#104 - The Secret to Maximum Efficiency Bid Optimizations on Amazon PPC: Macro vs Micro Bid Changes

#104 - The Secret to Maximum Efficiency Bid Optimizations on Amazon PPC: Macro vs Micro Bid Changes Speaker 1: Today, we're talking about a bid strategy most of you don't know about. Speaker 2: This is a total game changer for thinking through PPC performance and getting substantially better results. Speaker 1: We're excited to help you level up your Amazon PPC game with our overview of macro and micro bid changes. Let's dive in. Alexa, play That Amazon Ads Podcast. Unknown Speaker: Which one would you like to hear? Speaker 1: The best one. Unknown Speaker: Okay, now playing that Amazon ads podcast. These gentlemen are completely changing the game. Speaker 2: After listening to that Amazon ads podcast, my ads are finally profitable. Unknown Speaker: I also heard they're pretty cute. Speaker 2: All right, Stephen, what are we talking about? What's micro and what's macro bid changes? Speaker 1: Yeah, so this is something that during a lot of the onboarding and training calls that I've been doing with folks, I just realized this is such a valuable framework to think through your bid optimizations because Too many people are fixed and rigid with how they optimize. It's like, it's always the last 30 days. It's always going to be these types of adjustments. And it's just, it's not dynamic enough. And there's a big difference between, you know, those kind of routine maintenance changes and those we were calling the macro bid changes where you're kind of using longer timeframes. It's your whole entire account. And, you know, it's best for like routine maintenance. The micro bid changes are usually smaller adjustments. They're quite a bit more reactionary. They tend to be a bit more drastic, but we're going to break each of those down and talk about when and why you're going to want to use each one. But before we kind of dive into all of that, I just wanted to reference a few other episodes that I think could be helpful, just some related topics. Episode 35, we talked about date range selection. So that's a great primer just on thinking through date ranges. Should you use seven days, 14 days, 30 days, 60, 90? So a lot of people just might need a refresher on that. And also optimization frequency, which is episode 47, which is like, how often should I be optimizing? Once a week, twice a week, once a month, twice a month. So all that kind of stuff. There's also a question that we've been getting, which is, After I optimize, how long do I need to wait before making the next optimization? That's a topic we will have to cover separately. We don't have a whole episode on that yet, but it's kind of related to today's topic. And finally, episode 97. Which is on the art and science of Amazon PPC, where we really try to emphasize a lot of the art side of things. And what goes into that art of Amazon is not, it's not rigid stuff, right? It's not super scientific, but what's the art is like date range selection, optimization frequency, how long you wait before making optimizations, how many campaigns you should be optimizing, all your campaigns, some of your campaigns. So that's kind of all that's related, but Let's just start talking about these topics. Macro, bid changes. These are the full account optimizations. Andrew, what kind of usually goes into that? Speaker 2: I would be looking at using longer timeframes because we want to have a high amount of data confidence in those large sweeping changes and adjustments that we're actually doing. So we want longer timeframes to bring in more data to all those different calculations. This is really important for things like campaign placements. You don't want to use a really small seven day look back window for optimizing placements. You'd much rather use a larger timeframe because you've got a lot more confidence going into those calculations, figuring out how much you should be upbidding certain placements. And if you look at a very short date range, those changes might be a lot different than what you're looking at on a longer timeframe. So you want a high amount of data confidence and really start to use those longer timeframes whenever you're making those big sweeping changes across the account. Speaker 1: Great. So let's pull open a screen share and we can take a look. So over on, uh, here we've got AdLabs pulled up. So we've talked about this before. I'm just going to go to, let's do year to date. Let's do, let's pick full weeks. So on the date range selection episode, we talked about how it's usually a good idea to start a little bit more zoomed out and then Let's get to a weekly view. So we've got spend, sales, a cost conversion rates, right? So actually, this is a really good example. So on the date range episode, we talked about like, Longer timeframes are going to give you more data, more data confidence, just more overall clicks and sales to go off of. Shorter timeframes are going to give you better data recency, relevancy. So if you're seeing from here, for example, that conversion rates have really started picking up since April, that would be a good indicator that the conversion rates For the last bit have been above average. So you probably want to factor that in that conversion rates are on the rise. You would want to include that in your bid optimizations. Let's also look at ACoS. So yeah, ACoS is coming down. So in this case, you know, if I'm trying to optimize this account on doing a macro change, I definitely want to do a longer date range, but I still want to have a relevant timeframe. So probably starting from like April 13th, for example, so we can go from, let's come here to April 13th. And so that's about a month and a half of data, right? So if you look at this month and a half, conversion rates are up 26%, okay? And ACOS is coming down quite a bit. Spend is up for this brand in particular. If anyone hasn't seen Sneaky Sneakers before, it's a real brand and real Amazon data. It's just we've obfuscated all of the campaign names to maintain the client confidentiality. So in this case, I can come here, I can select all of my campaigns and we can go to optimize the bids and let's just say the target day cost is 25%. Now we're optimizing everything. Uh, you can have your campaigns into like different optimization groups. You might be targeting multiple different target day costs, whatever. That's not the point here. What we do want to emphasize is on these larger changes when you're doing like 30 days or more, Because you're optimizing everything, this is 246 campaigns. I'm not sure how many targets are going to get triggered here, but it's probably going to be several thousand, maybe even like tens of thousands of keywords that are getting their bids adjusted. Because this is such a large sweeping change, You might want to just be a little bit tighter on the max increase and decreases for your bid optimizations as well as your placement optimizations. And something else is that when you are trying to optimize placements, you definitely want to use a very long time frame. So yeah, 10,000 bid adjustment opportunities here. So as you can see, we're kind of optimizing everything. But if you have some, let's see some of our max like decreases here. And let's actually filter this down to just the keyword entities. So entities, just because I want to take a look at just what's kind of going on with our keywords and product targets. So if you've got some, Actually, I'm going to also filter down for high ACOS, just to kind of get us somewhere where we can kind of prove the point. So if we are looking at, and if you're like, why are you increasing bids on high ACOS? We've got a whole episode on why you might want to increase bids on high ACOS sometime, but it comes down to math. So in these cases, these are all high ACOS keywords. And let's see how many of those actually we are reducing bids on. Delta is less than zero. So for only the bid decrease opportunities, there's 84 of those, which is a total of $3,000 of spend. I forgot to note how much total spend we were impacting and optimizing from like the total of the campaign or of the account. But you see with all these items, if we are to reduce bids on all these keywords, which are driving, you know, $9,000 in sales, let's actually just back up. So it's a relatively small impact actually. So that's okay. It's only around like 10% of the spend and things like that. It might be a little bit bigger if I do like reduce ACoS. But if everything's already good, like if things in this account are already performing pretty well, let's get to this again. And we're just looking at some bid decrease opportunities. In this case, I'm including a few other items, but $14,000 of spend out of so it's around 35% of the total account spend. If I even just like clear these filters and just look at like everything that potentially could be optimized, that's like $38,000. So something that's very important to note is. All of the adjustments that you're spending through, that you're pushing through on high spend keywords, low visibility keywords, everything, you're increasing, you're decreasing, you're making a lot of changes. I could just push through all 10,000 changes right now and because this is such a large amount of the total account spent, right? Yeah, 41K, so 38,000 total versus the 41, that's 93% of the account that's being optimized, okay? If I had max increase set to 50% and max decrease set to 50% and these set to 100% and 100%, like, That's a lot of volatility that you're introducing into the account and that could be very drastic for performance, especially when you're using multi-keyword campaigns, auto campaigns, broad phrase match. There's a lot of things that change just in terms of the performance, the search terms that you're getting, the placements that you're getting when you make big changes on those keywords and placements and everything like that. It's totally fine to optimize the whole entire account. You know, we're going to push through optimizations on 90% or more of the total account spend. But in those cases, you want to do tighter adjustments. Now, Andrew, for you, you know, this max increase, decrease, what are you normally doing just like kind of as a default on the macro side of things? Speaker 2: Yeah, whenever I'm looking at the longer timeframes, usually between 10 and 15% on the bid adjustments is kind of where I like to stay, depending on what I'm trying to do. And then yeah, the placements, I usually just kind of leave those as default. But I always have in the back of my mind and have done a little bit of digging into the placement performance of the different things that I'm optimizing. And so I have an idea of about where things should be on average. So that way I can kind of if things aren't matching up, I can kind of jump in and adjust some things if needed. So that's usually where I'm sitting in terms of like defaults and in a general sense. And yeah, typical ones that I'm working with. Speaker 1: Yeah. And this is, by the way, what your placements should look like. So if your target day cost is, say, 25%, you should be hitting 25% across the board. On all of your placements and the spend allocation should be scaling with where the conversion rates are at. So if you have the lowest conversion rate on product pages, that should be the lowest amount of spend and you should still be getting a 25% ACoS there. And if top of search is your highest conversion rate, you should be spending substantially more there and be hitting pretty close to that same target ACoS. So in this case, one other thing that, you know, obviously not all of these campaigns are going to be perfectly optimized. If we drill into some of these campaigns, we're going to see, you know, this one here has a pretty high ACoS and that placement is increased. So like this campaign in particular probably needs some placement optimizations. So There's two things you can do here. Number one is if the placements overall look really dialed in, and of course you could break things out further into like looking at specific portfolios or optimization groups, but a lot of the time if the placements are pretty dialed in, you could just not touch them. And just disable that setting if you're using AdLabs, of course. I mean, if you're doing this all manually on bulk sheets, it's all the same principles, just you're doing it on bulk sheets instead of here. But you could disable that if you are going to include it, though, and if the placements are already doing well, then you might just want to tighten up some of these adjustments and constraints. Something else that I think is very important to note is The tighter the constraints, what you're getting in that, like if you're just doing 10, 15% increase, decreases, you're going to have less movement. So it's lower risk, lower reward, right? So if you're really trying to make a big impact on the ads, then you're going to want to use, you know, bigger adjustments. So if you're doing smaller adjustments, you might need to potentially increase the frequency a little bit. So I might push these changes through today with smaller constraints and then You know, later this week, if I didn't really move the needle, I might do another round of changes again, also with smaller constraints, or maybe open them up a little bit. So you just got to be thinking through all of that. But yeah, that's the macro. I think that's pretty straightforward. Most of the time when we're talking about on all of our episodes around bid optimizations, we're usually kind of assuming more of a macro approach, which are these longer timeframes, Usually never less than 14 days, probably more around the 30 day to 45 day mark. But let's start talking about these micro changes, because that's where I think things get quite a bit more interesting. And for that, I'm going to go to the switch to from May through June. And this is actually for my macro changes. This is usually a date range that I really like to use, which is, you know, at the start of a new month, I like to include all of last month and a little bit of this month. But let's just say for this account, I hope we get it. The account's doing a little too good. But let's say, let's start from May 7th, actually. So going from May 7th until June 4th, actually, we'll stay till June 3rd, because we're missing one day of data there. So let's just say our target ACoS is 20%, okay, to give us a little bit more of something that we need to react to. So if we see that ACOS is on the rise, and what really makes a good example is if you see like a big spend spike in the last few days. So people will see this all the time where, okay, I hope you guys all start listening very well because this is the thrust of the episode. And unfortunately, a lot of people drop off around the eight minute mark, which we're probably past, but this is crucial, okay? This is the most important part to listen to. So if you start seeing a spend spike that happened within the last like two, three, four, five days, whatever, and your ACoS is going up, and you need to react to that because it's causing a problem, like you see like you lost money for the last seven days, or if your PPC manager, your client is asking you what's going on the last seven days, you need to react to that. Optimizing on a 30 day timeframe is going to be too slow because where the problems arose in the last few days is not going to appear on a 30 day timeframe. So if we were to shorten this date range to, when did we see that? That spends like May 31st. Let's start from May 30th. So just looking at the 30th, so just like these five days, You see that we have ACOS on the rise, 30%. This is what we need to react to. When you're using these large timeframes, it's not going to be recent enough in terms of capitalizing on recent trends and shifts and things like that. And so this is where we really need to do something about it. Now, there's two other ways. Two other things to note and one is that we're not going to optimize all of our campaigns. That's a big no-no. When using these smaller time frames, we're trying to react. It's usually just a small handful of campaigns or keywords that caused a problem in the last few days. So that's all we need to do. We're just going to clean up a small handful of items. We're definitely not going to touch placements because we're not going to have the placement data confidence that we need on such a short time frame because You need, for all three placements, a solid number of clicks and at least one order to have reliable conversion rates for us to calculate it. When you're getting down to these really short timeframes, it's probably just not going to be there. Now, Andrew, I'm just going to keep going unless if you want to interject or anything. Speaker 2: I have something in the chamber, but keep flowing and going into this. Speaker 1: All right, I'm going to wrap up one more thought, which is just how do you, if you're not selecting all campaigns, how do you know which campaigns to choose? There's going to be two ways to do this. Now, a lot of people make the mistake of picking just, you know, the highest A cost campaigns, but let's actually throw in our percent of spend column. So you can see why this is not going to be super helpful. So this is the highest day cost campaign. Sure, we can clean that up, but that's only like 1% of spend and this one's like 0.25, 0.26% of spend. These are not going to move the needle on your account. So you definitely want to look for what are the highest spending campaigns that are, if your target ACoS, for this example, we're saying 20%, this is the highest spending campaign. It's pretty close to target. That's not the problem. This campaign and this campaign that are 28 and 30%, these are having a much larger negative impact on the account just because of, you know, that's, um, let's see, that's a thousand dollars of spend right there. One out of six is about 16% of the spend right there. These should actually be summing. We got to tell the devs to sum those up. But like these, just these four, five campaigns. So these five campaigns right here are constituting 1.7 out of six. Almost a third of the total account spend and they're averaging too high of an ACOS. So these would be the campaigns that we need to clean up and fix. And that's gonna have a much larger impact on the total performance of the account compared to if you just picked, you know, like these guys that are all like pretty low spenders cumulatively. So I'll pause there, Andrew. What are the thoughts that you have? Speaker 2: Yeah, no, that makes total sense. Like you're trying to just identify where the actual problems were that if you fix it, it's actually going to make a difference on a total level. So figuring out basically what campaigns are overshooting their target in the smaller timeframe, where was the actual problem that arose? And then you're kind of just going in and making some fine-tune adjustments to those. Now, a couple questions come up, but first things, Stephen, what are your thoughts on optimizing placements on this? Are you going to be optimizing placements on that short date range? Are you going to be excluding those, just looking at kind of pulling down those problem keywords that are within these campaigns? Speaker 1: Yeah, so I will answer that, but I'm going to say one more thing, which is that If you see a big change in spend, this is another pretty big indicator because sometimes the highest spending campaigns weren't the ones that actually created the change in a cost, right? If a cost is on the rise and we could actually maybe even do a little bit of a longer Time frame because I'm really trying to see a big change in that ACOS, which what do we get here? So yeah, this one we see, you know, 10% increase in ACOS, 8% increase in spend, drop in sales. So sure, I can sort by what's the largest spending campaign, but what sometimes is even better is to see which one had the biggest change in spend. So you can also sort by the delta, which sometimes it's going to be the same, like the high spenders are also the ones that had the biggest change in spend. But not always. So just two different ways to look at it. But coming back to that shorter date range that we had, and to your question, Andrew, no. So when we come to clean up these campaigns, we said it was these five, right? So it's one through five. So these are making up 28% of the spend. We could probably also include one more. So that's Yeah, big chunk of spend. Because it's such a small timeframe, when we come here, we are not going to optimize placements. Just because it's going to be way too small of a timeframe, the placement calculations will likely be incorrect or just not very statistically reliable or relevant. Now, When setting the max bid decrease, this is where we're going to make much larger changes, maybe 50% decrease. I actually prefer to leave it off. And the other thing that you got to keep in mind is when you're doing these reactionary changes, you usually want to be unilateral, meaning one-sided in the type of adjustment you're trying to make. You are either, if you're doing a small timeframe, it's because you're reacting to a problem. And that problem is either going to be high A costs or low spend, low sales. So you're either trying to really quickly drive up the performance, or you're really trying to pull back on some underperformance. So in this case, let's just say we're trying to really focus on getting that ACoS down. So we will go to, because we're seeing, shoot, spend spiked. It's up almost 25%. We got to react to that. So we're going to come here, optimize bids. I'm going to turn off the max bid decrease. We're going to go ahead and optimize everything. Technically, we could actually just focus on these high A costs and high spend no sales keywords because that's likely where most of the problems are. Actually, we will go ahead and do that. So we're going to come here and preview these results. Now, again. You can sometimes increase bids on high cost keywords, but in this case, the problem that we're currently experiencing is related to a high cost problem. So we really only want to focus on decrease opportunities. So we're just going to say spend is less than zero. Sorry, the delta is less than zero. Look at this. It's really not a lot of keywords, right? It's only seven keywords. And you know what? I actually had the wrong target ACOS in here. 20%, I'm going to say reduce ACOS. And let's hit this one more time. And then put that filter back on for, got to save filter here. Here we go. So most of the time, the problem keywords are going to be relatively few. So if we're trying to really just clean up the spend, you see it's Yeah, just these like top five keywords is the biggest problem. It's $300 to spend. And we can just come in and just reduce the bids on this small batch of keywords. And that's going to have a huge impact on the overall account performance. Now, you could select all of these. It's 84 in total. There's quite a few of them though that are just like relatively, you know, low spend. So those are just some smart bits things that are going on. We could fit this album for just the high spend and high cost keywords. So in this case, it's 21 keywords making $500 of spend, which that $500 of spend, we were saying like it was $6,000 of spend total in the account. So that's 8% of the account spend is averaging an ACOS of 33%. We're about to pull all that down towards 20%. And that's going to be very surgical just in the precision that we're doing here to quickly pull down the ACOS, clean it up, With minimal negative impact because we're not trying to adjust the whole account. If you try to do this to the whole entire account, it's very likely that you're going to underbid on a few keywords and you could lose way too many sales overnight. So, um, yeah, Andrew, what, um, any thing I'm missing here? Speaker 2: Yeah, no, that's a good approach. And yeah, very similar workflows in my campaigns, you know, just trying to figure out what those problems are. And usually it's just a handful of keywords that are really driving up the bad performance. And so just kind of going in surgically, adjusting those and pulling those down is going to have a bigger impact on the shaping and molding the overall performance of the account. I have one thing that's come into mind here, just on the, you know, the topic of macro versus micro. A brand that is highly seasonal. They're going into a peak season. They may want to make macro changes because it's a lot of their account. Their whole catalog sees a really big pickup in a summer timeframe, let's just say. Um, to where you would want to optimize like that whole account and make a bunch of big sweeping changes to kind of ramp things up. Do you prefer to use a longer timeframe and set a higher target ACoS or do you prefer to use a lower or a shorter timeframe and kind of keep ACoS targets the same and try to bump them up, maybe run them a couple additional times to kind of get the bids ramped up into that season? Speaker 1: Yeah, that is a good question. Um, I will answer that and also, Really quickly add one more thing to this micro thing, which is that, because on the topic of trying to increase, so let's just say like, even on a micro scale, like let's just say, looking at some of these small spending campaigns, let's pretend these campaigns here are part of your, some product launch or something like that. And you're really trying to ramp up the sales on these. Then even on a smaller time frame, you can come in here and you can be quite a bit more aggressive on the increase side of things. You can just like turn off. This max increase, the reason why a lot of times the max increase decrease, I just leave it off entirely just because you can always edit it on the preview screen, right? And there's also these bid floors where you can add in like, you know, you can say no bid should go below a dollar or something like that. That would be pretty aggressive. The dynamic ones are just, it's gonna be calculated based on the performance of the ad group, things like that. So on a smaller timeframe, You might wanna say, hey, just our low A cost, low visibility keywords, or even just the low visibility keywords. I just really wanna pull those up quickly. And let's just see what we get here. Filter out for the reasons, low visibility keywords. So on these ones, you can be increasing quite a bit. Some of them can probably be increasing substantially. So this one's going from like 11 cents to 34 cents. That seems pretty drastic, but that's just coming from those dynamic bid floors, which is just saying, hey, on average, I think we use 0.5X. So on average, you can afford 64 cents or 68 cents. I mean, 68 cents is kind of like an average bid for the account. So, you know, setting the bid to like half of that is still pretty low, but it's a much faster jump compared to this, uh, just 11 cent bid. So it's just going to help you get up there a lot faster compared to, you know, normally we're just increasing by like, you know, 5% or something like that. So Just one other thing to note on the longer timeframes, yeah, so if this brand is going into a bigger season and we still want to have some bigger impact on the account in terms of like we're really trying to ramp up the spend quickly and aggressively. So you can definitely tweak this where there's a few ways you can tweak it. Number one is you can open up the max increase. So you can make this like 50% max increase and 25% max decrease. So now you're biased towards more increases than decreases. And then kind of same with placements, you can open this up to like 50% max increase and 25% max decrease. So now everything is, you're going to still like go through and optimize everything. It's going to be a macro change. So you're doing the whole account. You could do only the increases, like when you get to the next screen, filter out for delta greater than zero and only push through increases. The problem is, if you only increase and don't decrease, Sometimes that could just increase the ACOS. So what we want to do sometimes is like you have a low conversion rate keyword. We want to reduce the spend on that. You have a high conversion rate keyword. We're going to increase the spend on that. And now your spend is roughly the same, maybe a little bit higher. And then you're just getting more sales because you shifted the spend from one keyword to the next. But just having a little bit of asymmetry And the max increase decreases can give your optimizations a bit more of a bias in the direction that you want to go. And then, I mean, final thing is like, you can literally just turn all of these things off. Like you can just turn everything off, and then just add them on the next screen. I like to do this a lot just because I want to see where things are trying to go. Like here, I mean, it looks like this account has like some top of search only strategies going on. So, you know, if you're doing that, you might just want to like exclude those campaigns, but looking at the keywords and product targets and just trying to see like where these are trying to get to. You're gonna see these high spend keywords having really big drastic cuts, right? So when you're looking at this, you might be okay with all these high spend keywords going down there. Maybe not. This is all just part of the art. Maybe you're deciding after looking at this, it's like, you know what, I don't really wanna decrease anything by like more than 50%. So that's just pretty quick, easy to clean up. You just go delta is less than, it's less than minus 50. You can just take all these things, edit those and say, set the bid decrease to, uh, decrease by 50%. And then, yeah, so now that will be the max decrease. So you can, you can further edit and fine tune things here. Um, and then on the increase side of things, you know, you're probably okay with all this. So this would be a pretty good mix of keywords where it's like, You're cutting a lot of bids by minus 50% stuff that's spending, not converting. And then you've got a lot more higher increases. So maybe you could put some caps on here where it's like, I don't really want to increase by more than like 300% or something like that. But they're all pretty, like if we just look at, these are all pretty reasonable. So I mean, yeah, I wouldn't try to cap any of these increases if I'm being growth oriented. I think everything looks pretty good. Yeah, this is a big, this is a larger like macro change. You'd still want to probably do something with these placements. Probably don't want these placements having such high jumps, but anyways, I'm just starting to ramble now. I think you guys hopefully get the point, which is, uh, longer timeframes, you're doing larger changes. You do want to have some constraints in there. And then the, on the micro side of things, it's quicker cleanups, fewer campaigns, fewer keywords, most of the time. And this account didn't really have that big of a spend spike problem, but most of the time when I see there's a big spend spike, it was like came from one or two campaigns and it was only one or two keywords in each of those campaigns. So this, this account doesn't really have that problem. But that's most of the time what it's gonna come down to is just like one problematic keyword that for whatever reason had a big spend spike. Sometimes why that could be is because maybe that keyword was bidding like $10 for whatever reason and the CPC was just $1 because you were the second highest bidder, right? We talked about this on last week's episode on can you bid against yourself? So maybe someone else suddenly started bidding $10 and your CPC just jumped to $10 overnight. So you didn't even make any changes to the bid, just you suddenly started paying a lot more on that keyword. And for that reason, the spend spiked, the ACoS spiked. And if you just quickly react to that one keyword, that could clean up the performance of the whole account. And that's just way better and safer than trying to do some drastic reduced ACoS optimizations to the whole entire account. And then you risk losing a bunch of sales overnight. Speaker 2: That's such a powerful framework for how to think through what you should be doing, how to make these changes, whether you should be doing micro, macro changes, just really got to know when to do what. I think you laid that out very, very beautifully. And for the audience, if you got some value from this, make sure you subscribe. We'll put out weekly content just like this, so you'll get notified when you're subscribed. And we'll see you next time on That Amazon Ads Podcast. Speaker 1: See ya.