Advanced Amazon PPC Bid Optimization

Advanced Amazon PPC Bid Optimization Speaker 2: What's going on, Vagination? Welcome to The PPC Den Podcast. I'm your host, Michael Erickson Facchin, and you are in for a treat today. I can't say enough good things about the woman who's going to be on our show today. And I can't say enough good things about the share that she's going to be sharing. If you are a fan of bid optimization, and I know you are, if you are a fan of intentionality in your campaigns, and I know you are, if you know that you should be bidding differently for competitor terms, as opposed to your own branded terms, You're in for a treat. The woman on the show was a longtime listener of the PPC Den podcast. I've known her for years and I've watched such an amazing growth in her and her agency and the way that she approaches bid optimization. So it's been a while since I've had an episode about advanced bid optimization, but this is it. We're going to be getting into it today. Unknown Speaker: Here we go. Speaker 2: Priscilla, I'm so excited to have you here and share the topic you're about to share. I've known you now for many years now. Do you remember how we met? Speaker 1: It was at iBadger. Speaker 2: That's right. I've gotten to know you fairly well over the years. We mastermind together almost every Wednesday in core, and I just have to say some of your shares are, I pay very close attention whenever you're sharing anything. I've learned so much from you over the years, so I'm so happy to have you here and feature you on The PPC Den podcast. Thanks for being here. Speaker 1: Yeah, and likewise. I listened to your podcast when I knew nothing. So this is very full, full circle moment for me. Speaker 2: And I think it's so, I remember many years ago, you sharing a, it was almost like the beginnings of what we were going to talk about today, which is, I think I'm calling it like advanced bid optimization because I can, I know, I can tell you, I talk to a lot of people and there's different, one of my favorite episodes I've ever done of this show was like the seven bidding personality types of like how people interact with their bids. And it's like, you know, I gave everyone like names of like, you know, a neglectful Neil who like never touches his bids or anxious Andy who like freaks out and like will increase a hundred percent or decrease a hundred percent. Right. You know, all of these bidding personalities. And I think that you are going to introduce a new bidding philosophy to the good people out there in Badger Nation, which is this sort of, let's just get into it, this sort of like bidding protocol that is among some of the most advanced that I've seen. I spent a lot of time thinking about bid optimization. I wrote our bid algorithm. I've taught other people how to like think about bidding too. And what I think you're about to share is among the most advanced Of everything. So I think it's really cool what you're going to share and I think people should listen to this in 1x because there's going to be so many interesting points here to this. But before we get into it, talk to us about how this came about. Doing bidding Is a big umbrella of tasks, but then choosing to create your own sort of internal algorithm that you deploy is its own other thing. So like talk to me about like your bidding journey to get us like take us on the journey from like never doing bidding to starting to do bidding to developing your own bidding protocol. Like what was that process like? Why did you go on this journey? Speaker 1: Yeah, so I, as you mentioned, I reached out Years ago, needing an automation tool, actually, but before I reached out for an automation tool, I was doing bid automation very manually. And so I'd listen to your podcast, research peel and stick method, like all the things that you recommended. I was implementing, but it was very manual. And one day I hired this virtual assistant who lasted about two weeks, to be honest, but she gifted me the greatest thing. And I asked her why she was, you know, taking more time than needed to update some of my very manual tasks. And she says, well, I had another client that had an automation for it. And she sends me, I was like, well, can you send it to me? She sends me this massive file with macros that would print this thick. And I decided to print it out. And study it for the next month, nine months. So it was like having a baby. I studied it and I think it opened up my mind to all the different if this then that combinations, stacking, And I mean, even people who are very experienced in Excel can get themselves into a rut if you put too many factors together all at once. It's confusing. There's a lot of parentheses. But I started to open my eyes to automate that bidding structure. And then I tried out AdBadger, but I wanted to see what was happening because then my wheels started spinning. I started thinking, OK, but what about this and this and this? And so I started to then just I built it in Excel and had many different classifications for bids, for targets, for performance, and really built this infrastructure where I can, instead of having static parameters that one threshold might not work for another, I created a system where I have a moving target. Especially when you consider events and you consider growth of the brand, you know, a 10 click threshold won't work today like it will tomorrow if things change, which everything is changing. That's what I'd like to talk about today because it's been such a good hack in our system. And I think it'll bring, it'll help people avoid a downward spiral of always trying to meet your ACOS metrics. And if you continue to lower your ACoS, you're going to lower your visibility. That was like the biggest thing. I was like, wait a minute. This is great. I've got a good ACoS, but my overall sales are crack, you know, so. This really helps. Speaker 2: You know, over the years, one of the ways that I continue to try to improve the Bids by Badger bid algorithm is I'll often talk to people. We'll often like look at an Amazon account together and I'll ask them, what would you do in the situation and why? Some of the most insightful experiences there were like, oh, I would look at Inventory levels, or, oh, I would look at placement data, or, oh, I would look at, like, what kind of match type is this, or, oh, I would look at, like, I would also want to see, like, last 7, 14, 30 days of data here. Oh, I would want to see, like, what product is appearing for. I'd want to see, like, you know, the organic sales. Like, all of these extra things that are not visible in the default Amazon ad console, all of these sort of, like, calculated fields, all of these, like, classifications of what the Keyword type is, is it close, loose, complement, substitutes? Is it expanded, ascent, targeting, all this stuff? You bring in that and you can almost, if you begin to see the matrix almost, if you begin to see all these factors influencing what it is that you're looking at, you can begin to get really deep and really nuanced in bidding as opposed to just simple like, good A cause bid more, bad A cause bid down. Because you can get into those, you know, you're sort of missing something if you're only doing it that way. It's really fascinating because like not every click is created equal, like not every order is created equal. There's a different value in A competitor click or like your branded click or your hero keywords or your, you know, all of these different things, like not everything is created equal. It sounds like the first place to begin is sort of the classification system. So talk to us a little bit. So let's put this on screen and thank you so much for sharing this and inspiring us, inspiring me to begin to think about The classification of a lot of these entities, of these things that you can bid on, right? So we've got a whole bunch of abbreviations here. I'm going to try to make this a little bit bigger here. Okay, so walk us through what it is that we're looking at here, because this is fascinating. Speaker 1: This is a very simplified approach. The first abbreviation is H for hero. So let's say you have a very high volume account. This hero may be anything over zero orders or it could be anything over 50 orders. You'll have to kind of cater it to what your volume is. But your heroes are... Yes. Speaker 2: Because you just brought up an exceptional point and you mentioned it previously. I wanted to highlight that again because you just mentioned it, which is you customize this per company that you work with, meaning you have a different definition for how you would define a hero for brand A versus brand B, which I love so much because I tell this to people all the time. Two companies, even sometimes selling the same exact thing, need different strategies just based on their life cycle, based off their brand awareness, based off all these other things. Amazon marketing is very hard for that reason. You can take some extremely similar products that need completely different strategies, which is something that really only comes with time. I just wanted to also highlight that too, the wisdom that I didn't want to skip over that. So that's amazing. Speaker 1: Yeah. And a good way to identify that is usually the Pareto principle 80-20 rule. So you can look at what 20% of your campaigns That make up 80% of the sales, the paid sales, and look at what that order threshold, that minimum is. And that could be a good way to say, okay, your heroes are anything over X orders because that makes up the majority. And you could cater it. I think that logic simplified is not really reflecting Right. There's a lot of things missing there. So this is definitely not everything, but same could be that it's got orders, but we don't have enough clicks or we don't click through rate and the conversion rates are not what they should be. So they're not at that threshold. To be able to say, okay, this is a hero product. And one of the things I do for my hero products, a lot of people will keep that bid the same. I actually increase it. I get a little bit more competitive, especially since supposedly Amazon's not supposed to. Maybe I'm the reason why CPCs are increasing. I'll push it up a little bit. Another really cool category that stumped me for a while is high volume, unprofitable. And again, the logic here is very simple. There's a lot more to it, but if your ACoS is like, A hundred plus, and you're spending a lot more than you're generating, but it makes up for, and sometimes I could do it as a percentage of your total paid sales, like maybe two to 5% of your total paid orders. That's something to keep in mind and see where there's efficiencies, but not negate it right away. Speaker 2: I wanted to pause here really quick, because you just said something that's Also part of the magic of this is that you have this simplified column over here of logic, which is like high volume, unprofitable orders, orders greater than zero, spend greater than or equal to 15. That's not the actual formula you use. You're actually customizing this. So you would say like you would have a formula instead of like spend over 15, it would be like spend over 1% or 2% of the total spend, right? You're using a lot of other metrics to inform your classification system here. So I imagine the formula that you are using in that column where you're... Because the process here is what? You download a bulk file or some file from Amazon, and then you have this sort of classification column that you drop in. And that classification column is dynamic. It's not Exactly this, but this is sort of the spirit in that column over there, right? Speaker 1: Exactly. Yep. And it becomes a moving target when you do a percentage of total for that timeframe. Speaker 2: Because it changes every week. Speaker 1: Right. So a percentage of your total for that timeframe could be really high one week and a campaign just popped into that new bracket. High volume, unprofitable. So how are we going to deal with it? And then the others are pretty explanatory. Bleeder, extreme bleeder, unprofitable, bleeding clicks. One thing I like is good click-through rate with low clicks. And good click-through rate with high clicks. So there's a lot of potential in the good click-through rate with high clicks, but why haven't we generated a sale yet? And so I know you'll probably introduce this, but part of the dynamic bidding is that I don't just keep it at 10 clicks. I look at what the average metrics are. How many clicks do I need to gather in order to generate a sale per match type? So within these categories, we're looking at for a good click-through rate, Low clicks or good click through eye clicks. I'm looking at every single match type in that category and saying... Okay, for broad, to generate a sale, I'm at 20 clicks, actually. So let's maybe put a 20% overage threshold. And once we get to that point, then I can determine whether this needs to shift or how the bid needs to move. Speaker 2: Every one of these classifications, there's more to classify here. And I think that's part of like the thing that There's a huge takeaway here, which is, again, if you only look in the Amazon ad console and you only look at what the keyword did in the last 30 days with general metrics that people see, without a classification system, without consideration to a lot of outside other things, like what the match type is, is it a good click-through rate, low clicks type thing, without being relative to everything else that the account is doing, Like all of these things are informing the decision-making, which is amazing. You classify things, and this is only the classification process. As you classify things, we can imagine that competitor targeting is probably the toughest, maybe one of the lower conversion rates, probably highest clicks needed to make a sale, probably one of the highest spend needed to make a sale. Speaker 1: Exactly. And to your point, like not every click is the same. The weight of a one Competitor click is different and it's sort of needed. There's a balance there, which that's a whole other conversation. But phrase, you've got nine clicks to produce a sale, but then you've got 29 clicks for, I probably should have shaded this so it's easier to match, but low bid auto. 29 clicks, LBA, that's low bid auto. So we don't really increase the bids over 15 cents or so. And again, that could be variable depending on your client, but we're going the very slow and slow approach there. And then, you know, loose match is you need 26 Clicks to make a sale. Factoring all this allows you to have that moving target and not stick within a certain click threshold and a cost threshold without balancing out how your account is making money. Speaker 2: So there's a couple of things that I want to just ask at this point. So this is fascinating that every match type has been classified, like clicks needed to make a sale, spend needed to make a sale, orders, units, so on and so forth. So you have this classification, which is It's sort of different than this classification. So you have like the performance classification of like what a bleeder is or top unprofitable or hero. You have this classification. And then you also cross this with match type. Are you saying that when you're doing your bid optimization, you're factoring both of these things in into separate columns that are doing this sort of advanced mathematics to come up with, okay, well, if my low bid auto is doing this, I can look inside that and cross it with some of these things and that will influence the ultimate bid that I set for any particular keyword or target. Speaker 1: Exactly. The only tough part is that Excel will slow down a lot if I make these dynamics. So that is the one challenge for me being an Excel girl, not a software guru, is to identify how we can streamline this. Speaker 2: You need a faster computer. That's the answer. Speaker 1: Black Friday's coming up. That is how we do it. And then we blend it. So there are many different scenarios that come up within the performance categories to address these different match type criteria. Speaker 2: Great. And also shout out because you hear sometimes like, oh, this Google sheet is too big. It like is crashing my browser or oh, this Excel file is too big. It's like crashing Excel. We are digital marketers and like I think digital marketers need faster computers, just generally. That's a caveat of mine. Speaker 1: The question is, should I get a Mac or should I get a PC? Speaker 2: I'm an Apple Android user. Speaker 1: Oh, that's rare. I'm a PC iPhone user. Speaker 2: There you go. So you have these cross sections and if anything, I want people to be inspired by a couple of things here. Number one, if you are just optimizing your bids using the default interface, Look at what advanced marketers are doing. Like, you know what I mean? Like, I hope that everyone sort of increases their pacing here of how they think about bidding. Now, I also think it's sort of like a stair-step approach, meaning like if you're starting bid optimization, this might be a difficult place to begin with. But at the same time, it's like you want to be thoughtful when you do bidding. You want to be intentional when you do bidding. That's what I see when I look at this. I'm like incredibly intentional bidding. Like you're looking at The true depth of everything, you know, it's like you are, what do they do in the South Pole? They like drill a hole down, like into the ice, like all the way to the bottom of the earth or something. And like, you're going so deep into every entity and you're able to learn so much about every single entity that you're looking at because you're doing this cross-sectional analysis. Speaker 1: I mean, it's just a matter of being more database and I, you know, I get my MBA. I was the nerd in the front of the class who soaked in all the information. And I feel like most other people were just doing it to get the paper. And I learned a lot. I mean, even my forecasting I use from my statistics class, my regression forecast analysis, because it's all database. I think it's because I don't put enough faith in my gut. There's always so much to consider and although I am building these what-if scenarios, I think that this leads into this next slide where I look at summaries of my bid updates and it helps me understand, am I moving in the right direction? Speaker 2: So walk us through what we're looking at here. Speaker 1: So this is segmented at the top by ad type, sponsor products, then sponsor brand, sponsor brand video, sponsor display. And so you could see for this particular client, like we've got everything broken out by match type so that I can see how many targets do we have per match type. And what the spend, sales, ACoS, and I honestly don't really look at ACoS as much. I look more at, is my bid going in the right direction by that category? And actually one other display I probably should have shared with you is by performance category. So if I have a hero category, how much is my bid increasing or decreasing? And that's really helpful. Speaker 2: You have another one of these summary sheets for the performance category. Speaker 1: Yeah. And it'll blend any match type. And I do break it out by ad type, so sponsored products, all my heroes, what direction is my bid going, by how much is my bid increasing on average for all my heroes in the sponsored product ad type. So really diving into the data helps me to understand, okay, so my average bid change, I guess in this instance for like broad plus, that first line, I increase my bid by an average of 2%, which is... Speaker 2: I see the bid change column percentage, you have it at 2%. Speaker 1: Right. Usually I have the amount, and I guess I don't have it in this particular dashboard, and I have my target CPC, expected value per click. There are other metrics that we use to gauge success or not based on the current cost per click. But it helps me understand if I'm heading in the right direction. If I see, oh man, I've got a conversion rate of 10%, why did I decrease that category? And so then I'll go in and update and whatnot. Or maybe I need to be a little bit more conservative because overall my traffic has been declining. Maybe I need to be more aggressive or less aggressive in my bid decreases because my overall traffic is kind of not heading in the right direction or whatnot. There's a million scenarios. Speaker 2: Awesome. So I want to click into some of this here. So if we just take that first row there, ad type, broad match modifier, that's broad plus. It'll tell you how many you have, how much you spent. How many sales generated, what the ACoS is, what the ROAS is, how many impressions, clicks, orders, click-through rate, conversion rate, CPC, and then you have like the old bid, and then you have a new bid. So you did a 2% bid change, meaning it's capped at 2%. Like that, in the logic, it will never increase something more than 2% for that category. Is that how I'm reading that? Speaker 1: It can exceed 2%. So what the bid cap does is shows us that based on a number of parameters, we're not going to increase by the amount the formula suggested if it exceeds a certain amount of total change between old and new. This one seems very, very small. So it might just be because that particular category of broad plus may be in the performance of You know, unprofitable. And so if it's in an unprofitable category, some may have increased, others may have decreased. And so that's why I wish I would have shown you the performance. Maybe I could share it with you after. But by performance type, that'll be a better indicator of for all campaigns within this particular ad type, you're moving in this direction or that direction. Speaker 2: It also looks like every bid change percentage is a positive number. Why is that? Speaker 1: Because probably most of the campaigns are heading in the positive direction. So there may be within this, like for example, let's look at something with zero. For product targeting expanded, that one's probably not doing well, which is why the bid change says zero. And so you probably have more campaigns or targets that are underperforming than are performing well, or maybe it's 50-50. That's why that's zero. But if we were to look at this at a performance level and look at all the heroes, it would show you that average direction we're heading in and by how much and by what percentage. Speaker 2: Let me ask a question about something interesting here. Sponsored brands, product targeting expanded. I'm only pointing this out because it's 47% ACOS and I'm curious how you think about this. If I look at product targeting expanded, there's a lot of spend there. $1,000 is the second highest spend category. It's also the 47% A-cost. It's one of the higher A-costs. And if I follow that through, I see a 4.1% conversion rate, which is above, below the average for sponsor brands. And it looks like the CPC $0.37, old bid $0.88, and then it was increased. But then I also see paused targets. Does that mean some things were paused? Target CPC is 19 cents. Talk us through that row, how you think about that. Speaker 1: Yeah, so if the bid is increased, then there's probably a couple of sponsor brand product target expanded that are performing well. So that's shifting up the average. I categorize my campaigns by campaign purpose. So if I have a brand that is very well known, There's quite a bit of spend that we're going to need to allocate to brand defense, right? Those should operate at a better conversion rate and click-through rate because it's the brand. In this example, there's nothing there, but just giving you an example. And there's competitive. How much growth does that client want to see? And given those directives, I know whether to push and not to be intimidated by higher aid costs if the intention is to be more aggressive. So, then I could tweak my bids based on that. There's cross-promotion. Discovery has the most amount of data here on this specific screenshot. So, for example, you can see Discovery. This is before I had good click-through rate, low clicks, and good click-through rate, high clicks. So, right here, I only have low click-through rate under Discovery. I had zero orders on it, and my bids were telling me to decrease by 68%. So this is why these snapshots work for me because I can then go in and say, that's too drastic or that's not enough. And so knowing the clicks and then incorporating my moving target parameters is why that was helpful. So I don't decrease the bid just because it's a low click-through rate. I can determine the performance category and then also the campaign purpose, organic rate ranking. If there's a set of campaigns that we're really pushing organic ranking, then yeah, the ACoS is going to be a lot higher for this particular one. The underperforming, there's one target, it generated one order, $84 in spend for $122 in sales. That's a 69% ACoS. That would be poor performing. But I'm not going to decrease the bid drastically if I'm really trying to push organic ranking. You could see the average bid adjustment was only 2%, right? So I'm not pulling the trigger and going, ah, you know, things are not working exactly how they want. Let's give the data some time and make database decisions. Speaker 2: Amazing. So you've also mentioned throughout this show that there's like, oh, I in a new version, I did this in a new version. I did this. What version number do you think you're on? Speaker 1: I've lost count. I've lost count. I wish I was more organized than I sound, but I used to keep track of all my version updates so that and I still do to a certain extent. I just feel like they're kind of sporadic and this is when I updated it to be more aggressive. But what I determined was it's very different by client. So if I have a bid optimization template for one client, I'm going to have a change log. It's going to be different from another one. Speaker 2: I think the big takeaway that we've really hit very well on this episode is that there is a huge depth to bid optimization. And I think you're really highlighting a lot of this here, where you have, on the surface, it's just the metrics in the ad console. And then you get a little bit deeper. And I love, like, these are concepts that I talk a lot about on the show, which is like the intentionality of any campaign that you're running, where it's like, why am I running this? And I think these are things that people intuitively know when they sort of like transition from one campaign doing everything for me to thinking, hmm, maybe I should move my competitor keywords elsewhere so that they don't crowd my normal terms. Oh, hmm, maybe I should transfer my branded terms elsewhere so they don't like crowd my Organic ranking terms, like you begin to segment for all these different purposes, right? I think the thing that your model really captures well is a competitor sale is not worth the same as a branded sale. And you should consider those things differently. Meaning, if I were to say, hey, $10,000 of competitor terms of revenue, how much would you pay for that? How much would you pay to grab $10,000 of revenue from a competitor, bidding on competitor terms and competitor products? That's worth something way different than you would never want the same ACoS for competitor branded terms. As your own branded terms, you should have very different performance expectations there. And such, you should have very different bidding protocols for each of those things. And I love, I hope that it's inspired anyone listening to the show to sort of think and pause about how they think about their bidding, maybe tag or identify campaign purpose or individual keyword or target purpose. Because I think you will end up with something very rich And I think the other thing that's also hidden in this whole analysis is like, you're customizing this per market. You're customizing this per each one of your clients, which I think is also very valuable because different clients will have different ideas for how much they want to spend on their goals. Like, do they want to do a lot of conquesting? Do they want to go after competitors? Sometimes I talk to people and they say, Hey, there's a competitor that I really want to get after. So like I really want to invest in competitor conquesting and specifically this competitor. So like they'll have a very specific goal or I'll talk to someone, Hey, I'm spending $200,000 a month on my advertising. 67% of it is branded. I need to pull back on branded. So coming up with different ideas about all of this and what the implications are is super valuable. You also wrote somewhere, you incorporate stock levels as well. So all of these different things can be incorporated in the file, in the Excel file for That. So like that will also influence. So I take it like the bidding formula will look something like if running out of stocks, like if stock or days left is below X, then apply some kind of bidding discount where you want to bid less on it or something like that. Speaker 1: Right. We'll do kind of like a soft pause, which sometimes we will pause the target and sometimes we will just kind of lower the bid so we don't take it out of the game. Speaker 2: Take it out of the game. I want to ask you just some other logistical questions here. What file are you using to get all the data? Are you using a bulk file? Speaker 1: Yes, I use a bulk file and all the search term reports. So first, it would be search term for sponsor products, sponsor brands, and then I think the match type for sponsor display. Speaker 2: How do you think about Placements, top of search, rest of search, product page, B2B, AMC now. Speaker 1: I haven't delved into B2B just yet, although that is on my radar. But placements, when it came out, when I think more of the parameters came out to rest of search and product page, I was only able to really look at it from the hero performance category, right? I wasn't going to change placements and automate placements based on any other category. I've had something really well, I actually developed another criteria list where for heroes, if placements were top of search were X amount of sales compared to the others at X amount of ACoS. I would formulate my bids to decrease the bids so that placements could increase or I would just bump up the placement or I'd zero it out. So I have a bunch of those hypotheticals that I wrote out and I formulated into the bids. Now, we still do more manual placement updates and review, but at least for my hero products, I can inch up or back down depending on performance, and that's kind of written into the formulas. Speaker 2: When you say manual placement analysis and manual placement updates, what do you mean? Like you're going into a campaign, you're looking at what it's doing, and you're making a decision there to increase or decrease? Speaker 1: That's also... You know, now that I think about it, we don't have a summary for placements and we should be looking at that so we could fine tune our formula. So right now we're just kind of looking, this is what the update was, and then we'll go in and just kind of visually see whether, we'll look for the red flags. We're not checking every campaign, but we'll look for the red flags, go in and adjust as needed. Speaker 2: How often are you doing this like how often are you repeating it and what's your look-back window? Typically, like are you doing this weekly you download a 30-day report every time you're operating? Speaker 1: I do last 14 days minus the previous two days. For clients with really high volume, I'll only do a week. I'll do the last week, which I know is probably a little controversial, but I think it just, it depends on your consideration window. Like we had mentioned, if you have a lower price point, high volume, and you can identify from your clicks needed to make a sale that you're generating enough data that I'm going to want to keep it very, very short to make those decisions. Speaker 2: These formulas sound very long and complicated and I think the longest formula I ever made was we had like an Ngram spreadsheet and this was in the days like before AI assisted formula building. So I'm curious how you are writing your formulas these days. Are you writing them yourself? Are you using AI to assist in Sort of encapsulating a lot of these, like this metric over here, if it's this metric, if it's this threshold, if it's over here. Are you using AI at all in the formulas? Speaker 1: Not as of late because I routed my formulas to pull from a driver tab. And so if I need my driver tab to change, I just update that and it goes. But I do in other spreadsheets have written them with ChatGPT. If I'm not familiar with the formula, I'll Kind of let ChatGPT lead the way, but because they're very specific places that I know exactly what I need to update. But yeah, I mean, the sky's the limit. I just think I usually segment out my formulas, which I think anyone should. If you're trying to, you shouldn't build a whole formula that's this thick right away. You should say, okay, what's the parameter for here? What's the parameter for here? And then build it after the fact, once you know that those details are correct. Speaker 2: Got it. So you had, we saw some examples in your spreadsheet where you have many columns and then, you know, you calculate the first thing, that's a column. Then the second thing, that's another column. Instead of one gigantic thing that can be untraceable, like a big bowl of spaghetti. Yeah. Well, this was Exceptional. I think if anyone is not inspired to think about bidding differently, you've got to find a new profession. This is amazing because this is what amazing bid optimization looks like. Every time I talk to you, every week where we mastermind together, I'm always learning something from you. And what a treat to have you on the show and share so much insight. Priscilla, thanks so much. We have a link to your Brand new business website in our description. People can contact you. They should contact you. What do you think your next update to your bidding formula will be? Like, what do you inspire to sort of, oh, I want to do this or, oh, I want to do that. What's your next iteration? What's your next version? Speaker 1: I think B2B has gotten a lot of talk. I still don't think it's large enough, at least for the clients that I have, to formulate. And I think some of these audience levers that Amazon has put out, we're testing, but we haven't necessarily built that in. So we're still at the target level versus target audience type. We haven't gone down that road yet. I think that would be probably next and really looking at that data because we have it by match type and we haven't gone down the audience. AMC is also very uncharted territory, which I know everyone's talking about AMC. I just don't have enough clients that warrant AMC just yet and so we'll see. Speaker 2: Well, Priscilla, thank you so much for joining us today on The PPC Den Podcast. Everyone else, I'll see you next time here on The PPC Den Podcast. Thanks, Priscilla. Unknown Speaker: And picked keywords. I've got my bids. Set placements too. Now bad mistakes, I've made a few. I've had my share of wrong keywords. Hello. The PPC Den, my friends. And we'll keep on the music. You are the PPC. Time for medicals, cause we fixed the game, baby.