What to know about Amazon’s A9 Algorithm ⎜ Seller Sessions ⎜ EP 159
Ryan Cramer: What's up, everyone? Welcome to my corner of the internet. I'm your host, Ryan Cramer, and this is Crossover Commerce, presented by PingPong Payments, the leading global payments provider helping sellers keep more of their hard earned money. Hey, everyone, happy Friday. Welcome once again to another episode of Crossover Commerce. I'm your host Ryan Cramer, and this is my corner of the internet where I bring the best and brightest Amazon and eCommerce experts in the industry, or at least I like to think so. I look up to them, and I want to make sure that you're getting the best knowledge that you can. Since time is money, getting all the information we can out of our guests on every single episode is my goal to make sure that you can apply that to your business, make it more profitable, grow it, whatever your goals and ambitions might be. But that's what eCommerce is all about, of course, and that's what Crossover Commerce is about, as well. This is episode 159. But before we get into the thick of it, Crossover Commerce is presented by PingPong Payments. No, we are not a table tennis company. We are a cross- border payments solution. What does that mean for you? You as a seller, you're paying a manufacturer distributor in a different country, which 95% of entrepreneurs are, you're going to be needing a solution that's going to save you time, money, and effort, not waiting around for currency exchange to happen. You need to pay them effectively in their localized currency, and that's where PingPong steps in, at a quick, effective, and cost- effective way. To do so, you want to talk about and use PingPong Payments, go ahead and sign up for free in the link below, or you can obviously go to pingpongx. com, and just mention Crossover Commerce. Go ahead and sign up. It's free today, and you're going to be saving tons of money to apply towards your business. Might as well go ahead and do that, but of course, after this episode, that you might be watching today or listening to us on your favorite podcast destination. So thank you, PingPong, and of course check them out for free, it's free to sign up today. So that being said, again, everyone who's watching and listening live to us on Facebook, LinkedIn, YouTube, or Twitter, we appreciate the follow and the tune in on this beautiful Friday, at least it is where I am. It is fall this week. Or it's the beginning of the fall season. Q4 is next week. What do we need to worry about today? But first and foremost, I think I want to take even a step back further from that and say what do we need to know about Amazon and how is it operating on a day- to- day basis? I think a lot of people want to think they have figured it out. There's lot of nuances about Amazon that someone may or may not necessarily know, so how does it actually operate? It's not eyeballs, it's not people that are looking around scouring the internet every single day ranking based upon what they feel is objectified or opinionated. It's actually an algorithm, and that's what we're going to be talking about today with our special guest, Danny McMillan. Danny is the founder of Seller Sessions as well as tons of other great content out there. Seller Sessions Live, which is coming up actually in the UK. I should say he is in the UK, but it is happening in London here in the next couple weeks. But then also doing things like Seller Sessions tours and polls, but also hosting a fantastic podcast, 600 plus episodes with him and his team over there, both former friends of the show. Again, friends of the show Izabela and Sharon Evans, as well. So that being said, he wanted to come on. He was gracious enough to spend some time. I know he's in the UK right now, so it's a little bit later on a Friday, so we're going to give this an action packed so much content episode that we're going to pile into as little time as possible, so you can take that information and apply it to your business today. With that being said, we want to go ahead and welcome to Crossover Commerce, Danny McMillan. Danny, welcome to Crossover Commerce.
Danny McMillan: Thank you for having me. How're you doing?
Ryan Cramer: I'm doing well, sir. Now you're in the guest chair, but as people may already recognize, you're one of the most respected thought leaders and content creators in the space in terms of consistency. But if people for some reason haven't already come across your content or information, what's that background for you as an entrepreneur, individual, how'd you get in the eCommerce space? What's that little, brief bio, if you will?
Danny McMillan: Yeah, I started selling on Amazon in 2015. In 2017, I launched Seller Sessions. We're up to, I think, about 750 episodes, and obviously brought on Izabela Hamilton and Sharon, so they do Mondays and Thursday respectively. And then I do Tuesday and Wednesday. I also run DATAbrill, so I co- founded that with Dr. Ellis Whitehead, which is a PPC technology company. We're right all there on technology et cetera. And we look after seven and eight figure sellers there. Then I've got SellerPoll, which is the yearly award for the Amazon community, so this is coming to year three. I've been mounting that too. And Branded by Women, which we do once a year to promote ladies in the space that are doing great things. And, of course, Seller Sessions Live, which is the yearly conference. And the rest I'm off on tour speaking around the world et cetera, et cetera.
Ryan Cramer: Yeah, you contacted me earlier about that tour. I'm super excited for you in that regard. You don't have any time on your schedule, my man. I don't know when you sleep. I don't know when you operate, but what's that like to have all these kinds of pokers in the fire, if you will? Is it just because you want to get content out there, or what's the drive behind it?
Danny McMillan: You do things that you love, right? So I have three rules. I actually say no to about 99% of stuff. And the way I was able to structure those nos is does it make meaning? Will or does it make me happy? And then the final one, will it make me money? They're the business ones. And normally what you find is that you can align all three of those because if it makes meaning, you'll make money in most cases. Right? If it makes you happy when you're having a shit day, you get up and you get the job done. Right? And you can't be in business unless you make money, so if you can find... Look at Branded by Women. It makes money, but it won't make the money that other events make. But we're trying to change and persuade different behaviors inside the community. So it makes meaning because we're making change. Right? And then what happens is the money comes later. But the fact is that I enjoy doing the work. I'm surrounded by brilliant women. And so everything I do is in and around those things. I'm past the age of chasing the pound note. Yeah? So for me, I don't get stimulated by chasing money. It just doesn't stimulate me, so I have to find things that stimulate me. So if something makes meaning, it makes me happy, and it makes me money, I'm not working. So every day is a vacation in some senses. Yeah, you get stressful days, but if you can find stuff that you love doing, it means because you enjoy it, you tend to excel at that. It doesn't mean you're the best or anything else, but you excel at that because that's your drive for getting up every day. You have to have a purpose in life. A lot of people don't always find what their purpose is. I know what mine is, so I'll go heads down every day. And I'm happy to do the grind.
Ryan Cramer: That's amazing. Well, I think those are good words for obviously any entrepreneur too because I grew up, actually, and I'm still trying to figure out what are those things that you want to. And I hate the question of what is your ultimate goal in life? But I think a lot of times people can actually find, we use the word, what are you passionate about and what do you like doing on a day- to- day basis and encompassing what your happiness is around. Like you said, this podcast has been a passion project of mine of talking to people like yourself and thought leaders in the space, people who walk the walk and talk the talk. It becomes so much more easier when you find that one thing every day where you can wake up, you can say, " This is going to be a great day regardless of all the other shit that's going on in my life." I actually get to talk with him on a Friday, or I don't want to wake up, or I don't want to do the prep work for that. So stuff like that, I always find super fascinating that people just go, and again, it's the old adage of I get stuck in a 9- 5, but for people who are in the space, it's actually more than a 9-5, it's that constant grind.
Danny McMillan: There's no work- life balance. It's all- consuming, so this is why you need to integrate into your life. Because people go, " Oh yeah, who chimes off at 5: 00 on the dot and don't look at their e- mail?" Even people in corporate, they don't do it. There is barely a separation between the two, other than if you work for the man, then you'll get a wage. You'll get a salary at the end of the month. Right? But the thing is an entrepreneur, the buck stops with you. The burden of performance is on you. Success is only rented. Every day you're paying the rent, so what you've got to do is you've got to be good with paying that rent. And if you're happy to pay that rent, but you're doing things that you enjoy. See, there's a difference between doing something you're passionate about and that's the only reason. That's why I see this part has three rules. Right?
Ryan Cramer: Right.
Danny McMillan: Does it make meaning? Does it make me happy? Does it make money? Whereas some people might say follow your passion, but you need to work out those logistics in between. Does it solve a problem? Does it help people? Do you see where I'm coming from? Not just chasing your passion. So you could have a passion project where you want to knit jumpers. The chances of doing that and making that a success by hand and only you is going to be very limited, so you have to pick things that you can be passionate about that have meaning, but of course, they're going to make money. It doesn't have to be today. I like to do everything I do is long- term. So if I look at something like Branded by Women, I'll go and raise the shit ton of sponsorship money. Yeah? And we go hard on the market in the videos, production, and everything else. Right? The conversion rate is a lot lower. Right? Most conferences could... Most of the tickets are bought by men, but now them habits are starting to change. So it might take me to five, six, seven, it doesn't matter how long. But you know you're going to get there because that's part of your purpose. You're on a mission, and you just get it done. And that's really what it comes down to. You just, you find it, you embellish everything about it, and then you go out there and go and get the job done. That's always the best.
Ryan Cramer: Is that what makes the major difference between entrepreneurs, successful entrepreneurs and the ones who fail more often than not? Is it that drive?
Danny McMillan: There're far more successful entrepreneurs out there than me because the nature of me not chasing the pound note. I could earn a lot more money by saying no to 99% of the things. Because money isn't my driver, so you'll need to define your success. Right? I know lots of people that have made a shit ton of cash, and once they got there, they wasn't happy. They were still finding their meaning and their purpose, and if you spend all of your life being super disciplined to chase money, and then still be unhappy, you've still got to find your life's purpose. Do you see what I mean?
Ryan Cramer: Right.
Danny McMillan: So it becomes part of the journey. The goal is the journey.
Ryan Cramer: Absolutely.
Danny McMillan: Right? So every day you get up, you might as well enjoy it because being a business owner has its downsides as well. Again, burden of performance is always on you, accountability is always on you, and the rent has to be paid every day.
Ryan Cramer: Right, well most entrepreneurs are making their earnings and paying their rent via Amazon. So I guess going into our topic, I know your time is so valuable. Everyone's time is so valuable, so taking that step back, if you will, of not going into so much deep content, but your passion is figuring out Amazing right. Amazon is this big, I won't call it, it's an ecosystem, but it's this cerebral living breathing thing that constantly is evolving over time. It's almost like this big AI computer that is constantly changing and evolving and adapting. People don't even know exactly what an A9 algorithm is, but it actually is what operates and is helping in that background of making sure everything is on a platform or playing field, if you will. So for the person who may not understand, oh, I didn't' realize that's what Amazon operates under, what is the A9 algorithm, if you will?
Danny McMillan: Well, A9 is what you use for, in basic terms, A9 is what you use for organic searching. When you go to the search bar on Amazon and search. Obviously the term PPC results as well because of the ads. But generally that is what A9 is there for. It's a search engine, just like Google, but rather than Google bringing together the world's information, this is about refinement of bringing the most relevant products to the end user or the person who's ready to purchase. So when you type and search on Amazon, that search's intent when you're very determined on what you want to buy, so it's very clear. Whereas Google you're there, you'll do the research to find these kind of places in terms of commerce or just information. Whereas on Amazon, you know exactly what you're doing. You shop on there because you want to go on there for convenience. So ultimately it comes down to relevance, obviously conversion rate, et cetera all goes into the information I've got here. So the way I've done it is that a few months ago, I'd come across a load of science literature. Right? So there's Amazon. Science. You can go on there, and there is the Joy of Ranking, which was a video that I found on YouTube, which is by Daria Sorokina. That was published in 2016. So she's done a visual and audio presentation from that. And then I've got the science lit that I've set some slides up, which are called open source because I believe discovering that research paper and everything else and extrapolating that information isn't really by my design. I don't do courses anyway, but I think it should be in the hands of each and every Amazon seller as the information is freely available out there. And if you sell on Amazon, it's going to be your bread and butter. Now don't expect you to understand the hunger score and how that operates and various different ways it works on the platform, but there are a lot of key things that you should be aware of that you can take away from this chat today, and which will help you move forward in terms of what you're doing because obviously there's... People come up with theories, and they come up with their tests. And they have a theory behind those tests that they execute, whether they come out good or bad. They use observations. I've just taken the data from the team, which is public information, then broke that down. I've actually got like 200 slides worth of that. The hardest job I had was to take this information that would send a glass eye to sleep, but had to make it palatable, and most importantly, digestible for an Amazon seller to go, " Oh yeah, I get that bit. I get that." And so that was quite a hard task to maneuver that content one, so people still hold their attention to want to hear about, and two, deliver it in a way where that could become useful for them and to give them antidotes of how that can benefit them.
Ryan Cramer: Very cool. So breaking that down, that sounds like something you're taking so much time and effort. What were the most surprising things? Is it something that shocked you more or did it verify more of what you thought? Like everyone has this idea of I really think that this is, their own hypotheses if you will, but then I'll take it the step further and say I'm going to prove my hypotheses. You took the data that's raw, it's open and out there. Amazon's somewhat public about some of those API accesses and whatnot. But you-
Danny McMillan: But it's not API. You look at advertising with API, you've got UWS and other APIs.
Ryan Cramer: Yes, exactly.
Danny McMillan: But not-
Ryan Cramer: Certain access.
Danny McMillan: ...A9, you know?
Ryan Cramer: Exactly, so this is something you broke down. Did you find something super surprising or maybe shocking? Or was it verification of what we always thought?
Danny McMillan: I think the biggest takeaway, if you were to turn this show off now. If you were just to say, right, what do I need to know? Right? Obviously sales, that's a no brainer. But of the biggest things from that, and it keeps referencing back to is add to cart. This is why you've got a plethora of people doing add to cart in the industry. Whilst that's not in the terms of service as manipulation because the verbiage is not there to support it because it hasn't reached ubiquitous status. Because although we work in this kind of community of Amazon. It's actually small in comparison to every seller they have on the platform. It's not just us. Right? So we know a lot of people in the community. The community is of a certain size, so you become an echo chamber. It's the same with Black Hat and things like that, which I don't encourage. But Black Hat is only a tiny problem on Amazon in comparison to the global sizing. Does that makes sense?
Ryan Cramer: Yeah, 100%.
Danny McMillan: So it's the same with A9. Not many people on the scale would actually know too much about it because it's like a black box. But for everyone in our community, they get excited. That's why they listen to podcasts, they go in the blogs, they go into the forums and stuff. And so this is why this information will be important to them because it's part of bread and butter of what they do every day.
Ryan Cramer: Right, it's a leg up on the competition, if you will. It's not a bad thing, it's just actively using the tools in your favor as they're deemed available. What are those crosstalk things, if you will?
Danny McMillan: Should I go through some bits here?
Ryan Cramer: Yeah, absolutely.
Danny McMillan: Because I pulled out some quotes as well rather than doing slides and stuff because it's more on the conversational basis.
Ryan Cramer: Of course.
Danny McMillan: So according to Amazon's own data, 70% of Amazon customers never click past the first page of search results. Right?
Ryan Cramer: Makes sense.
Danny McMillan: 35% of shoppers click on the very first product featured on the search page. Right? The first three items displayed in search results account for 64% of the clicks. So if you think about desktop, and you see the way that PPC moves, this is why Amazon's become the pay to play platform. And on the desktop often, your own organic's going to be pushed down close to being below the fold. So your visibility starts to really disappear. Right? So if you're not on page one, according to Amazon's own data here, 81% of clicks on brands are on the first page of search results. So going back quickly to the biggest part I would say in terms of a signal because it's a conversion that's on a continuum, right? So the sale, the final sale is the conversion. It's what you build into. Now people manipulate using add to cart, but add to cart will take you to the bottom of the first page. It won't take you up because you need actual physical sales binded to those keywords to activate to get it up the page. And if you notice, when people do Search Find Buy and stuff, it's very easy to from up in the gods are 432 to here, the bottom of page two in a matter of days. Right? You see it, just ping, you can't look at healing ten or whatever. It's very easy to get there because realistically all the activities taking place, or nearly all the activity is taking place on page one. Therefore, anything short of page one on the balance of probabilities is a complete waste of your time. You're literally ghosted.
Ryan Cramer: Yeah. Yeah, that's fascinating because ranking- wise, that's difficult to, that would make me think about my ranking strategy completely different. Obviously how to get it up there and let the organic nature take over it. Page one, page two are obviously a holy grail for where you're going to be and in service search functionality. Makes sense.
Danny McMillan: So the other thing is as well, there's been a big thing, not big thing, but generally pops up every so often is a blog post or a feature talking about A10. Now A10 does not exist, and it's nowhere to be seen in the science literature as well. And the quote here, " Small improvement and relevance can positively impact millions of shoppers." So small, minute changes can impact a million. So why would you throw away nine and then go and start with A10? You wouldn't. It doesn't exist, and there's no literature that I found on it. So I think that's quite important as well to understand. If you go back to 2016, they were using 100 machine learning models. Another thing to be aware of, across the Amazon platforms, I'm quoting here, they're managing different catalogs, indexes, and ranking models, so effectively, not all the markets are built the same in terms of ranking and search terms. So what that means by that is when you get an update in the U.S., it doesn't update in Germany. They've got separate teams. And on those teams, they run quality control tests, and they take a period of traffic at the time before they push out an update. So it's platform- dependent. So whilst you've still got A9 being the main driver overall on a global level that are used on the platform. The updates and stuff that you see and not all, they're not all coming from one place, therefore it goes across everywhere else. It's platform specific, yeah?
Ryan Cramer: Okay. Makes sense.
Danny McMillan: So what happens is the data's collected from products including sales, reviews, and the other bits I'll get into in a moment. And what they do is they've got training data, which is collected from customer behavior logs because behavior drives the impact of relevance and ranking on Amazon. So I'll give an example to quote. We use search to collect our training sets. Several times per day, we compute the unique set of keywords issued for each context of interest. The context could be a combination of marketplace category and some user features. Within that is context of query and textural similarity relating to the query. This is why behavioral features drive Amazon's rankings. Now there are some quotes, and I'll break these down into layman's terms. So to quote here. " Since the most relevant products appear higher in ranking, the observed click through rate, given position, captures not only the position bias, but also the typical relevance at this position." So what that breaks down into layman's term and into English is what the position bias means that you have a higher click through rate at the top before fewer and fewer clicks come down. So it goes back to those points earlier on. Of the activity, 64% of clicks are done on the first three products at the top. Now when it comes to the relevance, being high on the page means you tend to have a higher click through rate than the lower placed item even if the lower placed item is more relevant. So that can be a number of things. That could come down to the unit price of the product because people can be looking for deals on Amazon. Yeah? For training the ranking models, we use labels based on customer actions, such as purchases, add to basket, which is add to cart, or clicks. Right? So you've got two sets of labels. You've got positive and negative labels. So positive labels are clicked, add to cart, and purchased. They also bundle consumed for digital categories like Kindle, Amazon Music. So, example. You can't see someone drinking a can of cola that they bought off Amazon, but Amazon just needs the experience utilized by how do we watch the movie, where have you shared the link to that movie, and how many times you've watched it. Yeah? So they've got that additional data there, and that will feed into the system of the positive and negative labels. It comes back to the whole point of relevance. So ignored results I'll get to because I see you're starting to look a bit confused now.
Ryan Cramer: Right, no, I'm not confused. I'm thinking through this a little bit more too. And I was going to verify. So would that be, you said, sharing digital content like how many times you're consuming that content. Would that go as far as sharing that listing on social media or on the website or anything like that?
Danny McMillan: Yeah, so what you're saying is they've got negative and positive labels. Amazon wants to know just as much about what you don't like as what you do. And that's why they allocate you two sets of label. So the negative labels fall under a sub- category of ignored results. So product was shown but no action was taken. If the customer is on the results page and picks number three, then number one and number two would get a negative vote attribution. So let's just quickly tie this back into optimizing a listing. See optimizing listing is the flat vector rather than 3D. And what I mean by that is you do the title. You're going to do the bullet points. You do the backing keywords. You do the description. Right? You've got photos that are engaging. So you're thinking a bit in that way. If I've written my bullet points right, am I focusing on the benefits over the features and stuff like that. Well, the next step you've got to think about is the level of engagement. Right? And so this is where conversion on a continuum comes into play. Let me just pull up my notes. Just blended score. I'm skipping some of this stuff because it can be quite complex.
Ryan Cramer: It's all adding to the basics. So you're talking about the basis of what every person does. crosstalk
Danny McMillan: Right, yeah, so when we look at a listing, we look in the flat sense, not the 3D sense. Right? So if a conversion is on the continuum, if you've got a negative and you've got positive labels, if someone comes to that listing and bounces straight off, there's no time on page, there's no scrolling up and down, there's no having them view the images, there's no clicking to look at the reviews, so that listing's got no engagement. Right? But then a very similar listing that's very relatable to that because the ranking has to be adjusted in terms of position. If they learned on that one that they click and look at the images, there's a micro credit. Then if they scroll down, they're reading the bullet points, they got the time on page asset there as well. You scroll down a bit further and you go back up to the top and then you add it to cart. Now, they didn't buy that, but there still needs to be a calculation in between. So the point I get back to having 3D is that when you're writing a listing, you need to add a new layer now, which is how are people engaging with this, and what can I do to maximize the engagement on the page?
Ryan Cramer: Yeah, makes a lot of sense.
Danny McMillan: Rather than can I write really great sales copy because that's one dimensional. They're thinking of it only as the text not necessarily thinking of the step back of the entertainment.
Ryan Cramer: Right.
Danny McMillan: Right?
Ryan Cramer: Making an impression instead of action, yeah, exactly. It makes sense.
Danny McMillan: And then you realize if people clicking on your listing and they're not buying, that will fall into the negative especially if they bounce up. So they might like your images but don't like the price, even though you can see the price from the search page and the thumbnail. But the point is that you have to start thinking about here in a different light is how can I make this as engaging as possible. So that once you've got your listing up, you should review it. You should get other people to look at it. You want to look at your EBC and everything else. Is it engaging enough? Is it too long? Is it too short? Am I using body parts? Is it too much block text? Do you see what I mean because we know that sales drive the platform? crosstalk It could be searches and ten. And we know that we need to add the right keywords to find on the product, but we shouldn't stop there. We've got to think about the engagement as well because you can still maximize engagement on the page and collect micro credits versus them bouncing off. So it gives you a new way to look at a way that you're optimizing.
Ryan Cramer: Well, is that where most people maybe not optimize as much?
Danny McMillan: Well, people don't realize about negative and positive labels because they're not necessarily aware of it. They haven't read the literature. And that's totally understandable. But by reading the literature, you get to understand. Go to places like Amazon. Science and go and look at the Joy of Search, which you'll find on YouTube. It's all out there, just go and have a look for yourself and then look at your interpretation because what you're taking is research papers that have been provided from the source for you to interpret what is going to be required. And so we're always talking about relevance in a flat sense. Is that keyword relevant with that? Well, yeah, it could still be relevant, but what about the engagement on the page? The higher engagement, the more relevant it is. So that you're now thinking in a 3D motion of an extra layer that you're working with.
Ryan Cramer: I like it. Yeah, I think that's a great takeaway.
Danny McMillan: So all effectively you're doing is you're sharpening your tools to look at new dimensions of how you optimize your listing. You're just adding to that. Lots of people know how to do titles really well. They know how to get a copywriter to write great copy. But there's always that bit more to do. And so it goes back to the point. If you look at the correlation between people using PPC using rank. People will turn around and used to say, " Oh, it stopped working," and everything else. You can't pummel your way to the top anymore. Those days are gone. You need a stupendous conversion rate. Right? If you've got a poor conversion rate, then you'll drop down, so it has a negative impact. It goes back to positive and negative labels. So you need to have a very well-optimized listing and start, and especially in competitive categories, you're going to need to add some conversion history because you're starting with a dead start with a new product. So let's say you go, " Right, I want to run keywords on headphones." You've got no chance because all the people that granted in history, not only that, there's a chance you won't even get an impression share. The keywords even if you try to bid$ 20, you're still not going to get any love. So it comes back to the game is you want to maximize the conversion rate as high as possible. Now here's another problem. When you login to, which we may need to cover on another show, is that me and Ellis broke down conversion rates from the business reports. They're way out. They're way out. Some people might say I've got 40% conversion rates and it's really 26%. So we've got a few calculations. Let me see if I've got the calculations here, just it might be in these slides.
Ryan Cramer: Right, and when you find that, I also want to one, I was listening to a talk you gave, I thought it was fascinating when people are ranking products at such a high conversion rate. You said the negative impact on your listing can be if your conversion rate drops. So if you're doing search find buy or if you're doing those rankings of giveaways, it's 100% conversion. All of a sudden, even if it drops to 85% conversion, that's a negative impact on where it was versus where it is.
Danny McMillan: It suddenly goes from give 100 people a keyword to search and buy. All hundred people do, that's 100%. Then let's go back to the average on the platform, say a 20% conversion rate because the search's intent is not the that high. So you end up going back to real purchases on the platform. And this is why a lot of people drop. So people may blame their product and say, " Oh maybe my product's not good enough." It's not necessarily the product is not good enough. Your conversion rate is not sustainable at 100%, and that's where the problem lies. But it's still a great tool for people to use. People use it still effectively, but it's always good to understand why.
Ryan Cramer: Right, I was going to say you might have to adjust or tweak some, like not make it 100% but almost hedge your bets and make it a little bit lower but consistent so you over time.
Danny McMillan: Yeah, I don't think I have them here, these slides. We'll have to do it another day, but I've got two calculations. One to work out your true, organic conversion rate, and one to work out your true PPC because it's things like Subscribe& Save and stuff that you'd need to make adjustments for the calculations.
Ryan Cramer: Because, yeah, that would inflate your conversion rate higher than what it traditionally is.
Danny McMillan: Yeah, indeed. And yeah, going back to, say, high- velocity products. I might actually... Can I share my screen and show you-
Ryan Cramer: Yep, let's go ahead.
Danny McMillan: ... theslides here because I would think it would be easier to do so.
Ryan Cramer: Yeah, and just to reset. For everyone who's listening and watching it, we have Danny McMillan. He's kind enough to talk about the A9 algorithm, and again, that's searching functionality, which helps rank products on Amazon and just the science behind it really, to get through the data, the conversion rates, all the fun assets that no one really want to deep dive into but actually helps understand where Amazon's coming from to help almost reverse- engineer the ranking process and getting products listed on page one or page two and being successful on there.
Danny McMillan: Let me share here. Window, crown type. All right.
Ryan Cramer: There you go, go and add there.
Danny McMillan: Right, so problems ranking with high velocity products. This comes back to the ranking algo. There's so many add position placements worthy of eyeballs. So for high velocity keyword, we've established listings that are already converting. It can be extraordinarily difficult breaking and getting impressions, even with phenomenal conversion rates. So this is why you're hearing people bid$ 20 a click without getting impressions. So it's depending on your stamina if you keep going over time. Amazon's algo may determine to favor your campaign, but there's no telling how long that will be. But it could give you a nice boost down the road. So let's look at the answer. Is there a special calculation? The answer is yes, no, and some common sense. So what this is here is an example of a chart, and we're using two extremes. So keyword one is showing position one, position two, position three for fictional sales of 1, 000 units a day running PPC on this high velocity keyword. Fictional sales position two 500, and then position number three at 250. So if you're looking to spend money on PPC, and you're like, " How do calculate for this?", then you'll maybe look at the lower hanging fruits. You might go, you know what, there's a mid- tail keyword. Let's start with that one. So keyword two, 50 units a day. Well, that's still quite a lot. If you're looking for a top of the placement, with a conversion rate expressly 20%. But if you look down, position number five, you go all right, five units a day. That's achievable. So when you break it down, so to get to position number four, you might aim for at least five orders a day. If your conversion rate is 20%, then that may mean five orders, which is then divided by a conversion rate, which is equal to 25 clicks.
Ryan Cramer: Sure.
Danny McMillan: So if the cost per click is$ 5, that position, it will cost you$ 5, divide by conversion rate, which is equal to$ 25 per acquisition on average or$ 125 per day, which comes back to the point of gaining impression share and the cost of running the expensive top of search campaigns. So usually in a market that's red hot, generally you'd have a higher cost per click. So how do we launch PPC and then taper off knowing when our numbers land for break even? Well, they'll come with their pros and cons. Either way, you're going to need to pick your poison. So if you two are just to bid out the gate, it would almost certainly affect the ad placement. You drop the bid, that means someone in the ad placement algorithm and they're bidding higher than you, generally it's going to be higher than you. That's where you come and see the downside of adjusting the bid. But if you start pulling back on the daily budget versus decreasing the bid, you're less likely to lose that position. So at least with budget caps, your campaign will run in most cases, but you'll lose advertising for the rest of the day after you've burned through the budget. So it's another way of doing it to create this pull back. So you've got hybrid version. Twice a day, you manually address top of search boost in your campaign. So you start the day at say 900%, and then 12 hours later, you drop it to 0%, and then you work out if it makes economic sense. So you're going to play around with these factors and then you find the sweet spot, and then you generally get your ranking cost under control as you taper off over the six to eight week period. So to wrap this part in summary, competitive conversion rate in orders matter highly, highly to optimize listing matters highly, and you've got to remember going back to earlier point in the show, your conversion rates is on a continuum, so engagement, add to cart, clicks, and don't see conversion as a static object.
Ryan Cramer: Right, that's amazing. I love that. That's all really great content, and I know the takeaways. There's so much that we just added in that 39 minutes or so that we've just been talking. But for everyone who want to learn more, maybe, Danny, do you have any parting words? I know we jampacked it in there just so people can take it and apply it today. What are those parting things that you want to make sure that people takeaway and do that for Q4? Is there anything specific regarding this?
Danny McMillan: Yeah, it's very straightforward. We're all trying to learn all the time. What I love about Amazon sellers is they're very engaged, and they're very into marketing and everything else, and they're passionate about it. And I think what happens is when Amazon are providing conversion rate data that's incorrect, it really shifts the basis of your business until you recalibrate it and understand your true conversion rate. It's the same with PPC. We have to come up with a solution. It's like how do we put something together. Now, that's very abstract what I've done because I said 1, 000 units a day because I'm being over extreme. Because what people do is they have a low budget. If you've only got so much budget to burn on PPC and it doesn't work, that product's dead unless you can borrow a load of other money. So what about if you had the format? What about if you go low hanging fruit, mid- tail keyword. I need to get X about the day. I'm going to express this as 20% conversion rate. This is going to be my daily cost. I will use cash budgets, and I will run this for eight weeks to see where it takes us. Then you can put that as a line item in your spreadsheet and go, " Here is a basis of my marketing cost," rather than go, " Shit, I need to rank this product. Let's throw a load of money at it," and then find out that doesn't work.
Ryan Cramer: Right.
Danny McMillan: So it's trying to come up with things like that that can help sellers to go off and do their own set of calculations because there's no hard and fast rules on the exact science of how it works because there're too many variables to calculate. So we try to simplify a process that you can use, it's a rule of thumb rather than go in, " Right, I'm going to spend ten grand on PPC," but don't know how long the budget's going to last and what kind of results did it get. So you target the wrong keyword, then it's going to cost you a lot of money.
Ryan Cramer: Like you said, low hanging fruit is going to just boost that conversion. It's going to bring validity to that product. I think Amazon then naturally starts to bump it up even more, and it's really hard and fast to go the number one ranked product or ones that have 100,000 searches in a given month, those are difficult to achieve with a historical products, just other ranking products, and reviewed products. But ranking new products, get that low hanging fruit, picking off and working up to picking off the middle of the little guys, if you will, will really help you achieve long- term success. So, I think that's amazing. Thank you for packing that all in for our listeners.
Danny McMillan: You're welcome.
Ryan Cramer: But you have so much. We talked about Branded by Women. We talked about Sellers Sessions Live or Sellers Sessions, the polls and all the fun stuff. Sellers Sessions Live is sold out. It's going to be happening October 7?
Danny McMillan: Two weeks, yeah. So two weeks today it kicks off.
Ryan Cramer: Two weeks.
Danny McMillan: Yeah.
Ryan Cramer: So yeah, is that exciting for you? What's the first event you're putting on?
Danny McMillan: It's the seventh time I've moved the date. Because my glass I normally half full, I'm like, " Don't worry about COVID, we'll get it sorted. It'll clear up soon." And you keep setting the date because your intention is good. You want to get people back together, but you have to do it so it's safe. So the date's moved seven times. And right now we've got this little window left, and anything can happen in the next two weeks. So I'm just going to be prepared for the unexpected, but I hope we get it off because I feel the world needs to return in a safe manner. Unfortunately there've been a lot of people that have been lost to COVID and I'm very sad about that and condolences to their families, but the damage coming soon, the psychological damage of being locked down for a number of years and having a lot of restrictions, that is going to have massive impact as well. And the quicker the world becomes semi- normal, and we get these people back in a room, when we get them inspired again, the better.
Ryan Cramer: Right, there's that fine balance of safety but also inspiration and achievement and drive again.
Danny McMillan: I have to be clear, and I've told the delegates, " Look, we have to be responsible." Right? Because when people come back from Prosper, people did catch. They were there a week long and it was five days, so you're constantly packed in. This is a one day event. But still, you have to take precautions. And people have to be responsible for yourself. And I don't get into that vaccine/ anti- vaccine. All I care about is the safety of the people in the room. So I've asked people, your choice. You're not obligated to. If you'd like to wear a mask, wear a mask. I'd like you to do lateral flow test before you come, test yourself. Be responsible for yourself. You can't tell people what to do. But ultimately everyone's safety is under my name. My name is above the door, so I want everyone to be as safe as possible. So you're still taking, it's still a risk in COVID times, but I think it's very important for people to get in a room. People are desperate to get out to events. In America, it's been running for a lot longer. There's been a lot of events going on. There's nothing in the UK. I've done Poland just recently. I've done Estonia and places like that, but there really hasn't been much in the way in the UK. Now, we're coming into the winter season now, and so a lot of these events are going to start kicking in, but it's been a long while. It's been nearly two years for people.
Ryan Cramer: I know. It's seems like an eternity, but I think that especially when you look at it, and the fact of working in an office every day, you're seeing people quite often. As an entrepreneur, it's almost one of the loneliest positions you can possibly have because it might just be yourself. It might be you and a partner and that's it, and it's just you talking to each other all the time. That engagement, that social interaction is very important on a psychological level. Like you said, I personally appreciate the awareness of ultimately everyone wants to do it safely and make sure you can get in and get out, and no one's of any harm or any health thing. So I appreciate, I'm super excited to see it kick off.
Danny McMillan: The other important thing about it as well, the release of oxytocin, human touch. People underestimate the power of that. And the difference from that separation from people from their loved ones and everything else. And that integration, going back into society of that, I think it's just really, really important, but it has to be done in a safe manner.
Ryan Cramer: Absolutely. Well, I appreciate what you do, and again, keeping people, like the podcasts available, and like you said, you're constantly pumping out content so people can stimulate that creative side. And I appreciate what you do for the community. I know it's a lot of drain on you for all the different events you're doing, but I think it's super inspiring and exciting for other people, like you said, across the pond, if you will to steal a term, and look at what you're doing and be inspired by it. So thank you again for hopping on. I know we already ran over 15 minutes longer than what we first started and said, so but hey, thanks. If people want to connect with you, or if they want to learn more about DATAbrill or Seller Sessions Live for future purposes or how to get in touch with you, what's that best way?
Danny McMillan: If you want to email me, danny @ sellersessions. com for anything related to that side, the conference side and the podcasts. And then DATAbrill, danny @ databrill. com for PPC.
Ryan Cramer: Awesome, well thank you so much, Danny, again for hopping on today. And we'll have to have you on because I know that was so much information we could down all these different paths together. I appreciate it. I know our listeners appreciate it, so thank you so much for doing that today.
Danny McMillan: Thank you. Take care.
Ryan Cramer: Awesome. Thanks, Danny. Again, everyone, thank you for hopping on again on a Friday. We like to have a lot of fun, but that was such an action- packed episode. We have a shorter, but there was so much more content that we got to do in that amount of time, so thank you so much for spending the time here on a beautiful Friday wherever you might be tuning in from. If you're watching this on LinkedIn, Facebook, YouTube, or Twitter, or again, if you're listening to this on your favorite podcast station, Amazon Music, Spotify, Apple Podcasts, Google Podcasts, Crossover Commerce is on all those places, so we're catching up on all of our updated episodes. This is episode 159. That means there's 158 other episodes that if you haven't had a chance, go back and listen to. Super applicable that covers anything from logistics to product listing to Amazon advertising to marketing to even Tik Tok. We talked about inspiration, localization. You name it, we're going to cover it on this show. So that being said, I'm Ryan Cramer. This is my corner of the internet. And this is my podcast, Crossover Commerce presented by PingPong Payments. We'll catch you guys next time on another episode next week. I will have three more episodes that are going to help you affect, grow, and develop your eCommerce business. Take care.
Ryan Cramer of Crossover Commerce talks with Danny McMillan of Seller Sessions one-on-one to review what you need to know about Amazon’s A9 Algorithm.
Crossover Commerce is presented by PingPong Payments. PingPong transfers more than 150 million dollars a day for eCommerce sellers just like you. Helping over 1 million customers now, PingPong has processed over 90 BILLION dollars in cross-border payments. Save with a PingPong account today!
Stay connected with Crossover Commerce and PingPong Payments:
✅ Crossover Commerce @ https://www.facebook.com/CrossoverCommerce
✅ YouTube @ https://www.youtube.com/c/PingPongPayments
✅ LinkedIn @ https://www.linkedin.com/company/pingpongglobal/
You can watch or listen to all episodes of Crossover Commerce at: https://usa.pingpongx.com/podcast