| Hey, Welcome to Part 3 of my Facebook Advantage Survival Guide - where we are navigating the minefield that is running ads in 2024. Perhaps a little unfair. Facebook remains the very best place to find readers at scale - hence our focus this year. But the landmines are real, which is why we are taking the time to cover this topic carefully. In Part 1 we covered the huge changes Facebook has made - which flummoxed experienced hands and left newbies with a vicious learning curve. We broke down how the algos work - and how they can work against us if we are not careful. In Part 2 we got more practical, showing you the exact settings to switch on and off in each stage of ad creation to avoid most of the nonsense and give your ads the best possible chance of success. Today, we are diving into the most gator-filled pool of all: targeting. But there is a treasure chest at the bottom of that pool... if we can avoid getting into a deathroll with any more terrible metaphors. Targeting 101 When I started advertising on Facebook maybe ten years ago, success was all about targeting - specifically interest targeting. Facebook had an exhaustive list of interests for advertisers to choose from, and millions of users which it paired to those interests through various means, he says somewhat euphemistically. For authors, this proved exceedingly useful as the list of interests contained the names of big authors with big audiences and a decent variety of genres and sub-genres - sometimes with surprising granularity, even if it was a little patchy in places and not exactly dynamic. You could mix and match too. To give a simple example, if you had already exhausted George R. R. Martin's considerable audience, you could target the Game of Thrones TV show instead. Or maybe the Lord of the Rings movie. Or a dozen other genre-flavored things. Throw a Kindle interest into the mix as well and - hey presto - you had the bones of a successful ad. The system wasn't exactly friendly to newcomers (or oldcomers for that matter) but you could generally control where your ads were displayed, and who saw them, so it was possible to muddle through and find success. And once you had a handle on profitability, you could then scale up from there and really grow your audience. Targeting 2024 In its continuing mission to simplify ad creation, Facebook has - irony alert! - made everything much more complicated. And more perilous too because things can go off the rails now in an instant. The crashes are more spectacular too and then to make matters worse you don't really learn anything from surveying the wreckage because a lot of the Advantage stuff now happens inside a black box. Why was Ad A successful while Ad B wasted every single penny? Facebook has now made answering both parts of that question much more difficult. To the extent that you can have a wildly diverging set of results when Ad A and Ad B are identical in every single aspect. Great job, world! Luckily, there are ways around all of this. Or most of it anyway… The Solution The way through this mess is quite straightforward even if the finer points can be tricky. As I said last week: 1. You need to switch off as much of the Advantage stuff as possible. 2. Your targeting/Audience needs to be on point to nudge/influence/corral the bits you can't turn off. We covered the first in detail last week, but the second will require some explanation. Best practices around all this are still being hammered out – not least because Facebook keeps coming up with new ways to try and force into using all these features. But I've been reading every bit of research I can find, as well as conducting lots of testing myself. Two things seem clear: (a) Facebook will aggressively override your interest targets if Advantage+ Audience is not turned off - one of many reasons I strongly recommend doing this. I thought they would still act as something of a steer to the algos when Advantage+ Audience is off, but I was surprised how aggressive it is - basically rendering your targeting inputs more or less irrelevant. (b) Facebook can still override your interest targeting selections when Advantage+ Audience is switched off... but it pays a lot more attention to your inputs. Let's zoom in on this because - clearly – that nuance is key. When does Facebook ignore me Last week we spoke about how Facebook can override demographic controls like gender if you leave Advantage+ Audience switched on. Which is yet another reason for switching Advantage+ Audience off. Many advertisers assume the same is true of interest targeting, where if they are targeting Nora Roberts readers who own a Kindle, and they have Advantage+ Audience switched off, Facebook won't engage in AI tomfoolery to expand their reach into areas which are unhelpful. This is no longer true as of 2024. Facebook can and will go beyond your targeted interests if it thinks it can get you cheaper clicks even if you have Advantage+ Audience switched off. Some voices seem to be saying that we are basically powerless now and should throw our pants out the window and embrace the algo with wild abandon. Big mistake. We might have lost absolute control in terms of targeting, but we still can influence when and how Facebook expands our targeting – and we can do that rather heavily. But only if you have Advantage+ Audience switched off. If you leave it on, Facebook will virtually ignore your suggestions and just find you the cheapest clicks available, anywhere, from anyone, whether they are a romance reader or not, own a Kindle or not, read books or not. But if you turn it off, Facebook pays much more attention to the inputs and keeps things mostly going in the direction you have nudged it. And, as an example will show you, it looks like you can upgrade that nudge to a solid shove too. Let me give you an example. Game of Drones Let's pretend I am a fantasy author. For simplicity's sake, my main comp author is George R.R. Martin, so I generally like to run Facebook Ads targeting his audience – or the subset which owns a Kindle, to be specific. So, in my interest targeting, I choose George R.R. Martin, and I narrow that audience with the Kindle interest. My targeting inputs could be completely ignored if I leave Advantage+ Audience switched on, so I double check it is off. George R.R. Martin has a large audience and my ads are reasonably profitable, but eventually Facebook will seek targets outside of my inputs, especially if I am spending a lot of money, running the ads for a long time, targeting his readership in a smaller market where the audiences will be much smaller too, or perhaps targeting devices with smaller user bases. Facebook gives us little information about who it expands the targeting to, but we can get an idea by looking in the targeting suggestions box, where it seems likely the system will expand the targeting to include first JRR Tolkien's audience, then Brandon Sanderson's, and then perhaps people who like Emilia Clarke – as in the actress from Game of Thrones. You can see how it might not take that much for an ad to go off the rails. However, if JRR Tolkien and Brandon Sanderson are also good targets for me, and I target them together with George R.R. Martin, the system is much better at staying on track. Going by that targeting suggestions box again, it seems likely the system in this case won't expand the targeting into those who like Emilia Clarke, but instead into areas more suitable like Lord of the Rings and Terry Pratchett. Which is considerably better. Your Takeaway is Ready, Sir It's important that you understand how the system works, the ways things routinely go wrong, and how to dodge those common pitfalls. There's a lot of complexity to wrap your head around, which makes me super wary of "simple" solutions to all this stuff. I started working in digital advertising back in 2004 and run huge budget Facebook campaigns all the time and I'm still processing some of the changes over the last year or two and tweaking my approach. That said, following this framework will help you a lot with your targeting: 1. Turn off Advantage+ Audience. (And always double-check that it is switched off when duplicating an ad, restarting an ad... just make it a habit to always check it's off - and that goes for all the Advantage settings I spoke about last week. Facebook is getting increasingly sneaky about forcing us to use Advantage features, but we can almost always slip the noose). 2. Test potential interests individually. (Don't run tests for too long, especially with smaller audiences, as Facebook will reach outside your targets eventually and pollute your data.) 3. Ditch the losers, put all your winners into a supergroup. 4. Save that Audience! So you don't have to input everything again the next time. 5. For bonus points, you can keep testing after you have your set of winners and add them into your supergroup in an ongoing fashion, once they have proven themselves worthy of joining the band. Returning to the above example momentarily, my fantasy supergroup might contain George RR Martin, Brandon Sanderson, and JRR Tolkien. If I wanted to expand things further, I could test Terry Pratchett next, and then if the results were good enough, I could add him to the supergroup and save that new audience – and keep growing my targets in that manner. While we can no longer eliminate all targeting expansion, my testing seems to indicate that we can limit its missteps by feeding the algo targets which we know work for us. Yes, this might make testing a little trickier, but once you can get over that hump and have a few workable targets, things should start getting easier for you. Please note I'm using that word in a relative sense lol. Finally, let me underline the recommendation above to save any useful Audiences. In tandem with that, I also suggest familiarizing yourself with the Audience interface generally. Getting a good handle on your Audiences - and the various types of Audiences too - is key to scaling up. And we'll talk about that next time. Dave P.S. Writing music this week is Pete Seeger with Mr. Baggins. | 
![[THE LATEST] Inside hell's kitchen 👩🏽🍳](https://lh3.googleusercontent.com/blogger_img_proxy/AEn0k_u0jeTUdCRRPeVkTYAJE5GlteipYH5EXBsD7LGCXv6I6EiQg1TytcGD77jIHSrZNbwfDZSmF9V9alp4cTQDV5JHwE9TigFtu0olXqysrnYfMZ_UA2_pa4B9Iaug7C7ptEtpzcZ6db1R3gQayVh0dx0l0RWnvgfPld3zGZrSO3Q9eFRUOnKusL3X-HFcYVeDPqyhoZK-zQACpMWcliKRoyixOMptCjhaxQJ8qgsV9QFC6-u5uiv9VDS9kPNgLETeSF5DUEpSNynRSSXsmH6ghJ0=w72-h72-p-k-no-nu)
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.