Facebook Ads in Learning Limited!? Here's The Fix!

0:00If you've run Facebook Ads before, there is a good chance you've seen the dreaded learning limited status applied to one or more of your ad sets. And the big question is, what should you do as a Meta advertiser when you see learning limited? I'm going to show you exactly what to do in this video. So, let me start by quickly explaining what learning limited actually is, and what does it mean when that status, that delivery status, is applied to one or more of your ad sets. What Meta is effectively saying is that you aren't generating enough results within a short enough time period for Meta to be able

0:30to fully optimize that ad set and get out of the learning phase. So, I'll quickly explain what the learning phase is because we need that for understanding learning limited. The learning phase is usually a relatively short period, say 24 to 48 hours on average, although it can vary, after an ad set is first launched or significant changes are made, like introducing new ads into that ad set, where Meta will do a lot of testing and experimentation to try and work out how to get you as many of the result that you've optimized that ad set for. So, they're going to try and work out who within your target audience

1:01is most likely to purchase, for example, or what time of day then someone's most likely to become a lead, or should we put ads on Instagram or Facebook? Should it be on stories? Should it be on reels? All the various variables, there's loads of them, to try and get you the best results possible. Now, Meta continues that process, that doesn't stop, the learning process doesn't stop, but the learning phase, that initial 24 to 48 hours on average, is where mo- like the the heaviest testing is um where the results are typically the most volatile. You can have fantastic results for a few hours and then bad and vice

1:35versa. A lot of Meta advertisers have experienced um this process. Now, typically, once you get to, say, the 48-hour mark, Meta will take that ad set and take it out of the learning phase, and you will see the delivery status active. And then the campaign, that ad set, is just going to run, and like I said, it'll continue to optimize. But sometimes, instead of going into active,

1:54you will see it say learning limited. And Meta's saying, "We did not generate enough results in order to exit the learning phase properly. We remain learning limited. We're not able to to optimize this campaign." Now, if you're familiar with the way that Meta advertising works, you might think, "Well, that sounds absolutely terrible. I know that Meta ads perform much better

2:14if Meta's able to optimize properly." And all things being equal, that is true. But, I do want to sort of start by allaying some fears and saying that you can run a successful ad set in learning limited. You can generate a good return on ad spend. You can generate a good cost per conversion. Ideally, you would exit learning limited. You would get that ad set into active status. You would likely see an improvement in results by doing so. But, it's not the end of the world if you're in learning limited. And I think it's important to understand how bad something is to then know what are the

2:47appropriate measures to take when we're deciding what we should do to actually avoid it in the first place, okay? So, I wanted to clarify that. Quick one before we continue. If you're running Meta ads and you'd like me to take a look at your actual ad account and tell you what I would change, test, or scale, I've built something called the Ben Heath Inner Circle. Inside, you get direct access to me for less than $60 per live session, where you can join, ask me questions, and even share your ad account for live feedback. And these sessions are scheduled across multiple time zones, so you can join from wherever you are in

3:16the world. So, if you're dealing with performance dropping, creatives not working, or you're unsure how to scale, you can bring it to a live session and we'll go through it together. If you want to check it out, there's a link in the video description. Let's get back to the video. Now, to get specific on what Meta say needs to happen for you to exit the learning phase and not end up in learning limited, but in active, is that they want 50 results per week. So, whatever it is you're optimizing for in the ad set, and that's important because that's the determining factor. So, if you've got an ad campaign that's a

3:44traffic campaign, and the performance goal is maximize number of link clicks, then you need 50 plus link clicks per week in order for that ad set to go into active mode, properly exit the learning phase, and not run into learning limited issues. If you've got your campaign set up differently, so let's say you're running a sales campaign, and you're optimizing for the purchase conversion event, then you need 50 purchases per week or more in order to

4:10not run into learning limited issues. Now, obviously, it's a lot easier to get uh link clicks than it is to get purchases for many businesses, so that's why this is an important part, and we'll we'll dig into that more um again later on. And that's the theory that Meta gives us, but in practice, I've seen many ad sets properly exit the learning phase and not end up in learning limited, even if they're generating fewer results than 50 per week. There doesn't seem to be an absolute number where it will or won't. It seems to vary between different ad sets, between different ad accounts. Sometimes it might be if you generate 20 results per

4:45week, um you'll exit and you'll be absolutely fine. You'll be in active. Sometimes it might be more like 40. But Meta sort of say 50. And I don't think I've ever seen an example where a campaign or an ad set has been running more than 50 results per week and has ended up in learning limited. So, I don't think that's ever happened that way. 50 sort of if you get there, you're

5:01safe. You're you're all good, okay? So, let's talk about fixes. If you do find yourself in learning limited, you're thinking, "Well, this isn't ideal. I'd like to get out of it. I think my campaign perform better." You're probably right in that assumption. Uh what do what do I do to to get myself out of learning limited? So, what Meta would typically recommend is just spend more. And that makes sense, because if you spend more, you get more results. Depending on how you've got that ad set optimized, whether your your performance goal is purchases or leads, you spend more, you're probably going to get more purchases, you're probably going to get

5:30more leads. That gives more conversion data to Meta. Meta's better able to optimize the campaign. Now, we do have to be somewhat wary whenever the ad platform says that the fix to your problem is to give them more money. Doesn't mean it's the wrong advice. We should just always be potentially a little bit skeptical and think, "Okay, is that actually serving what I'm after here or what they're after?" We also have to factor in what's right for the business, not just what's right for optimizing the ad campaigns. And I get a lot of questions like Q&A's all the time. I get a lot of questions from

6:06people asking things like, "Is it okay if I turn my campaign off for a week? We've run out of stock and turn it back on again." And they're worried that from an optimization standpoint, not having the consistency of leaving the campaign running, so turning it off and then turning it back on, might hurt the meta algorithm gods and you know, they're going to run into issues there. I'm like, "Well, ideally the campaign would keep running. It would be consistent as opposed to going off and on. But you've run out of stock. Like the business

6:33priority has to take precedence there. It has to be more important. You don't want to be generating clicks from customers coming through to a page where they can't buy the thing. They're likely to be annoyed. Like that's just not That's not smart." And I think the same applies here when it comes to getting out of learning limited. If you don't want to increase your budget because you couldn't handle more sales or more leads right now, then absolutely fine. Don't

6:55do it. I also wouldn't want you to just whack your budget up to get out of learning limited in a campaign that isn't performing well. So if your ROAS is way below where it needs to be, your cost per conversion is too high, you're not profitable yet, you're you're going through that testing phase, you're learning yourself as an advertiser, maybe experimenting with different offers, different creative, all the elements that we know can massively improve your results if you get them right, but there often is some experimentation and a lot of tests don't work, some do. So there's a process to go through for most advertisers. If

7:25you're in that process and it's not profitable, don't go and spend a bunch more money on it. And thinking that, "Oh, if I just spend enough to get out of learning limited, Meta will then be able to optimize my campaign and it will then be profitable." That may happen if your results are already decent, maybe not quite where you want them to be, but okay, you might get a 10% lift, 20% lift from doing that. But, if you're operating at a 0.6x return on ad spend, and you need to be a 3x return on ad spend or higher, that's not the fix. I would rather you kept a

7:58smaller budget, improve the elements of your campaign with that smaller budget, improve the ROAS within learning limited because you can do that for sure, and then scale once the campaign's better and improved and and and doing it that way. So, panicking and just whacking up the budget probably isn't the right move for most advertisers, like I just described. But, in ideal world, we would get out of learning limited and get that delivery status to just active, give Meta the conversion data it needs to be able to optimize. So, what would I recommend you do instead if I'm not just saying just go and spend more? There are

8:29other things we can do. Well, one I've already touched on, which is just make the campaigns perform better, because even if you stay at the current budget level, but your campaigns uh do a better job of generating the result that you're after, they generate more purchases, they generate more leads, well, you are going to get closer to getting out of learning limited if not get out um with the extra results you're generating, the extra conversion volume. So, that's going to be better offers, better creative, all the things I talk about in my videos to help produce better results. You may also find that if you've got better results, you then want

8:58to scale, which will also then help you get even more leads, even more sales, and make it more likely that you'll get out of learning limited. Another thing that I'd recommend just about all Meta advertisers do at this point anyway, but it'll also help you with any learning limited issues you're running into, is to consolidate campaigns and consolidate ad sets. Okay, so let's say you've got five different ad sets, um and each one is generating roughly 15

9:23purchases per week. If you consolidated that down into the one campaign, one ad set, and let's assume that you generated the same number of purchases, that would be 75 purchases going through that one ad set instead. With the five separate ad sets, very likely to run into learning limited issues with probably all of them with those sorts of numbers. With the one ad set generating the 75 purchases per week, very unlikely you'd run into learning limited. In fact, I'd say almost impossible for you to run into any learning limited issues. And that's one of the big reasons why I'm a big fan of consolidation anyway, even for

9:58advertisers that aren't seeing any learning limited issues. Um we know that that learning process continues in an ad set even once it goes active. So, I want to force as much conversion volume, as much data through the one campaign, through the one ad set. Sometimes you still need multiple, of course, but we're trying to consolidate as much as possible to help with that optimization process. You also avoid issues like auction overlap, things I talk about in in other videos a lot. Um but yeah, so consolidation is the way to go definitely if you've got learning

10:24limited issues. One step that you may want to take, and I'll talk about the criteria when you do and don't want to do this, is to change what it is that you're optimizing for at the ad set level. So, in an ideal world, you would optimize for the thing lowest down the sales funnel as possible. Basically, tell Meta to go out and get you what you really want, which for most businesses usually going to be purchases

10:44or leads. But, there may be instances where it's completely unrealistic for you to basically ever get to over 50 purchases per week. Perhaps you sell something that is really high ticket. Um maybe you're a relatively small business, and you're like, "We'd be thrilled with five purchases a week. We'd be thrilled with 10 purchases a week. There's there's no way we're ever going to get to to 50." And many businesses are like that. Well, in that instance, because there's not really a prospect of you getting out of learning limited and for Meta to be able to to work out the delivery and optimization like we'd want it to, you might in that

11:18instance be better off optimizing for a conversion event that is higher up the sales funnel, but that happens more often. So, the classic example would be switching from purchase to add to cart. Ratios between add to carts and purchases are going to vary. Let's say you get five times as many add to carts as you do purchases. Well, if you're operating consistently with seven purchase per week, but 35 add to carts, well, in that situation, it might be better to instead change the performance goal, um the uh the conversion event that you're optimizing for to add carts instead of purchase. Could you see a worsening in that ratio by doing so

11:53between add to carts and purchases? Yes, most likely is in fewer of your add to carts will go on to purchase cuz Meta's optimization is very literal. If you say you want add to carts, that's what Meta's going to get you. But, you might be overall better off just because Meta's better able to optimize that ad set. This would very much be a test and see situation. I'm not making that recommendation it'll always work. I'm saying under these circumstances, you may want to test this and see overall results. When we ran it optimized for add to carts, we got this result over X time period, let's say a

12:26month. When we ran optimizing for purchases, we got Y results. Which one was better? And then you can go on that. If you're in a situation where you're generating say 20 purchases per week and it's absolutely feasible for you to be able to scale in the near future or improve the campaign and get out of learning limited, I would not recommend changing this. I would recommend optimizing for the end thing that you want, which in

12:49this instance is going to be purchases. Hyros is the best tracking and attribution software for Meta ads. It's the software all the big players in this industry use, including myself. I use it to track results for my own ads as well as all my organic results as well. So, if someone books a call for our done-for-you services, for example, I can see exactly where that person came from with a lot of accuracy. That allows me to make better decisions for my marketing and my business. Hyros has been able to track that from our ad spend in this campaign, we've generated 96,000 but 58,000 of that was not reported by

13:19Meta. So, if I was just looking at my data within my Meta ad account, I might have come to the conclusion that this campaign is doing okay, but in reality, this campaign is doing a lot lot better. Now this is a relatively extreme example because of the nature of our business. So most of our product services have recurring billing elements. Meta might be able to see what we make initially from our ad spend on that initial transaction but not those subsequent payments. That's why it shows the importance here of being able to accurately see what did we actually get from our ad spend in order to be able to

13:48make better decisions and optimize correctly. So if you want to improve the performance of your ad campaigns without having to touch your ads, go ahead and sign up to Tyros. There is a link in the video description. The next thing I will say and again this applies broadly to app meta advertisers but also very much helps with learning limited issues is you really don't want to be constantly

14:06tinkering and changing your campaigns. The learning phase is re-entered when you make a significant adjustment to an ad set or when you first launch one. So once it's live, anytime you introduce new ads, it's going to go back to learning phase. Anytime you make significant changes to the settings at the ad set level, you're going to go back into to the learning phase. That

14:26doesn't mean we shouldn't make changes. Very important that we do. Learning limit is not the end of the world. Being in the learning phase is not the end of the world. But if we're going to set things up to get the best results possible, we can be a little bit smarter and still test things, still make adjustments without it having that much of a negative impact. So basically set yourself an adjustment schedule. I'm not going to make any changes to this ad set

14:52more than once every 7 days. And if we've got changes to make, we could be working on things within those 7 days. We could be producing new ad creative etc. etc. But we're going to make sure we do it all in one go. So instead of drip feeding in a new ad creative every day, why don't we put in the five new ad creative that we produced at the end of the week. And then we'll do so again at the end of the week. And just by being a little bit more a little bit smarter, a little bit more planned out around that optimization schedule, um that will help and you

15:18provide your campaigns with more stability. I have seen many instances where ad sets will exit learning limited and turn to active even though the number of conversions they're generating per week hasn't increased. Partly, that's just because as time goes on, you generate more results, you generate more conversion data, so Meta is able to learn. But, also stability just really seems to help. Meta is able to do more testing, they're able to plan more. When I've talked about testing before, I think one of the things that's helped people understand what Meta is looking to do and why you meddling with things gets in the way is that Meta will

15:53plan ad impression delivery frequency. So, what I mean by that is they've worked out that on average for someone to buy your thing, they need to put an ad in front of an individual four times, and they need to do so within 48 hours. So, Meta might plan, right, we're going to put an ad in front of them when they're on one of the Meta apps Tuesday morning, then again Tuesday afternoon, then again Wednesday morning, then again

16:13Wednesday afternoon. If you mess things up and change things around, they might be halfway through that ad impression delivery plan for that individual, and it's been messed up, and that resets their engagement with your ads. You've paid for half of it, you haven't got the result yet, and it's much harder for them to plan the subsequent ad impressions that are going to be required to get the conversion. That's why stability matters. We have to change things, we have to introduce new ads, we have to be able to test, but just try and do so in batches and less frequently than constantly being in there and messing around with things. Now, most

16:48Meta advertisers that struggle with learning limited are operating with a small budget. And if that's you, I'd strongly recommend you check out this video next, cuz in that video, I give you the adjustments and changes you want to make to your Facebook ad campaigns, your Instagram ad campaigns, if you are operating with a small budget, cuz there are things that you want to do differently. Go ahead and give this

17:06video a watch.