Article

23 Meta Ads Pitfalls – And How To Avoid Them

Advertising Analytics Consumer Behavior Data Marketing Meta Testing

September 9, 2024

As I’ve sorted through the hundreds of emails I’ve received, the most common topic request is for help with Meta Ads. More specifically, here are the actual requests from readers:

Can you write something on Meta Ads Accounts gone wrong? We used to see tons of sales, but since the Spring our account has spent more with fewer sales. How do we figure out where we went wrong?
What are the pitfalls and mistakes that accounts make that stagnate growth and kill profitability?
What should we (a lead generation brand) be doing to maximize our potential on Meta?
What are the biggest issues you see in Facebook Ads accounts?

If the commentary on X, LinkedIn & at conferences is any indication, these readers are not alone in their frustration with Meta: there’s a lot of brands, agencies & individuals struggling to replicate their pre-iOS14.5 performance.

I’m in the fortunate position to be able to analyze 100s of Meta Ads Accounts – from small spenders (<$10k/mo) to positively massive accounts (>$1M/mo). One thing I’ve noticed – regardless of the size or scale of an account – is that there are a number of commonalities across under-performing accounts. Broadly, these can be divided into 4 groups:

Group #1: Preparation & Research

There has been a sustained – and surprisingly effective – push by Meta to drive brands and media buyers to adopt broad targeting via the “Power-5.” While this isn’t new (the Power-5 was announced in 2018/19), the adoption, particularly of broad targeting, didn’t surge until 2020-2021.

The theory underlying the move to broad targeting is that Meta, with its advanced targeting algorithms, unfathomable amounts of data, and machine learning, would be able to find a brand’s target audience simply by analyzing the creative (and the performance of it). For awhile, especially for brands with broad appeal, broad performed exceedingly well – to the point where a significant number of accounts (40% to 50% of the ones I’ve reviewed in the last 12 months) have dropped all other targeting and run exclusively using broad (these brands maintain geo and appropriate demographic targeting, but don’t use LALs, Interests, etc.).

A casualty of this has been audience research and insight – after all, what’s the point of analyzing your competitors, alternatives and audience, understanding the challenges, pain points and desires of your target market, segmenting your existing lists for LALs, and making strategic investments in maintaining all this data if you’re just going to let Zuck take the wheel?

One of my guiding principles in anything I do – from media buying to investing – is to zig when everyone else zags. Incrementality has always been, and will always be, in the places where most people are not. That’s not abstract philosophy; it’s mathematical reality.

I have a working theory on why broad targeting – especially at the outset – worked well: there’s a decent-sized segment of the Meta user base that has an astonishingly high proclivity toward buying stuff, submitting lead forms, requesting demos, etc. For lack of a better term, the individuals in this segment (writ large) are the digital equivalent of the Home Shopping Network population: if it’s something they aren’t precluded from using, and it is served in the feed long enough, they’ll think about buying it. This isn’t an exact analogy, but it’s sufficiently accurate for the purposes of an explanation.

Meta knows who these users are. A material segment of these users do not fall into Meta’s pre-existing interest audiences. For Meta, that’s a problem – advertisers weren’t reaching the full audience they could (because interests stacks, LALs, etc. limited distribution), so there were highly-valuable segments of the user base, representing exceedingly valuable impressions (i.e. expected value per impression exceeded the cost per impression). Meta’s solution was broad targeting – a solution that worked exceedingly well at the outset (adoption was slow, so competition for those high-value impressions was also low, resulting in case study worthy ROAS / CPA numbers….which had the benefit of increasing adoption, which led to rising CPMs on high-value, low-visibility audiences, and so on and so forth). This is a quick, if imperfect, summary of how we got to where we are today: the majority of advertisers using broad in some capacity.

But, here’s the rub: when every advertiser uses broad, the available advantage declines. The purchase-happy segment of the Meta user base is only so large. The end result? Meta now has to work harder and longer to find the commonalities among each advertiser’s converting users. The machine, smart as it is, has picked all the low-hanging fruit (and most of the mid-level fruit, too). The proverbial HSN audience is saturated.

This is where savvy advertisers (“smart marketers”) can add value. In the film The Imitation Game, there’s a marvelous scene in which Benedict Cumberbatch’s Alan Turing breaks the German Enigma. He does it not by raw power (testing every possible combination each time), but rather by focusing all of the power he has on the most probable possibilities (in this particular case, by identifying several words that must be included in each transmission). Marketers who have invested in research and audience insights can do the same thing with Meta. We can use LALs (particularly well-segmented ones based on users who buy a specific type of product or brand, or those who exhibit certain behaviors (subscriber vs. one-time customer; spend over $X during Y period, etc.) and interests to point Meta’s (insanely powerful) machine learning in the right direction.

Are we going to do 2015-style granular targeting? Absolutely not. And even if we tried, it would likely work about as well as if Alan Turing and his merry band of mathematicians tried to crack Enigma using a quill and an abacus.

Instead, what high-performing accounts do is:

Conduct diligent, ongoing customer insights research.
Use those insights to craft both relevant targeting (audiences) and creative
Leverage existing customer data to create both exclusion and targeting lists
Segment existing customer data into relevant, discrete segments – buyers of X product/brand; prospects who requested a demo, but didn’t show up; individuals who exhibit Y behaviors. Then use those segments as seeds for LALs.

We’re far from alone in observing that well-constructed LALs, Interest Stacks + Hybrids (a LAL with an interest stack overlay) tend to produce more efficient outcomes at acceptable scale than broad alone. And, when you critically examine how Meta’s ad platform works, that outcome should seem obvious, not anomalous.

Group #2: Budgeting, Metrics & Data Flows

Data is the lifeblood of every Meta Ads Account. If your data is flawed, faulty, broken, incomplete or otherwise busted, don’t expect your ad account to perform to its fullest potential.

There’s a reason Meta has invested billions (yes, with a “B”) into building out their data capabilities – from CAPI, to support for third-party attribution (a masterstroke), to native integrations with virtually every widely-used CMS. It isn’t because they want to make life easy for marketers; it’s because that data represents a durable, scalable competitive advantage. That same logic applies to individual advertisers (at a much smaller scale): the ones with the best data will have an advantage.

To be very blunt: exceptional data is table stakes on Meta. That alone isn’t making or breaking ad accounts. Where the true advantage lies is in how that data is supplemented: the budgets, metrics and forecasts.

Virtually every high-performing Meta Ads account I’ve reviewed has the following three things in common:

A Well-Designed Forecast:

Forecasts are, by their very nature, exercises in being wrong – and that’s why most brands don’t bother. It’s always why I insist on having them in-place. There is real, significant value in the exercise of creating a forecast: the pursuit of the relevant inputs, the discipline to craft the model, the understanding of the business/organization that comes from seeing how people, money and product move through it. The best forecasts are living, breathing entities. They evolve with the organization. They serve as a check on spending and a “GPS” (of sorts) for marketing investments.

The magic happens when a well-designed forecast is integrated into ongoing reporting, such that campaign performance is evaluated relative to forecast at a granular level. Instead of trying to compare CPAs for the past month vs. the prior month and/or vs. the prior year, we can now compare actual CPA vs. projected. We can repeat this with all constituent metrics of CPA/ROAS, too – CPM, CTR, CPC, CVR – to gain insight into what’s driving the under- or over-performance. That data, in turn, leads to smarter optimizations and more focused teams – after all, if CPM is on-projection, CTR is double and CVR is only 55% of projected, the CPA would look on-target. But, by having this insight, the savvy Meta Advertiser can see that real work needs to be done on the Conversion Rate (whether that’s improving landing page, clarifying the ad, changing the offer, adjusting the message, etc.) in order to either (a) beat forecast or (b) avoid disaster if/when a CTR regression strikes. Without a forecast, that likely never happens – leading to a situation where the advertiser is reacting/panicking, vs. proactively addressing potential issues.

The Right Budget:

Nothing kills Meta campaigns more than under-funded campaigns. If you’re selling a $100 product with a $50 CAC target, setting a daily budget of $25 is the advertising equivalent of slow, miserable, painful, frustrating death. A good rule of thumb is to set each campaign’s daily budget equal to (# of Ad Sets * 7.25 * tCPA). So, in the above example, if the campaign has 2 Ad Sets, you’d set the campaign daily budget to:

2 Ad Sets * $25 * 7.25 = $362.50 per day

If you can’t afford that? Run one ad set. Still can’t afford it? You’re not ready for Meta Ads. There’s no point in throwing a fraction of a single CPA into the Meta machine if your goal is sales, growth or anything else positive.

Distinct KPIs For New & Returning Customers:

There’s an all-too-frequent conversation that occurs between CMOs + CFOs (or COOs):

CFO: We’re spending a ton on marketing. Too much, in fact. We need to right-size expenses in order to hit our profitability targets. Where are we most efficient? Let’s focus on those areas and cut the rest.

CMO: The most efficient campaigns are remarketing, current customer upsells/cross sells, and our lapsed customer re-engagement.

CFO: And if we reduce everything else, but increase our spending on those

Campaigns, could we still hit our forecasted revenue or lead targets?

CMO: Yes.

CFO: Let’s do that.

And – that’s typically where the conversation ends. But, there’s a second part to this dialogue that should occur:

CMO: But – if we do that, we’re effectively mortgaging our future for a short-term Revenue boost.

CFO: Explain.

CMO: Remarketing, existing customers, lapsed customers – these are all people that have an existing relationship with our brand. These people all know us. They aren’t incremental; our incremental customers come from prospective marketing, organic search, referrals, and the like. If we cease those campaigns, new customer growth will stagnate and we’ll end up losing market share to our competitors while alienating our current customers.

CFO: That seems suboptimal. Bad, even.

CMO: It gets worse. With no new customer/audience acquisition, we’ll inevitably fall behind on product (since we won’t have as much data coming in), while the value of our brand – defined as our ability to command premium pricing, accelerate deal velocity, and generate new customer acquisition without paid channels – will collapse. We’ll be forced into a commodity position, where we will inevitably get crushed by larger companies.

CFO: That seems terrible.

The solution to this is twofold: (1) set distinct KPIs for new customer acquisition vs. existing audience activation/retention (i.e. remarketing) and (2) find new ways to reduce the cost of bringing in those new customers (i.e. moving away from discounts to other, less costly offers that are more likely to resonate with your target audience – based on your audience research).

In more ad accounts than I care to count, acquisition is being consumed by remarketing, because (shocker) remarketing is more efficient (higher ROAS / lower CPA). The only way to grow a company is by bringing in new customers at or below an acceptable target. No matter how you slice it, if you don’t have a good handle on both (i) what that target is for a given customer/client/prospect profile and (ii) what your actual cost to convert a net-new customer is, you’re effectively flying blind.

Set clear, defined KPI sets (efficiency + volume) for both new customer acquisition and existing audience activation.

Group #3: Account & Campaign Structure

Much has been written about Meta Ads campaign structure – in fact, I wrote an entire article about it a year(ish) ago. And while I’m a noted advocate for cost controlled campaigns, I’ll be honest: I’ve seen lowest cost campaigns work. I’ve seen bid caps perform. I’ve seen cost caps perform. There’s no single “ideal” account or campaign structure.

That being said, there are – objectively speaking – better practices.

Over the course of evaluating 1,000s of Meta Campaigns, spanning 100s of accounts in just about every imaginable industry, there are 8 common campaign structure issues, pitfalls and blunders that tend to surface in poor-performing accounts:

Too Many Campaigns: “mini-campaigns” / “micro-campaigns” / “niche campaigns” – whatever you call them, they’re problematic. These campaigns tend to be under-funded (see above note on budgets), redundant (if you’re going to create a distinct campaign, it should be for a distinct purpose – a specific offer, a target audience, both – and both the targeting and the creatives included in that campaign should reflect that audience/offer) and self-sabotaging (these campaigns fracture optimization data, making it more difficult for Meta to parse signal from noise). Bottom Line: if you’re going to create a campaign, create it for a purpose, make the necessary investments (budget, creative, time) to allow it to fulfill that purpose and operate it with a purpose.
No Testing Strategy: This one is endemic across the Meta Ads ecosystem: far too few brands actually have a coherent, defensible test-learn-evolve strategy. This is not merely running A/B tests (though those can be useful); it’s understanding how to test creatives, identify winners and scale accounts.
Optimizing For Bullshit: Without fail, at least one campaign in every low-performing account is optimizing for bullshit – clicks, engagements, video views, unique reach – instead of the thing the advertiser actually wants (sales, leads, inquiries, phone calls, applications). To be exceedingly blunt: Meta will give you exactlywhat you ask for – so just ask for what you want. Just as not all clicks are created equal, not all engagements, or views, or reach is equal, either. If you optimize for clicks, Meta is going to serve your ad to people who tend to click on ads – but probably not people who like to click on ads and buy (because smarter advertisers are willing to pay more than you for those people).
Trying To Out-Smart Meta: For as long as I’ve worked in this industry, there have been people claiming they can hack, beat, loophole, or otherwise out-smart Meta, or Google, or Microsoft. Whether that’s running an old Top-of-Maze | Middle-of-Maze | Bottom-of-Maze structure, or thinking you have some magical psychographic voodoo, it never works. The true cheat code is being brilliant at the basics, committing to do the work, and then executing on it.
Not Taking Advantage of Whitelisting: If your brand uses influencers or creators, you should absolutely be using whitelisting. This allows the brand to access the creator’s audience and handle for the promotion – unlocking new potential customers/clients and maximizing creative-audience relevance (since the people following that creator will be more likely to respond positively to a creator they’re already familiar with and following).
Lack of Proper Exclusions: This should go without saying, but the number of prospecting campaigns I see with existing customers NOT excluded is…shocking. The same is true for other obvious “no-go” characteristics – location, age, job title, etc. If there are characteristics that you don’t want in customers/leads/prospects, and those characteristics are excludable on Meta….exclude them.
Not Managing Comments: This is an easy one: check the comments on your ads. If they’re good, respond to them (or just leave them alone). If they’re negative, hide them. It isn’t difficult, but it makes astaggering difference in performance. Also: use Post IDs for moving ads, so that the existing social proof transfers. Yes, it’s a PITA, but it’s worth it.
Forgetting The Fundamentals: While there are all manner of sins in this bucket, the big ones tend to be: (1) not having the right data in the account; (2) over-segmentation; (3) an ad hoc account structure with no rhyme or reason behind it; (4) running DPAs without (at least) price bucketing – preferably bucketing based on contribution margin and (5) not aligning creative – offer – audience in each ad set / campaign.

Group #4: Creative Strategy, Process & Implementation

Despite all of the above discussion – 3,000+ words of it – on things other than creative, here’s the simple truth: creative is still the single-most-impactful factor on Meta. You can get a lot wrong on the data, structure, budgeting and strategy sides, but still do OK if your creative is that good. In football terms, having great creative is like having Pat Mahomes as your QB: you’re not guaranteed to get the win, but you can be pretty awful in a lot of other aspects and still have a fighting chance.

That being said, most brands don’t have great creative. In fact, I’d go one step further: most Meta Ads accounts I’ve reviewed don’t even have good creative. Here are the 11 most common creative issues I’ve seen:

No Creative Strategy & Process: The simple, painful reality is that most brands don’t have a creative strategy or process. They make ads when they feel like making ads. There’s no structured process to connect research and insights to creative formats, angles and offers. And without that process, they inevitably fail to create the quantity and quality of creatives required to scale their account.
No Creative Velocity: I’ve written about this extensively, but for Meta Ads creative, quantity drives quality. You must have a critical mass of assets flowing into your account in order to scale. If you’re not adding new creative into your account regularly, you’re going to have a difficult time.
One-Size-Fits-All Creatives: The entire benefit of Meta Ads is the ability to target distinct groups of people with hyper-relevant messages – something that was never possible with traditional media (TV, billboards, radio, print). That entire advantage is forfeit when advertisers run the same creatives in every campaign.
Running Obvious TV Ads: Within each of the last 10 Meta Ads Accounts I’ve reviewed, I’ve found at least one TV commercial running – horizontal only, 0:30, no cut-down. While I have zero issue with brands maximizing their creative (those shoots aren’t cheap), I have a staggeringly large issue with brands failing to do even the minimum amount of customization to make that creative work for the platform on which it will be served.
Lack of Creative Diversity: If every ad in your account looks like it was created by the same designer, or shot by the same videographer – you have a problem. The same is true if your entire account is chock-full of statics, or carousels, or videos. The best performing accounts have a stunning diversity of creatives, from multiple creators, to multiple editors, to a balance of ad formats. If you’re wondering how to get creative, we use Soona and Trend for eCommerce product/photo shoots, plus have a full creative team and 3 photographers on staff.
No Authenticity // Overproduction: If every ad in an account looks like an ad (or worse, looks like an overproduced ad), it’s pretty much a foregone conclusion that the account is going to under-perform. Social media is an inherently authentic medium – it’s people connecting with people. In this kind of organic environment, over-produced, inauthentic ads stick out like Shaq in Times Square. That’s not to say every ad needs to be ugly and shot on an iPhone 14; but rather, there needs to be a balance.
Benefits vs. Features: Far too many brands are OBSESSED with features. Spoiler alert: precisely ZERO prospects care about your features. They care about what those features do for them (AKA Benefits). If the vast majority of your ads are talking about your latest [blah blah blah] instead of how your [blah blah blah] alleviates a burning pain point for your target audience, your account will struggle.
Not Mapping Creative To Audience: The world’s most brilliant piece of creative will fall flat if it’s shown to an irrelevant or otherwise uninterested audience. The entire reason why we invest heavily in segmented audience understanding is so we can better map our creative onto each group – whether by appealing to their desires/ambitions, tapping into causes they care about, addressing their challenges/pain points, or highlighting their fears. Every piece of creative in your account should be mapped to a specific audience – and that specific audience should be aligned with the targeting in that ad set.
Lack of Hook Testing: The hook is (arguably) the single most important component of your ad – but most brands refuse to experiment with different hooks. There’s absolutely no reason, in 2024, for every ad in an account to begin like a 1998 infomercial. Test different hooks. Test different narrative arcs.
Excessive Copying of Competitors: I’ll be the first to admit that I absolutely am in favor of competitive research. It’s a dog-eat-dog world, and refusing to do something because your competitor did it first is stupid (just do it better). There’s a ton of unmined brilliance in the Meta Ads Library, the TikTok Ads Library, and the Google Ads Transparency Center. But, as with all things in life, there’s moderation. Supplementing your creative assets with concepts adapted from your competitors? Smart. Making your entire creative strategy to copy whatever X competitor does? Likely to end poorly.
Not Tracking Relevant Creative Metrics: I’m always surprised when brands don’t have any creative metrics tracked – hook rate, thumbstop rate, hold rate, etc. While these are not KPIs (nor should they be considered KPIs), they’re immensely valuable in evaluating your creative performance.

In closing this exceedingly long issue, I want to point out that 95% of Meta Ads accounts have issues. If you’ve read this and found yourself admitting, “Yeah, that’s us. That’s us too. Uh-oh…” – you’re not alone. It’s relatively easy to point out everything that goes wrong; it’s exceptionally difficult to actually do the work every single day to address those issues and prevent them from resurfacing. This is my way of saying: don’t get discouraged. Admitting you have a problem or a shortcoming or an issue is the first step in addressing it.

Cheers,

Sam

Related Insights