NEW Way To Test Facebook Ads!

0:00Right now Meta is rolling out a new Facebook Ads feature that is going to make it much easier for us to test new Facebook Ads and Instagram Ads. Testing is incredibly important and we really needed something like this. Let me show you what I'm talking about. So, to demonstrate this new feature, I'm actually one of our client ad accounts because this feature isn't yet available in our example ad accounts that I usually use for recording purposes. And that's the first thing to note is that it's not available everywhere yet. It's available in about half of our ad

0:26accounts at this point across the board. I know that Meta is rolling it out. Some ad accounts have had this for for quite a while, but it's a fantastic feature. I really want Meta to roll this out to all ad accounts. We spend more than a hundred million dollars a year advertising on Meta and 30-40% of that is testing new ads. So, anything that makes the testing process easier makes our life a lot easier and this certainly fits into that category. So, to check to see whether you've got it, you want you want to go to the ad level. Most people think about testing ads as being

0:54something you do at the at the ad set level or potentially the campaign level with structure and things like that, but it's actually at the ad level. And if you scroll down within ad set up beneath multi-advertiser ads, we've got this show more settings option. So, it's quite hidden if you don't know about it. If you click on that, you will hopefully see push delivery to this ad. If you

1:14don't see it, you don't have it yet. Hopefully, you will you will get it soon. And you can see that Meta explains what this does. Help ensure your ad gets delivery by dedicating budget to it. Now, that's been one of the biggest complaints that Meta advertisers have had for a long time is that they're being told by people like me and by Meta, you need to create more ads, create a diversification, as many ads in there as possible, high-quality stuff's going to help improve performance. So, they spend a load of time and effort, potentially cost, making new ads. They introduce them into their ad sets and they get no

1:43spend. The existing best performers continue to get all the spend and they think, "What am I doing this for? This is a total waste of time. I've just gone and made these things that hasn't got any budget." Well, this feature is designed to resolve that exact problem because we can turn it on and then you can see the options, right? So, the first thing you want to do is is decide what percent of budget spent per day should go to test ads. So, the default here is 20%. I think that's actually a reasonable default. In terms of testing, people often ask me, "What percentage of your budget should go

2:14towards testing?" We normally sit somewhere in the 10 to 50, maybe even 60% range. It really depends on how good your results are. So, if your campaigns aren't converting at all, they're not working, well, 100% of your budget is testing. Whereas if your campaigns are performing really well and they're really profitable and you want to be able to test, but you also want to be able to benefit from the great performance you're getting with the existing ads performing really well, then we're going to reduce that percentage down a lot more. That's where we might be testing 10, 20%, something like that. So, really should be dependent on your objectives

2:49and how well things are performing. There's also a trade-off between short-term and long-term. So, testing more, spending more on testing is going to reduce your short-term return on ad spend, but makes it more likely that you'll find winners that allow you to then scale and and spend more over the long run. So, there's also that trade-off to uh to be aware of. So,

3:06there's no perfect amount of testing. Hopefully, that helps in terms of the the framework and the parameters that we're uh that we're looking to do. Now, like with all budget things, Meta has some flexibility. So, they say, "Look, we'll aim to spend an average of $30 per day in this case cuz the campaign budget is 150, but it may well fluctuate from day to day." Meta does that because they know on some days of the week you get better performance. So, they'll try and spend more on those days, less on on others as more data comes in. And then next we've got duration. So, how long do

3:32we want to force ad spend into the test The default is 7 days. I think that's reasonable, but this time frame should be adjusted according to to a few different factors. The most important being conversion volume. So, Meta learns when you generate conversions, when you generate more of the thing What that means is that a campaign that has generated 20 conversions in an hour is just as old, if we're going to use that term, as a campaign that has generated 20 conversions in 2 months. That's very counterintuitive and difficult for people to understand. They're like, how long should I leave the campaign? It's like, time is irrelevant. It's all

4:09conversion volume. How long did it take you to get those 20 conversions? In 2 hours or in 2 months? There are businesses that fall into both those categories and even more extreme, right? Depending on on your business, do you sell bespoke high-ticket services and you only work with a handful of clients at any one time? Are you selling something relatively inexpensive and it's a product or a software or something and you're selling tons and tons and tons of them? So, the most important thing is going to be doing your conversion volume. The fewer conversions you generate, the longer you need to run this for so that you get

4:39enough spend on those test ads actually know if they can perform or not. And likewise, the opposite. Conversion volume is going to be a function of what it typically costs you to get a conversion and your budget. The more you spend, typically, the shorter the time frame, the less you spend, the longer the time frame. And that doesn't mean, by the way, that we would change this from 7 days to like 10 and then we'd max out at 10 days. No, no, no. We might run this for a month. We might run this for a significantly less time if it's an account that's getting loads of

5:09conversions and we just want to test to see. Once you've run this, so let's say we put this back to 7 days, which is probably going to be an average and a default. Once we've run this and it and it comes to an end after 7 days, your ad will use your existing budget after it's ended. So, basically, delivery returns to normal. So, if this ad's a good performer in amongst the rest of them, then it's going to get spend. If you perform some spend, it hasn't performed

5:29very well, it's unlikely to get spend. Just like Meta does with regular ads, right? The better performing ads get more spend, uh the worst performing ads get less. And don't worry if you get this time frame wrong initially. If you use this feature, you push delivery to this ad, you start with 7 days, at the end of the 7 days you just haven't got enough data, you you make a call on whether the ad performed well or not, you can always redo, extend the time frame, etc. etc. And over time you'll get a sense of like what's appropriate for your ad account um and your business. Now, there are a few other

5:56things I need to run through about this new feature including alternatives, which ones you want to use and when. But first, I want to very quickly let you know about my new school community called the Ben Heath Meta Ads Academy. There are just under a thousand people in there right now and the feedback so far has been fantastic. People are finding it really valuable. I'm in there all the time and personally responding to every single post, question give it a thought, I would just love Ben to get back to me on on these questions that I've got about how I should approach my Meta Ads, you know, things that maybe

6:21should be tailored to my specific circumstances. Well, you can get that within the community. 110 plus exclusive videos you won't find here on YouTube, just exclusive to the community that teach you everything you need to know to get the best possible results with your Meta Ads. Weekly live Q&A calls with me. If that's of interest to you, there's a link in the video description. We're currently running a discounted introductory offer that will be going away soon. So, if you want to get involved, go ahead and do so now. So, what's the point of using this new feature? Pushing delivery to this ad forcing spend on the on new ads? Well,

6:50firstly, it stops ad creative, new ad creative from being stopped. I briefly touched on that already, but it can be really frustrating if you create a load of ads, they don't get any spend, you think, "What was the point? How do we know if this could perform well? We're just getting more and more budget spent on the existing ads. They're fatiguing, results are dropping off." I'm hearing

7:05this story all the time at the moment. So, it's great for helping to resolve that problem, give new ads actual spend data, see how they perform. It also protects delivery of the existing ads. I think this is really important cuz what a lot of advertisers will do in that circumstance when new ads aren't getting any budget is they'll turn off the old ads and they'll force all the budget into the new ads and that's really risky cuz if the new ads don't land as well as the previous ads were performing, you're going to see a big drop off in performance. So, I like that we can

7:34allocate a proportion of the budget to testing new ads without interrupting what the existing best performers are doing too much. And that gives us a nice balance between testing new stuff and making sure that we still generate leads and sales in the things that are really important to our business whilst whilst we're doing so. With this structure, you don't need extra ad sets, you don't need extra campaigns for testing. By using this feature, you can just force spend onto

7:59the new ads and see how they perform. And that's really important particularly for ad accounts with lower conversion volume because we don't want to fragment our data. We don't want a handful of conversions in this campaign, a handful of conversions in this ad set, handful of conversions here. In an ideal world, you have as many conversions going through individual campaigns and individual ad sets as possible. You can only do that by having fewer campaigns and fewer ad sets because remember

8:22that's what allows Meta to learn. Remember I said before about your campaign is just as old if it's got 20 conversions in an hour versus 20 conversions in 2 months. Well, we want as many conversions in an ad set and a campaign as possible so that that ad set, that campaign can mature, Meta can learn, they can optimize, they can help us get better results. So, I really like it for for that reason. Now, this feature is similar to the creative testing tool, which is also at the ad level. If we were to continue to scroll down within this ad and we get down to beneath where we add in the ad creative,

8:54we've got this option here called creative testing. Now, you can see this is grayed out because it says, "You're pushing delivery to one or more ads in this campaign. You can only set up a creative test after these push delivery periods have ended." And what that indicates is that these features are very similar in their functionality. If I go back up and turn off this, we will now be able to use the creative testing

9:14tool where I can set up a creative test. So, whilst this is similar to the push delivery to this ad feature, there are some significant differences. So, with the creative testing tool, this is really designed to test a batch of ads. So, you can test between two and seven. If you've got a whole bunch that you've created and you want to test, make sure they get spend that's pulled away from the existing, then this can be a great way to do it. I've got other videos running through the creative testing tool and the setup and, you know, how you can change these settings down here to make sure you you've got it set up

9:40properly. The issue with this is that a lot of ad accounts don't have the conversion volume to get enough data. So, I talked about previously making sure you've got enough time to be able to get enough conversions on the new ads to be able to actually assess whether perform well or not. Well, if you're doing that with say five ads at once through the creative testing tool, well, it's going to take a lot longer than if you're just testing one ad at a time, which you could do through the push delivery to this ad feature. So, in ad accounts where you're getting lots of conversions, dozens, hundreds of

10:09conversions per week, then this can work well and you're going to get data within potentially a few days, maybe seven, maybe 14 days, something like that. In a lot of ad accounts where they're generating a conversion or two a day, if you set up a creative test with five days, you're going to have to run that test for a month or longer to get enough data, maybe more, and that's just not feasible for a lot of businesses. It's going to take too long. So, you should actually should be better off with the push delivery to this ad feature. So, I think it really depends on the

10:33circumstances, it really depends on on the conversion volume. It's not budget, remember, cuz some businesses generate conversions for $2,000 each, some businesses generate conversions for a dollar each. Like, it's really conversion volume that that matters. So, this is a good alternative. I don't want you to watch this video and think, "Okay, we're always using the push delivery to this ad and not the creative test tool." Absolutely fine to use this, it's just you need to know the circumstances when this is and isn't appropriate. With that in mind, one of the absolute keys to scaling ads is working with the right data. Meta ads

11:03especially are driven by AI targeting. Bad data means not only bad ad decisions, but bad AI targeting. Meta tracking is consistently off by 20 to 30% and as much as 50% plus for long-term sales. This leads to that bad data. That's why I want to tell you about Hyros and their new form of attribution called AI swarm data. It fixes this gap and A, allows you to make the correct decisions with your ads, B, sends the right data to Meta. This results in an almost immediate boost to ad ROI via better targeting and better decisions, which is why it's used by major players like Alex Hormozi, Tony

11:34Robbins, myself, Four Seasons, Playboy, Watt, and thousands of large-scale e-commerce stores. Here's an example. Hyros helps point out gaps in tracking. If I couldn't see this data, I wouldn't be able to scale as effectively. More importantly, it feeds the ad platform the right data for better targeting, cutting your cost to acquire customers. That's why I'm comfortable recommending Hyros. They literally guarantee increased ad ROI or you don't pay. The usual result is around a 10 to 15% increase in ad revenue. And if you're an agency, not only does it get your clients better results, Hyros pays up to 40% for lifetime commissions on each client you bring. Even better, they will

12:06set it up for you, then help you optimize ad platform AI targeting with it. Use the link below. The first 15 people to do so will get a 30% discount. Okay, so let's run through some best practices when it comes to this new push delivery to this ad feature. So, the first thing to understand is the trade-off. If you go ahead and test a load of ads at once, then your results are likely to get worse in the short run. This is because most of your test ads will not perform as well as your existing best performers. A lot of advertisers are surprised when this happens, but if you think it through, it

12:36makes total sense. Let's say you've been making ads for 3 months, and currently all your ad spend is going into two ads, which are your two best performers of all the ads you've made over the last 3 months. Well, if you go in test six eight new ads, for example, what are the chances that those are going to perform better than the best two out of all the ones you created previously? It's actually really quite low, and that's why testing reduces performance, why you're likely to see a hit to your return ad spend in the short run. We do it because what we want to do is find

13:05those that can outperform our existing best performers, because that might be a total game changer for our business. If the best we know to do so far is a 3x return on ad spend, and we choose a new ad and that produces a 4.8x return on ad spend, that might allow us to spend three times much of the campaign. That might allow us to acquire many more customers, achieve much higher levels of of profitability. But, you have to understand that trade-off. So, the more aggressive you are with testing, the more you feel that hit. The more ads you test at once, the more likely you are to

13:35see a drop-off. So, if you really care about existing performance whilst being able to to test as well, I would say to use it sparingly. Test it with an ad here, an ad there, not too many ads at once, and potentially don't dedicate too much of your budget um to this. You also want to be careful using this all the creative testing tool around big sale events. Think like Black Friday, Cyber Monday, for example. Those sorts of campaigns, that sort of setup is not really well geared up for allocating a small proportion of the budget to testing, to finding out how this ad is going to perform over the long run,

14:07because there isn't a long run. We need to generate sales right now. In those types of campaigns, I'm much better just throw the ads in there, let Meta work it out really quickly. A huge proportion of your budget's going to go to just a few ads, the best performers. That's okay, because in those time periods, when we're running timely sales, we're not trying to learn for the long run. We're trying to be as profitable as possible over the short run. So, I wouldn't be doing much in the way of testing during those sorts of campaigns. There would be an exception if you've got a a repeating

14:36time-based campaign. So, let's say, for example, you're running a campaign to the end of the month, and then next month you're going to have a different sale, and next month you're going to have a different sale. Many businesses operate like this. Well, then you can learn things during this sale event that will help you in future sale events. So, that's a little bit of a side, but like the big ones, the ones you don't do very often, Black Friday, for example, then uh let Meta do its thing. You're better off doing that. I would also recommend that you use this feature for bigger changes as opposed to minor tweaks. There's not

15:02much point in forcing delivery on an ad that is just a different background color to a previous ad. With those, just put them in there, let Meta work it out and do its thing. I just don't think there's that much value in making sure that that has spent to learn how that performs, because we know it's going to

15:19be similar to what we had previously. It's just a background color change. There's going to be minor differences. If we're going to be using this sparingly, you're better off doing it for those. Okay, we change a different format, we change a different style. We've got a totally new hook, we've got a new avatar, we change the angle. Like the big stuff that can make a a really big difference. And then of course, make sure you give it enough time. Like if you're forcing delivery to an ad and you're wanting to know how that performs, if you run the test and it's only generated three conversions during

15:46the test, probably don't have enough data to be able to make a call on it. If you do that, run it again, run it longer. I often get the question of like, if I've got lower conversion volume, how do I assess performance earlier? You don't. You can't. You can't cheat the system. You can't use a different metric to try and work out if this ad's performing well or not. You have to do it based on return on ad spend, cost per conversion, your north star metric. And if that takes time, that just takes time. Otherwise, you're guessing and you could easily make bad decisions. Also, remember that creative

16:18diversification still matters. And a mistake that I've seen advertisers make quite a bit recently actually is that they're going to use features like this and they're going to do a lot of testing and they're going to try and split test their way to the perfect ad. And what I mean by that is the one that produces the best return on ad spend, the best cost per conversion on average to your

16:36total audience. But post-Andromeda Meta is able to do personalized ad delivery. So, the ad that performs best to one person is going to be a different ad creative to the one that performs best to someone else. So, you can easily get into the point where like, okay, well this ad's got a slightly better ROAS, so we're going to run with this one and we're going to turn these other ones off. I actually think that's often a mistake because we want to leave those on. Yes, a smaller proportion of your budget may be spent on those other ads. It may not get as good a return on ad spend or cost

17:04per conversion as the other ad, but it might allow you to generate extra incremental conversions because it allows you to tap into parts of the audience you would not have been able to do with technically the better ad on average. Hopefully that makes sense, right? We need lots of different ads. You don't just want to split test your way to the one perfect ad. You need lots of different ads cuz different people respond to different ads. Meta can use

17:28them in in conjunction with each other. It's where just going off the data and looking at the numbers can lead you down the wrong path, particularly if you're you're in that like super testing mode and trying to split test your way to it. Now, in the ad accounts that have this, we're using this push delivery to this ad feature quite a lot. If you don't have it yet, absolutely fine to use the creative testing tool. There are some differences I briefly touched on, but it still allow you to test new ads. And I demonstrate exactly how that works including all the settings you need to

17:54run through for it in this video here. It's also a great option if you operate with a larger budget or more conversion volume. So, go ahead and check it out.