Karen Grace said
Hi John, hope you're well.

I'm starting to wonder whether it's hurting me that I'm using the same images for my ads that I've used for a couple of years now, and I'm starting to wonder whether when I do ads, facebook is serving them to the same people, as an initial 'tier', and as a result those initial conversions are expensive and the algorithm doesn't then recover...

This is one of those situations where your logic makes sense, and some of that could be happening, but ultimately its not likely significant enough to make a difference. Because you're optimizing for conversions the algorithm will quickly look for patterns based on conversions and should theoretically filter out people who have seen the ads, if that segment is not converting. It's also just a reality that the audiences we use are typically so large that it's just not a significant factor with a budget like the one you are working with. That said, one would expect that you would need/want to change up your artwork every so often to keep things fresh for multiple reasons.

This is related to the fact that I'm not sure how effective my audience exclusions are, because a lot of them are based on the last 180 days. As you know, I've had long gaps between doing ads, and it actually is 6 months since I last ran some ads (!), so most exclusions are meaningless, except people who have landed on my subscriber success page which I do also exclude ! I've done a LOT of testing over the the last couple of years, so I do wonder whether a lot of the same people are seeing the same image they've seen a few times over the last couple of years, and glazing over. For example, on one ad I am currently running, I am using an image that was once very effective at getting cheap clicks, and has been tested with dynamic creative in the past on this same audience (as has the text & other elements), but clicks this time round are really expensive and unviable on this ad that I have now been running for 3 days- $2 a CLICK !

Like I mention above. It's possible but not that likely in my opinion. It's more likely changes in your audience and/or the algorithm. But hey, when prices go up you need to test new things. Changing up ads is part of the process. So try new images and see what happens.

I also realise that I haven't excluded people who have watched videos I have put on facebook, (I have made some custom audiences of them, but forgot to exclude them). I wonder whether those people assume I will entertain them on the fb platform, (and therefore don't bother subscribing), and so I do think it's probably best to exclude them and treat them independently ? - I do have ideas for how to do that! ...

Again, because you are optimizing for conversions. This shouldn't be a major factor. Some marketers swear by excluding warm audiences, I typically don't. It's not a bad idea to do so, but in my experience it's six of one and half a dozen of the other. By excluding warm audiences you get a truer metric, but by including them you often get some extra social love and beneficial engagements. All that matters is results. If one approach isn't working, try the other approach.

If you think my instincts may be correct that I might be hitting the same initial audience again and again, would you advise that I try excluding my video audiences as above or any other tricks to get beyond this possible issue? 

I am also really up for doing another photoshoot or two, to change up images that people see, because all my photos are a few years old now anyway (pure laziness!). This would involved a fair bit of testing again, but may be worth it- do you advise artists to keep their images fresh? I'm aware that text also has an impact, but I wonder whether images are the best place to start ? (I know it's the first thing I personally notice.)

Many thanks in advance for any thoughts.

Karen

  

I don't know if your instincts are correct or not. But if not excluding them is costing too much, try excluding them and see if your prices improve. If they don't then you know that's not the issue.

Changing up images from time is always a good idea. But changing them should not take a ton of testing. Usually the same targeting and creative will work similarly. You'll just either see a drop or increase in price based on the images. Best of luck