13 Unspoken Meta Ads Secrets (Until Now)
For the last six years, I’ve been neck-deep in social media marketing for brands of every size and industry. Early on, I held analyst roles, which meant I spent lots of time examining Meta ad campaign results to identify exactly what worked and what didn’t.
Later, I put that deep knowledge to work directing creative strategy, doing media planning, and managing big ad budgets on Meta.
I’ve got my 10,000 hours, and then some.
I’ve learned that there are several unwritten “rules” of Meta ads: lessons that most performance marketers learn the hard way, just because they’re not written down anywhere, and not often talked about.
Well, now they are.
Here are the things nobody told you about advertising on Meta that every seasoned performance marketer should know.
Tips for Meta Ads Nobody Seems to Talk About
- For Ecommerce, Meta spends more on weekends, but often still not enough
Weekend performance is usually stronger for ecommerce, but Meta doesn’t always scale accordingly. While Meta can now spend up to 75% above your daily budget (up from 25%) when it predicts an uplift, it still frequently underutilizes weekends. Manually increasing budgets on high-performing days, especially Saturdays and Sundays, can drive better efficiency.
2. Broad targeting is your bread and butter, but you should still micro-target
Meta’s algorithms perform best with broad targeting, where signal density drives efficiency. Avoid exclusions (age, gender, location) within Advantage+ Sales Campaigns (ASC) or broad targeting, business-as-usual campaigns. When you add exclusions to these kinds of campaigns, you feed less data to Meta’s AI models, which hampers their ability to learn and tune on your behalf.
That said, microtargeting still has a place. Instead of applying them to your broad campaigns, though, test narrow segments in separate, dedicated campaigns.
3. Interest-based target audiences will eventually stall
Interest stacks can stall for two main reasons. First, they’re inherently limited audiences. At some point, you’ll have reached saturation that pushes your campaign past the point of diminishing returns. Second, interests are more directional than precise - the data at the core of an interest-based campaign doesn’t reflect the exact reality.
While you may find that some interest-based campaigns present higher efficiency for a time, you shouldn’t rely on them for stability. Broad audiences keep delivery cheaper and more consistent, especially in ASC or Campaign Budget Optimization (CBO) structures.
4. Creative drives 80% of performance actually means you need more creative
Diverse creative is essential. Even if video underperforms on direct ROAS, including it improves overall ad set performance. Meta will serve different formats (video, static, carousel) to users most likely to respond. Include a range of creative styles to match your key customer personas, then let the algorithm optimize delivery.
5. Creative fatigue isn’t about a “magic” frequency
Many buyers let frequency climb to 5-6 before seeing decay. Some pull ads right as frequency hits 2.0.
Ultimately, the key is rotating winners, not pulling ads at 2.0 frequency out of habit.
To time it right, you need to keep a close eye (weekly is good) on frequency and click-thru-rate (CTR) trends for each ad set. We’ve been using a beta product we built at Prism that helps us uncover potential creative fatigue in just a few minutes.
6. Conversions per ad set matters
A simple one, but some people miss it: Meta’s learning phase requires ~50 conversions per ad set per week. Without enough signal volume, campaigns can stagnate, and even best practices won’t yield results.
7. Match optimization to your source of truth
Align Meta’s optimization window with the source of truth data. If the source of truth attributes only on a 1-day click, then meta campaigns should be optimized to the same. Including view attribution directs the algorithm to give you more conversions it can attribute to views. For lower-funnel campaigns, this misalignment can cause a drop in incrementality. Basically, you’re giving up ad dollars to convert more people who may otherwise still have converted.
8. Attribution Lag Skews Performance Read
Meta attributes conversions to the day of conversion, not the day the ad was served. After budget changes or launches, results often lag. Don’t judge performance too early: expect a delay in seeing the true impact of scale-ups or cutbacks.
Lots of people overlook this one, just because it’s not called out directly when you pull reports from Meta.
9. Watch for Over-Concentration in Ad Sets
Meta often pushes spend toward one or two top-performing ads, starving new creatives. Dynamic ads especially tend to chew through ad budgets. In this case it makes sense to break out these ads into their own ad sets to have a manual control of their spends.
10. Email and lifecycle flows prop up paid ROAS
We see it all the time: a brand reduces email sends and ads start underperforming.
If you ever see results suddenly drop, you can blame Meta’s algorithms if you want to have a venting session, but you should also check on changes to any simultaneous Email or SMS campaigns, even if you don’t think there’s a direct causal relationship. Meta is the acquisition engine; owned channels are the margin engine.
11. Scale budgets in smaller increments or with duplicates
We obviously vary our approach depending on client goals, the market, budgets, etc, but a good rule of thumb:
Increase budgets by +10 % a week. Or duplicate the ad set/campaign at a fresh budget, then prune losers. When you increase budgets in big jumps, it can make it really hard to tell what you’ve learned and damages efficiency as you miss chances to optimize.
12. Don’t day-trade your budgets
Small budgets amplify volatility. Small swings from day to day are just noise, and can be distracting. To find the signal in the noise, look at 7+ day averages.
If you want to fine-tune your ad spend on a daily basis, try dayparting.
13. Let the algorithm re-allocate spend inside a CBO, then break out star ads if they hog delivery.
When one ad soaks up budget, clone it into its own ad set so both the new test and the proven ad get room to run.
I’ve tried to avoid the most obvious tips that you’ll likely already learn elsewhere. In my mind, these are the lessons that I either missed completely, had to learn the hard way, or at least aren’t discussed enough given the impact they can have on your ROAS.
A big part of running successful, efficient campaigns on Meta Ads Manager is having some wisdom - some years of experience under your belt that helps you avoid mistakes, wasted dollars, wasted time, wasted clicks.
I’m sure I’ve missed some. I’ll try to come back and keep this list updated as I continue learning.
