausted it likely is.
TLDR;
- Your spending issue is likely caused by audience fatigue. The same 33,000 people have seen your ads for a year; the algorithm is now struggling to find anyone new in that group who is likely to convert, so it stops spending to avoid wasting your money.
- The recent shift to Audience+ gives Meta's algorithm more control, but it struggles when constrained to a very narrow, exhausted demographic-only audience. You're effectively tying its hands.
- The solution isn't to just widen the age bracket. You need to build and test completely new audiences based on interests, behaviours, and most importantly, lookalikes of your past customers.
- You also absolutely need to refresh your ad creative. Stale ads contribute to poor performance and can cause the algorithm to deprioritise your campaign in the ad auction, especially during busy times like after Thanksgiving.
- This letter includes an Audience Prioritisation Flowchart to help you decide which audiences to test first and an interactive calculator to understand campaign metrics.
We'll need to look at why your winning formula stopped winning...
It's completely understandable to think that if an audience worked before, it should work forever. Tbh, that's one of the biggest myths in paid ads. An audience is not a static pool of people; it's a dynamic group, and over time, they experience what we call 'audience fatigue'. After seeing your ads for a whole year, the people in your 53-65+ group who were going to buy from you probably already have. The rest have learned to ignore your ads. We saw this exact scenario with a women's apparel brand; their previously high-performing campaigns stalled out. By systematically testing new audiences, particularly lookalikes of their best customers, we helped them achieve a 691% return. The algorithm is smart enough to recognise when an audience is exhausted. It sees the engagement dropping, the conversion potential drying up, and it throttles your spend because its primary goal is to get you results. It's decided that forcing your ad on this tired audience is a bad investment.
The timing you mentioned—right after Thanksgiving—is also a major clue. That period is the start of the most competitive and expensive time of year for advertising (Black Friday, Cyber Monday, Christmas). The ad auction becomes incredibly crowded. When faced with intense competition, Meta's algorithm will always prioritise campaigns it predicts will perform the best. A campaign with a small, fatigued audience and potentially stale creative is going to be pushed to the very back of the queue. The algorithm essentialy concludes it can't win the auction at a price that makes sense for your objective, so it spends pennies, or nothing at all.
I'd say you need a new targeting strategy, not just a bigger audience...
The core of the issue is an over-reliance on a single, narrow demographic. Your past success has given you an incredibly valuable asset: a list of customers who have converted at $4-$8. This data is pure gold. Instead of just telling Facebook "find me people over 53," you can now tell it, "find me more people who look and behave exactly like the ones who have already bought from me." This is where you need to shift your focus.
We need to build out your targeting from scratch, using a structured approach. I usually prioritise audiences based on how close they are to the final conversion action. The further down the funnel, the better they tend to perform. It's time to pause that old demographic-based ad set and start testing new ones.
1. Start with Lookalike Audiences (Your Top Priority)
This is your strongest possible starting point. You have a year's worth of conversion data. You should take your customer list (you can export this from your sales platform) and upload it to Meta to create a Custom Audience. From there, create a 1% Lookalike Audience. This tells Meta to analyse the thousands of data points associated with your existing customers and find the 1% of users in your country who are most similar to them. This is infinitely more powerful than just targeting an age bracket. It targets based on actual buying behaviour, not just a birthday.
2. Then Test Interest-Based Audiences
While the lookalike is building and running, you should also test audiences based on Detailed Targeting. Don't just leave it open. Think about your ideal customer within that 53-65+ age range. What are their specific hobbies, interests, and behaviours?
- -> What magazines or websites do they read? (e.g., Good Housekeeping, The Telegraph)
- -> What brands do they like? (e.g., Marks & Spencer, Waitrose, specific clothing or hobby brands)
- -> What are their hobbies? (e.g., Gardening, cruising, bird watching, historical societies)
- -> Are they engaged shoppers? (A behaviour you can target directly)
Create seperate ad sets for each *theme* of interests. For example, one ad set for gardening-related interests, another for travel-related interests. This lets you clearly identify which pockets of your target market are most responsive, rather than lumping them all together.
3. Don't Forget Retargeting (MoFu/BoFu)
You also need to capture people who've shown interest but haven't bought yet. This means setting up retargeting audiences for website visitors, people who have engaged with your Facebook or Instagram page, or people who have watched a percentage of your video ads. With a small budget, you can group these together into a single "Warm Audience" retargeting ad set. These people already know you, so they often convert at a much higher rate.
You probably should refresh your creative too...
This is just as important as the audience. If people in your old audience were tired of your ads, new people in a new audience will still be seeing that same, year-old creative. An ad's effectiveness decays over time. For one of our clients, a company selling cleaning products, we saw a 190% increase in revenue just by swapping their polished studio photos for more authentic, user-generated style videos showing the product in action. What was compelling a year ago might be completely ignored today. You don't need a massive budget for this, just a few simple changes can make a huge difference.
I'd suggest creating 2-3 new variations of your ad. You can use the same core offer but present it differently:
- -> Different Images: If you used a product shot, try a lifestyle image of someone using the product. If you used a graphic, try a real photo.
- -> Different Headlines: Try leading with a question that addresses a pain point, or a headline that highlights your main benefit. Use a framework like Problem-Agitate-Solve. E.g., "Tired of [Problem]? It can feel like [Agitation]. Our product provides [Solution]."
- -> Try a Simple Video: You could even just use your phone to make a short, authentic video showing the product or talking about its benefits. These often outperform slick, professional ads because they feel more genuine.
The algorithm considers 'ad quality' and 'predicted engagement' when it decides which ads to show. Fresh creative signals that you're actively managing the account and gives the algorithm something new to work with, which can often kickstart the spending.
You'll need a new campaign structure to make this work...
With a $10/day budget, you can't afford to test dozens of things at once. You have to be efficient. I'd recommend a really simple structure to start with:
Campaign: Advantage+ Campaign Budget (formerly CBO) - $10/day
Using a campaign-level budget lets Meta's algorithm decide how to best spend the money across your different test audiences. It will automatically shift more budget to the ad set that's performing best, saving you the guesswork.
Inside this one campaign, start with two or three ad sets:
- -> Ad Set 1: Lookalike 1% (Purchasers). This is your most likely winner.
- -> Ad Set 2: Interest Group A (e.g., all your 'Gardening' related interests).
- -> Ad Set 3: Interest Group B (e.g., all your 'UK Travel' related interests).
Put your 2-3 new ad creatives inside each of these ad sets. Then, let it run. With a small budget, you need to be patient. It could take a week or more to get enough data to see which audience is working. Your goal is to find one or two new winning audiences that can replace your old, fatigued one. Once you find them, you can turn off the losing test ad sets and let the winners run.
This is the main advice I have for you:
| Problem | Recommended Action |
|---|---|
| Audience Fatigue & Over-reliance on Demographics | Pause your existing ad set. Upload your past customer list and create a 1% Lookalike audience. This should be your new top-priority audience to test. |
| Lack of Audience Diversity | Brainstorm and test 2-3 new ad sets using Detailed Targeting (interests, behaviours) relevant to your 53-65+ demographic. Group them by theme. |
| Stale Ad Creative | Create 2-3 new ad variations. Test new images (e.g., lifestyle shots), new headlines (e.g., addressing a pain point), and maybe a simple, phone-shot video. |
| Inefficient Testing Structure | Set up one new campaign using Advantage+ Campaign Budget ($10/day). Place your new test audiences (e.g., 1 Lookalike, 2 Interest groups) as ad sets within it. Let the algorithm optimise the spend. |
Getting a campaign unstuck and finding new pockets of growth involves a fair bit of this kind of structured testing and analysis. It's a continuous process of building new audiences, refreshing creative, and interpreting the data to make the right decisions. It's often where having an expert take a look can really speed things up and avoid wasted ad spend on tests that aren't set up correctly.
If you'd like to go through your account together in more detail, we offer a completely free, no-obligation consultation where we can review your setup and work out a more specific action plan. Sometimes just having a second pair of expert eyes on it can uncover things you might have missed.
Hope this helps!
Regards,
Team @ Lukas Holschuh
