Rabbbits Weeekly: Meta Mega-Thread
Or, the one in which I string a lot of articles and thoughts together to make Meta look really bad and then share tips to improve your ad performance on their platform.
Hey, it’s me, again. Just double dipping in your inbox this week.
This time, an all things Facebook Meta edition. Because they can’t get themselves out of the headlines and some of my thoughts/feelings/aversions about them have been coalescing. So I figured, why not dump it all in one go on my fellow (poor, unsuspecting) rabbbit chasers.
Unofficial Meta mantra (probably): Grow or Die
Facebook is a growth engine. Since the early days it has been maniacally focused on growth, damn the consequences.
More users. More features. More engagement. More ads. More money. More countries. More apps. More avenues. More alternatives to the real world.
I described the big 5 tech companies a while back to a friend (something) like this:
Google wants to the be the world’s brain
Microsoft wants to be the world’s (digital) infrastructure
Apple wants to be the lens through which you see the world
Amazon wants to sell the world
Facebook wants to be the world
If you view Facebook’s past through the lens of recent news and announcements (especially the Meta pivot) that last bullet feels more and more clear.
There is a lot of high-falutin’ corporate speak about “connecting the world” and building “technologies that help people connect, find communities and grow businesses.” And many of their efforts to paint themselves as a victim in response to some societal or privacy related change is couched in terms of how the change would, essentially, be bad for your neighbors (small businesses) or ability to connect better. But, ultimately, I think Zuck’s goal is to Ready Player One the world.
Sarah Frier, the author of No Filter: The Inside Story of Instagram, had a great conversation with Web Smith, of 2PM, on his podcast, Polymathic Audio. It puts some helpful context around the Instagram acquisition (way back in 2012), which was, in hindsight, one of the more important moves Zuck+co have ever made.
It may be hard to remember now, but Facebook used to suck at mobile. Hard. So they primarily bought Instagram as a mobile leapfrog opportunity with a side of competition minimization. They initially dismissed its opportunity as a growth engine (I can’t decide if this part was a mistake or a con by Big Blue).
The two platforms took polar opposite approaches to their growth philosophies. Instagram fell on the concierge end of the spectrum while The Facebook™ went for brute scale. As Sarah shares, Facebook thought “why would you ever cater to one user, celebrity or not, when you could be catering to millions of users, tens of millions, hundreds of millions?”
Even more telling, to Facebook, “growth was more about hacking us. Hacking the public’s attention and time spent on Facebook, so that we became more and more dependent on the platform. Done via notifications, emails driving clicks.”
And when growth is threatened or they can’t buy their way out of competition, then they go full politik. Leaked emails suggest Meta has retained a political consulting firm to give TikTok the mudslinging treatment. Using misinformation and misdirection to get Meta out of the headlines and popular opinion turned against their existential threat in The Trend Machine™. Some of the tactics (allegedly) include: xenophobia, blaming TikTok for trends that started on Facebook, data fears, harm towards kids, etc. The classics, basically.
(Disclaimer: I don’t use TikTok. A) I’m too old to understand it. B) I already feel like I don’t have enough time to do the stuff I want. C) I don’t really use any social media (see point B). D) I don’t know what the connection to the Chinese government is and that’s not exactly an entity I trust.)
It’s starting to feel like the metaverse is The-Company-Formerly-Known-as-Facebook’s last shot to stop the slow skid to stagnation we’ve seen befall many a giant company in the past. (Early returns aren’t looking great.)
Now, on to the:
Bad Reputation
Meta’s reputation at the moment is…not great. One of the things they did to try to counteract this, specifically related to radicalizing/polarizing content, was launching their Widely Viewed Content report. Which is really just opaque data propaganda. But even their own reports manage to highlight some concerns:
So, a page that broke your rules managed to claim the #1 spot for the quarter before you realized and removed it? It’s fine. It’s probably fine.
How did they get to a point where this report was needed?
(This thread is a good cautionary tale on why you shouldn’t build your business in someone else’s walled garden. You don’t have an audience if you do, you’re just borrowing one.)
There was also that time their downranking system broke and the spread of misinformation and other not good content spiked.
But cracking down on watchbait (clickbait for videos) will definitely help though. Right? That’s the problem. Gotta be.
They are adding some brand safety features for ad placements, so I’m sure all these issues will disappear shortly.
Even when they try to improve security they manage to screw it up. This time they added a feature called Facebook Protect, basically two-factor authentication but they rebranded it (because of course), except their heads up emails looked like a scam and now people are getting locked out for a host of reasons related to bad implementation.
Anti-Meta
A new social media platform is blowing up on college campuses (allegedly). You know, the places The Facebook™ got its start.
BeReal is positioned as the anti-Instagram. You get to post once per day, at a random time selected by the app. Take a photo and post it or don’t, that’s your choice. But it’s the only choice you get.
This is an interesting concept, and, like TikTok, might be different enough from the entrenched giants to gain a meaningful foothold. But only time will tell.
The Metaverse was Hacked
I mean, not really, but virtual world game Axie Infinity (one of the darlings of web3 and metaverse) was. $625 million in crypto was taken. Whoops.
Meta (the company, not the -verse) was hacked (along with Apple). Giving customer data to some peeps impersonating law enforcement.
(Speaking of the metaverse, Zuck recently said something along the lines of their metaverse dreams being roughly 5 years away. “5 years” is the standard tech timeline given when the real answer is unknown and probably way longer.)
The Engine is Humming Again
Facebook ads are suddenly performing better according to “many small businesses”. Maybe everyone just needed some time to adapt to the changes and get their systems in order. But definitely totally normal and nothing to think too much about, right?
Uhhh, well... Facebook lied about a metric in ads manager (again), inflating reach by as much as 400%. This is why I recommend you ignore what Facebook tells you and determine performance based on what happens on your site. (Metric monkeying is pretty common practice across these companies)
But wait! 3D shopping ads are coming, so that should fix everything.
But wait… It appears we’ve fully entered the “rearranging deck chairs on the Titanic” phase of the business lifecycle as their big new change to the ads platform is to slap “Advantage” before some of their ML powered features’ names. Instead of fixing the product that is generally hated by marketers, they’re playing the name game. I’m noticing a trend to their strategy: don’t fix problems when you can distract from them.
They are giving users more control over how ads can target them. But do any users know this tool exists?
Ultimately, this feels like another one of their PR stunts where they appear to be making changes but they are still ignoring the actual problems. Like the fact that user interest categorization appears fundamentally broken. Saying you hate cheese might fill your feed with cheese ads, because they missed the context (I know, shocking that Zuck+co could miss context clues). The research suggests that Meta is overeager to assign topical interests to users, which leads to inaccurate interest targeting options for advertisers. My guess is the approach is at least partly designed to grow the audience sizes for interests, damn the consequences (see: the unofficial motto up top).
Zuck is going to jail!
I mean, probably not, but a new UK Online Safety Bill includes the provision for criminal sanctions against senior executives.
This piece discussing the Bill more also makes some of the points I make in the Australia bit (below). Except with more useful words and fewer question marks.
Maybe Meta won't have to leave Europe after all as the EU and US seem to have hammered out a data storage agreement.
But the presumed forthcoming Digital Markets Act will likely shake things up for them.
But, The International Mint of Zuck’s Bucks is being investigated by the EU for colluding with Google to prevent other companies from competing with The Goog's display ad bidding program. The fact that this is because of Big Blue's terrible Audience Network placement is humorous because it is definitely not a feature worth antitrust headaches.
Oh, and they’re being fined in Ireland ($18.6 mil) for violating GDPR.
US regulators aren’t falling over themselves to defend Meta, so (like GDPR) this may be a sign of things to come on the company’s home turf.
Trouble Brewing Down Under
Meta is being sued for allowing scam crypto ads using images of famous Australians to run. I get the reasoning for the suit but I’m not entirely sure what systems (read: algorithms) Meta is expected to have in place to prevent this. What constitutes a “famous Australian”? What makes a crypto ad obviously a scam? Are all ads with images of people supposed to be flagged for manual review? Who is reviewing them? Do they know who is famous in Australia?
Steal Like a Tech Company
In a surprise twist, TikTok is stealing from Facebook. Well well well, how the turntables. Maybe Stories are still the future of social media / internet content after all.
In an effort to ensure we don’t forget who started the sticky-finger circle, Facebook is working on Reactions for Reels to grow engagement. Because, remember, if you’re not growing, you’re dead.
And they’re combining their various Insta video things into one (bye IGTV), so basically they’re putting TikTok in Insta. Again.
Over on The Trend Machine Lite™, Instagram product tagging is now available to all users, so maybe now is finally the time for social commerce?
Speaking of stealing, the book Steal Like an Artist by Austin Kleon is celebrating 10 years. (No affiliate link or anything, I just like Austin and the work he does.)
An Actual Good Thing
Instagram videos will now have auto-generated captions. So maybe we can have nice things every once in a while.
If you’re still here, you’re clearly desperate to find some shred of value in this thing, maybe these will help.
Now that I’ve slammed Meta repeatedly, how about some tips to better advertise on the platform? Because, let’s face it, it’s a powerful marketing channel.
First up, should you let Big Blue pick placements for you or take that into your own hand? I like the ultimate takeaway, go with automatic placements but keep an eye on things (especially the dumpster fire that is Audience Network) and go the manual route if you notice either some budget vampires or some high flyers.
Now that you’ve decided where your ads will show, who should you show them to? Here’s a quick look at what is supposed to be working for top of funnel efforts now (which is probably what your focus should be on here with all the data collection issues they’re having). In a nutshell:
Broad demo and/or interest targeting vs. lookalike
Larger % match lookalike audiences
Stack stuff that used to work together until the audience is big enough to hopefully work again
I say test these ideas out but don’t go all in from the beginning. I’m still seeing the standard lookalike approach crush in some accounts.
I highly recommend this thread of the Facebook bidding algo (you can set this under campaign bid strategy in the budget pane). Even if just to understand how it works. Here’s the gist:
Bid caps aim to maximize your conversions at your target CPA without regards to budget (a.k.a. where we’re going we won’t need spend limits)
Lowest cost bidding aims to spend evenly over the course of a day while maximizing conversions and CPA (which sounds nice but what if 80% of your conversions happen after work hours but 70% of your daily budget is already gone? (made up math, don’t check it))
The recommendation: give FB an overly high daily budget but use bid caps to limit off-peak (performance) spending
In the tweet thread’s example, he spent 7% more but increased conversions 15%
So while you can overspend with this method you should be overspending on more results at better cost pers.
I have run a small test of this in one account and the early returns tell me it’s worth trying out. The purpose was to increase traffic and we’re driving down CPCs to levels from the halcyon days of FB ($0.15 over the last 30 days).
My caveat: make sure your campaign is set up to optimize for an action that feeds back to the Facebook algo quickly. We started with a pixel conversion event and nothing was happening. Switched to link clicks and it turned into a rocket.
Back to your regularly scheduled smattering of marketing news and nonsense next week. Probably. I’m clearly off the rails with this thing already.