And just looking at ads manager compared with like the actual sales it can be like 50,60 70% off, which is crazy. It’s like a tracking thing in addition to on average, I have seen kind of a hit and results are just especially with purchases, a little bit less so with leads, that’s usually something easier to get people to do the first time they interact with you.

And then some of them were just driving like traffic and views and stuff. And that’s not really impacted in the same way. Because it’s not purchased as especially big purchases. $1,000 purchases, that’s what really is taking the biggest hit.

Yaron Been: I’m wondering what comes up that took the biggest hits was using more of lookalikes. I mean,

Was there any pattern that you noticed except for the price that was higher, and they were more based on a probably a longer life cycle of the customer before buying?

Is it also related to the way that you were using interest or look-alike in important to just going broad?

Jacob Landis: I’ve seen an impact with lookalikes, you have kind of like a set audience and especially people who aren’t spending huge amounts of money, it’s for them to perform best, you want to get like 1000 new people in over like 60 days.

So for some people, like through the Facebook ads, that’s challenging, and that number just got cut in half. So that’s taken a hit, I am seeing a hit with broad targeting that was working really well, with stuff that applied to everyone just really wide open audiences, if it has gotten better and better.

And now with those, it just seems like it’s not quite optimizing correctly, it’s, you’re still kind of paying a premium on Facebook to it’s a lot more expensive than other ad platforms. But the algorithm was so good and made sense. And now some of the time it’s just like overall getting fewer results.

But having trouble like sticking with positive results like volatility, so it used to be a week goes by, it’s doing well, the chances of B doing well a week from now are pretty high, some stuff kind of falls off a cliff quickly.

And now some of them are just it’s like a roller coaster-like it’s doing well. And then the next day, it’s just kind of a nightmare.

Yaron Been: I encountered the same roller coaster. But this is one of the reasons that I moved away from the lowest cost. I’m mainly using more conservative bidding, which I’m mainly using a bid cap lately. Not even cost control.

What are you seeing in regards to the bidding?

Jacob Landis: Yeah, so I’ve done a lot of testing and I think that’s one of the biggest things is, you know, go through $2 million of our data, and like research does the learning phase make any difference?

Not really, I did that for bids, as well. And found the best results with cost cap, that that’s kind of the sweet spot. Because cost cap, you kind of set the mount, you want to bid. And if it’s like, hey, if we find an amazing person just a little bit over, we’ll keep your average to that.

Whereas, you know, bid cap, it’s like, if it’s one penny over what you’re bidding to get in front of that person, it won’t happen. So yeah, the nice thing with those manual bids doing bid tests is you could spend very little on a bad day and a lot more on a really good day. So then it really manages your costs a lot better.

It’s just trying to find that balance especially if you’re taking on a client, you don’t want it to come in and be like, Okay, we wanted to spend $3,000 a day and we spent $27. Yes, we did, we had lower bids.

And so is that balance of trying to get consistent results. And even everyone has their priorities. Some people would rather spend less, or have their spending kind of fluctuate but get consistent results. And some people would rather kind of spend the same each day usually, usually, I do like a 50/50 split.

So we have and then for the lowest cost campaigns, if they’re doing well, I’ll increase them and incrementally like 20% throughout the day,

Yaron Been: For the recaps, you do for the cost control, or the cost cap to be accurate. So if it’s under-delivering so I mean, doesn’t waste money, invest money.

Do you increase the bid with the duplicate and with an increase of bid or do you just leave it alone?

Jacob Landis: So generally, what I found is like a campaign doing fairly well. I don’t like doing a lot of messing with it. Besides, you know, increasing the budget by 20%, I found doesn’t reset it, or you can pause it for up to a day or decrease it.

But otherwise, I generally won’t touch things that are a little less important with cost control ones but just as a general rule, if it’s not coming in quite as much of this duplicate, create a new campaign with a higher bid, and the same ones or maybe if I noticed some patterns, I’ll swap out a thing or two but keep it more or less the same.

And just do a new campaign anytime. Something has been working. I don’t want to touch it because I feel like it’s kind of been optimized and yeah, it has this thing going, and the more you swap out ads and make all these changes, the more messes with it.

Yaron Been: Yeah, this thing with Facebook ads is such a black box. Nobody knows. So you have to base your actions based on your intuition and your experience. But you don’t know because things also might change on the back end, the algorithm has changed.

So let’s say six months ago, the bid cap worked the best. And now maybe something has changed and the lowest costs work the best. And you don’t know. So you can’t keep on rocking the same way that you used to work before you Oh, you constantly need to adjust and test also things that didn’t work in the past because they might start rocking now.

So this is kind of an issue because a lot of people are saying, Oh, the lowest cost is the best or beat cup is the best or lookalikes but it’s constantly changed. And it depends on the market and the product and the price point.

So there are a lot of variables and a lot of unknowns in this black box.

Jacob Landis: Yeah, I always run a mix simultaneously, just so we can hit some consistent and then you know, scaling it up and down based on it because if our targets $3,000 a day, on a good day, I want to be spending 6000 on a bad day, I want to be spending 500 to like test and start turning stuff down or pausing it.

Because otherwise, if you just spend the same day after day after day, no matter the results, every bad day is going to wipe out every good day.

Yaron Been: I like it. I like the idea of mixing and having a lot of stuff and just letting the algorithm optimize and spend daily of whatever is giving you the best signals I wanted to ask you in regards so we talked about the inconsistency and the volatility and the fact that tracking is kind of inaccurate these days.

What can the store owner do to optimize the tracking?

Jacob Landis: Yes, there are several things, I have a full playlist that walks through step by step of like every single thing for iOS 14. Setting up conversion API is just like additional data points additional data that it can send with some people, or even creating new pages, or like specific product pages that we just send to Facebook.

That way, we’re not running into the like, if you see a purchase come in from a page that’s only getting traffic from Facebook, you know that traffic came from Facebook, so you can way more accurately track that. You can add UTM or like URL parameters that just say, hey, it came from Facebook. So that will be in the URL when people click on it.

And that tracks of RIA a cookie sends it to Google Analytics, I found that’s a little bit more accurate, or at least a useful data point to have in there. That data. There are some retargeting audiences that I haven’t heard anyone talk about I, I did find a way to have a retargeting, retargeting people who have opted out of being tracked. So basically what it is, you can retarget based on people’s actions on Facebook and Instagram.

So if someone watched a portion of a video or something like that, for Facebook, the really good one is you can retarget people who have clicked on a link. So we’ve switched over to where every single link goes to the website on a really valuable page. And because clicking on the link was behavior on Facebook, you can retarget that person. So retargeting them, it’s almost like retargeting people who visited the webpage.

Sometimes we even set up, you can do videos on the different products for people who clicked on a link. So it’s like, okay, this person visited the website. Then we show a video, let’s say for a specific product, then you can retarget those people who watched at least 50% of the video, and then you can send them to additional ones on the video chain. So you’re most creating these chains that people who continue to watch videos on a specific product.

I’m hoping Facebook rolls out a way to have multiple things where it can be like they watch half of this video, and they click the link to the website. I’ve messed with that a long time. And you can only do one or the other.

Yaron Been: But if you create two audiences, and you create like, in the targeting, and you include and you do two layers of like, an end operand. So you can create this.

I’m just thinking because I’ve never tried this, but I think because you can create a custom audience of people who viewed like 50% and a custom audience of people who clicked, and then if you layer them, one on top of each other in the targeting, maybe it’s a combination of people who did both. Am I correct?

Jacob Landis: Yeah, I messed with it and couldn’t get it. Because I want that audience to only be people who recently clicked on a link and watch a portion of the next video.

So you can continue to show stuff to people who clicked on a length and watch something, it’s just, I mean, you should still do the same retargeting audiences, still, half of the people are in that I generally like to kind of create these chains where people continue to get stuff because it’s always like, okay, trying to get into people’s minds, like there’s some reason that they keep coming back to this website, when not making a purchase.

So like, we interview clients and customers and look compared to the competition and figure out, okay, for some people, the price, you know, they don’t know if at all, pay often, be a good investment, or do what it says or like quality or other things like that.

So usually we do like testimonials or things where they get like this chain of different things for basically a busting every single objection that people ever have to the product. And you can kind of sort of do that with retargeting this way. But it’s a lot easier when it’s like people added to the cart, and now you can target them with specific to that, they initiated checkout retarget them with that you can do that with people who have opted out with iOS 14.

And you mentioned the UTM. And I think it’s very important because some people overlooked this, but for me, I think the fundamental part of tracking and doesn’t make sense not to have any UTM.

If you’re a real marketer, I guess even every store owner should have like a UTM for every traffic channel every medium. Not necessarily for everyone, for every campaign, but at least for every medium and source.

They agree. I do absolutely. I mean, there are programs like high ROAs, which we’ve implemented with some clients where I feel like it does a better job of tracking. But basically, you can do that same thing yourself.

They’re using those ATMs, they’re making it pretty automated, they’re making it so you can see the performance of each ad, I mean, it brings it into that platform. So it does have value. But some of these are like 1000s of dollars per month, you have to be spending a lot for it to be fully worth it.

But you can build your own. I’m trying to build my system where you can drill down to the like the ad set level ad level, it’s just kind of like including that or tagging that in the UTM. But even just having source equals Facebook, just doing that is going to give you a big data point, because people in their advertising, all of a sudden results are reporting correctly, or seems like Facebook’s off.

And they’re like if you don’t have that data in Google Analytics, it’s a very important data point, you’ll be missing when you’re creating your ad, your URL parameters source equals Facebook, just that is going to make a big difference.