OPINION: Common FB Advertising Advice That We Disagree With
We believe the best way to improve is by challenging one another and sharing learnings.
We recently reviewed a massive piece of content put out by a wonderful and smart company in our space and wanted to mention that while it does include some awesome nuggets that every ad buyer should know, it also brings out some pieces of advice aren’t necessarily best practices across the board (to us).
The biggest takeaway here is that as marketers, the answer should never be ALWAYS YES or ALWAYS NO. The answer should be “it depends.” Every single ad account created on Facebook is different, as is every single human being on the planet. As is every company’s customer base and what they need, want, or are looking for at any given moment. Just like you want something different for lunch today than you had for lunch yesterday, we’re humans. We change our minds.
So or anyone who ever teaches that one way is the only way to do something with regard to marketing, we’re … skeptical.
Ready? Let’s get nerdy.
FB Machine Learning (aka the all-knowing algorithm): The point here is that using machine learning can help to minimize human error, optimize spend, and reduce management busywork. While YES, machine learning can be important, we believe it’s also important to know when the algorithm/machine learning is working and when it’s not. Sometimes, the machine isn’t well-oiled and we end up wasting money on what a robot thinks will make a human spend money. Humans aren’t robots and they don’t operate via a formula. There’s still a part of marketing that is intangible, unable to be calculated, creative, and a mystery, if you will. This is why sometimes ABO is better than CBO or why you have to be in an account daily (or even multiple times per day) to pull levers based on the full picture (like MER - have you read this blog on why we’re moving to MER from ROAS?) instead of JUST what’s going on inside the Facebook platform.
CBO vs. ABO: Yes, CBO used to be a best-practice. It used to be how we ran 95% of our campaigns. We even created an entire Foxwell Digital Course on why CBO was amazing. However, post-iOS14, Facebook ads have lost a LOT of data. Because of this, CBO and the algorithm’s ability to optimize effectively has decreased quite significantly. While as an advertiser, we would love for Facebook to optimize budgets for us, that means to spread the wealth of the budgets evenly and intelligently, not dump all of the budget into one ad set that’s at a 0.40X ROAS in-platform. This is what we’ve seen since iOS14 was released, hence why we’ve been switching almost exclusively to ABO. Yes, humans are not as smart as machines, but we do understand a bit more about humans and about quick changes when the machines are on the struggle bus. Every account is different, so we’re only speaking for the hundreds of accounts. we see on a weekly/monthly basis, but if your CBO campaign is dumping a significant portion of your budget on a less than amazing ad set, try switching to ABO and see what happens.
Cost caps: Neither Cost Caps, Bid Caps, nor Lowest Cost campaigns are the magic setup for a 10X ROAS, but cost caps CAN work. They also sometimes don’t AT ALL. They also work insanely well in some accounts and not others, or work one week in one account and not the next week despite all variables being the same. This is why testing is important. And continuous testing is a must. What worked last year likely won’t work now. What worked last week may not even work today. With all that being said, cost caps also essentially spend whatever they want, so if you’re on a tight budget, don’t try these at home. In perfect world, cost caps (and bid caps) would work flawlessly and you would only spend when you could get a CPA at or below the set cost cap. In theory, YES, cost caps are the way to go (unless you’re scaling to the actual moon and want to light money on fire). BUT, that’s not the case for Facebook ads right now (maybe one day).
The learning phase: In our experience, it is NOT as big as it’s cracked up to be. A campaign/ad set doesn’t have to have 50 conversions to be successful. For smaller accounts, 50 conversions in a week isn’t possible. 50 conversions in a month may not be possible for some ad buyers and small accounts. Yes, the more data the campaign/ad set/ad/pixel has, the better. But the magic “50” number is not the holy water of printing money in your ad account – sorry. If you can, aim to get 10 or 20 conversions per week per campaign. Or if even that is too much, try to set your budget to where you can get even one conversion a day (set the daily budget at 10 or 20% higher than the account AOV). No budget is too small to run ads, so just be flexible with what works for YOU and YOUR account. Again, EVERY account is different. There’s no one-size-fits-all strategy that’s going to work for an account spending $100,000/day and $10 per day.
Combining prospecting and retargeting in a single campaign: Maybe it’s just the perfectionist and organizational nut inside us, but we recommend splitting out prospecting and retargeting in different campaigns. It helps tell the story better of what’s happening in the ad account and who is converting and when, in our opinion. There’s also the argument of Facebook attributing revenue to the last touchpoint instead of the first if you don’t set up your ad accounts this way – we’re already losing a sh*t ton of data in Ads Manager, we likely wouldn’t see this information anymore anyways, a conversion is a conversion (we’re just thankful it was reported), and last touchpoint IS still important, because it’s how they ultimately clicked “purchase.” More info is always better, but this isn’t a dealbreaker for us. But if you like it this way, we’ll silently judge your disorganization and will still celebrate you when it’s successful.
Budget breakdown: Let’s start this off by saying that literally every single account is different in how the customer journey goes and at what point they usually convert. We have accounts with amazing middle funnels because they have great organic web traffic or organic Instagram engagement, and accounts that the middle funnel is not a thing and you either buy in prospecting or you don’t buy at all. HOWEVER, with that being said, on average and as a general rule, we suggest having a budget breakdown something like 70% prospecting, 20% middle funnel, 10% bottom funnel, and then if you have a product that is highly repurchased/high LTV, you could take some percentages away from each of the funnels for a loyalty funnel of existing customers. Keep in mind that, in order to grow, you have to spend a significant portion of your budget in prospecting and acquiring new customers.
Naming conventions: Very important. Definitely all they’re cracked up to be. Do this. Our naming conventions look like this:
For campaign level: DATE_FUNNEL_SPECIFIC SALE/LAUNCH/GROUPING_LOCATION (if running multiple countries or geofencing)_AUDIENCE NAME_CBO/ABO_OPTIMIZATION. (091721_TOFU_Fall Launch_3/5/8% High Value LLA Stacks_ABO_Conversions).
Ad set level: DATE_FUNNEL_AUDIENCE_PLACEMENT (091721_TOFU_3% HV LLA Stack_All Placements).
Ad level: DATE_FUNNEL_Creative Name/Differentiator_Type of Creative (091721_TOFU_Dog with subtitles 30 seconds_Video).Exclusions: Also important, though don’t get entirely carried away. At a minimum, we suggest excluding purchasers last 180 days (or all-time) and website visitors last 180 days on prospecting campaigns. Then, as you get into middle and bottom funnel, we usually exclude 7 or 30-day purchasers or whatever else makes sense based on the audiences we’re targeting. Also, if you use an app like Motion, naming conventions are super useful (and 1000% necessary)
Automatic Placements: We usually use them, unless you have a reason to want to run an ad on one platform only. We will do manual placements if we have account data and knowledge to prove something substantial, like our Instagram conversion rate is 50% higher than our Facebook conversion rate, or I want to serve different ads to Instagram Stories and Reels than to the Facebook feed. This is where your human instinct and knowledge comes into play.
Worldwide targeting: Sure, this can work. However, we don’t suggest scaling to worldwide immediately. We also like focusing on two to five, even 10 countries that your shop sees consistent sales from instead of just going worldwide and getting some low-value traffic. However, if you can ship worldwide and you are scaling to the moon, sure, give it a whirl, but expect numbers to decrease on account of more eyeballs. Another note here is that if you’re going to run worldwide campaigns, you absolutely must continually check performance country by country. Otherwise, you might spend $11,000 at a $28 CPM for $900 in purchases. Yes, worldwide campaigns used to work. Now, post-iOS14.5, they can be dangerous. Learn, and adjust.
Conversion-only campaigns: We’re big fans of conversion campaigns, BUT we’re also big fans of doing whatever works in your account without strict absolutes. Sometimes video view campaigns (and retargeting those specific ad video views with a conversion campaign) can work, or even optimizing for ATC or IC to get more data coming in and then retarget them with a conversion campaign can work. Some super high AOV products have a hard time getting conversions within a 7-day window, so other optimizations CAN work, if your account has a good strategy surrounding it to support it. And, for what it’s worth, about 90% of our campaigns right now across all clients are conversion campaigns. They’re just not the ONLY way.
If you take anything from this novel, take this:
Marketing is never going to be a one-size-fits-all approach. No two accounts are the same. Nothing is going to work forever and ever. There are nuances everywhere in digital advertising. Testing is key, so is individual optimization and knowing what you’re doing.
Want to learn more about these cool (and nerdy) ad tactics? Check out our courses.
