#117 - How to Troubleshoot Falling Sales on Amazon (And How to Grow Sales Instead)

#117 - How to Troubleshoot Falling Sales on Amazon (And How to Grow Sales Instead) Speaker 1: Today, we're talking about the most important metric on Amazon, sales. Speaker 2: It's critical that every Amazon seller and advertiser knows how to identify and troubleshoot a decline in sales to get things back on track. Speaker 1: All of what we're about to share also applies to hacking sales growth. So 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 1: So this episode is piggybacking off of the previous episode about troubleshooting ACoS, understanding what actually makes ACoS go up or down and how you can analyze performance in your account as well as make optimizations to improve ACoS. This episode will likely be a little bit better because Andrew and I planned a bit more for this episode. So if you haven't seen that other episode, You can take a lot of the learnings from this one, go back, make sure you watch that episode on rising and falling ACOS because it's a lot of the same principles that we'll be applying. But when it comes to actually calculating the change in sales, Most people think they know this, but they don't actually put it into practice. And this is probably the most critical formula that I think everybody needs to know and also action. So Andrew, what is this formula here telling us? Speaker 2: Yeah, it's basically just giving us the formula for sales and understanding that sales is just a combination of your traffic times your conversion rate and really just illustrates that as those things change, that's going to directly influence your sales. So the delta right there in the formula that you're seeing, that just means the change. So as your clicks change, as your conversion rates change, that's going to directly influence how your sales go up and down. Speaker 1: Yep, and for this episode, we're just going to stick with PPC sales, advertising sales, but all of this translates perfectly over to total sales as well. You would just be replacing clicks with sessions. You can also use page views and things like that, but I think sessions and clicks, that's the only metric that's, you know, the same definition between Amazon Seller Central versus the ad console because tracking units and things is kind of difficult. When it comes to calculating conversion rate in Amazon, a lot of folks like to use the unit page view percentage. That's a pretty good proxy. The only thing to keep in mind is that there can be multiple units per order and there could be multiple page views per session. So it's not the same definition. If that's how you're defining conversion rates in Seller Central, it's not the same as what's in the ad console, which is orders over clicks. So there can be multiple units in an order. There can be multiple, yeah, there can be multiple units in an order and multiple page views, detailed page views in a click. So orders over clicks and purchases within Seller Central over Sessions is going to be your same conversion rate definition. So as Andrew was saying, the change in sales is just the change of clicks and the change in conversion rate. So question to the audience here, feel free to pause the screen if you need time to think about it. But what would happen to your sales if you got twice the amount of clicks? So this is the change, right? So like month over month, you doubled the amount of clicks, but your conversion rate got chopped in half. Well, your ACoS probably would have spiked, but what would have happened to the actual sales is there would have been no change because 2X times one half is one. And so that is basically, I mean, these are kind of big numbers. You're probably not ever going to see that much. I mean, you could. But the idea here is just you multiply or add these changes together, and that's going to give you the total change in sales. Anything to add here, Andrew? Speaker 2: No, no. I really just love this simplistic way of breaking things down, really piggybacking off that last episode where it just creates a really simplistic framework to understand and diagnose When something changes, how to understand and troubleshoot what it is. You know, I get clients messaging me every week, like what happened to the sales on this product or that product and having this clear framework to just be able to drill into, okay, was it a traffic problem or was it a conversion rate problem? Where was the change? And just being able to pinpoint and identify and illustrate what is driving the change in performance to the clients that I'm working with on a regular basis. So this is a constant question that comes up and having a simple framework for it is powerful. Speaker 1: Yeah, and in a minute here, we're going to dive into some actual examples. So please stay tuned. This episode is very important. A lot of people need this kind of in-depth analysis around troubleshooting sales and understanding what actually makes them go up and down. In this case, Yeah, sales would have stayed flat, but let's just say that conversion rate was flat. So if you got twice the amount of clicks with the same amount of conversion, you would have gotten twice the amount of sales. Conversely, if you only got the same amount of clicks, so 1x, no change, and conversion rate got cut in half, your sales would have dropped in half. So the whole purpose of this is you want to find out which metric moved by a larger amount. In this case, they kind of cancel each other out. But if your sales is going, you know, up or down, you want to look at, let's say went down, right? You want to look at where did that decrease come from? Did it come from the conversion rate side? Or did it come from the traffic side? And whichever of those is the larger decrease, that's going to be the first thing you want to look at to try to assess. So To be particularly accurate, there is one other variable involved here, which is also the change in average order value. The reason why we don't emphasize that in the formula is because AOV tends to not fluctuate as much. You'll see clicks and conversion rates can go up and down by, I mean, clicks can go up and down 10, 20, 30% within a month. Conversion rates can fluctuate, you know, 5, 10, 15%. AOV typically is only changing, you know, less than 5%. So it's not usually as big of a variable, but it's one other thing to check, especially around deal periods and things like that. So, and usually conversion rate and AOV are inversely correlated. So it could be, you know, you had a 5% increase in AOV, which led to a 10% decrease in conversion rate, and then that's gonna cause a negative impact to your sales. So just worth mentioning that. Speaker 2: Yeah, well said. I mean, I was going to say the same exact thing. Pricing, you know, most people aren't really changing their prices every single day, every single week. Usually that's like once every six months or a year, maybe a little more frequently now that, you know, there's tariffs and all this type of stuff. I've seen that a lot more people increasing prices, but, you know, you're definitely going to see that in the conversion rates overall. If you're increasing price, you're going to see conversion rates inversely correlate to that. Speaker 1: Now let's go on to change in traffic, the change in clicks. It's also important to understand, you know, if your clicks did decrease by, let's say it's down 25%, conversion rate stayed flat, that would mean your sales are down 25%. So you wanna try to figure out, okay, what might have happened with my clicks? I know that my clicks are down. What's gonna impact that? Well, we can do that same logic where we're just going to look at the impressions The change in impressions times the change in click-through rate. You're going to get that same kind of output. So another example, if your impressions quadrupled, so you're getting way more impressions, but your click-through rate dropped in half. Andrew, what's the answer? What's the change in clicks? Speaker 2: I would just have to take a wild guess and say it is a 2x in clicks. Speaker 1: Whoa, Andrew, you're a pretty smart guy. Speaker 2: Some people say I'm like the Einstein of my generation. Speaker 1: I don't know. Speaker 2: I don't know. Yeah. Speaker 1: It's a little generous. Speaker 2: It's just what people say. I don't know. Speaker 1: Well, anyways. Yeah. So hopefully it's making sense. If your impressions quadrupled, but your click through rates got chopped in half, you're still going to have twice as many clicks as you previously had. And so If you do have that change in clicks, let's say your clicks are down, you want to look at where did that come from? Did it come from impressions or click-through rate? If they're both down 25%, then you're going to have to look individually at each of those things. But most of the time, it's going to be one or the other that moved by a larger degree, and that is going to help you pinpoint what exactly you need to look at. Let's just say your click-through rate stayed the same, but your impressions dropped. Actually, I forgot what slide is. Yeah, so impressions comes next. So if your impressions dropped larger than your click-through rate, these are the things you're going to want to look at. It could just be the search volume. That's a pretty huge thing. So those search query performance reports available in the brand analytics, we have a whole episode dedicated to SQP, so you can check that out if that's new to you. But you want to look at what's happening with that search volume. And Andrew, I'm talking too long. You take us through the next two bullets. Speaker 2: Yeah, search volume is really going to be the main driver there. If that's moving a lot, your impressions could swing quite a bit with that. Another thing you want to check is your CPCs or your bids and your placements. Like was there a shift in where the impressions were actually coming from? Did your bids change? Your CPCs change too much where you started losing top of search visibility or Maybe you cleaned up some product page impressions and actually reallocated that to better performing placements. You can usually see a decrease in impressions there. And then one other thing you'd want to keep an eye on and one thing I've been really tracking a lot lately as I've been doing more ranking campaigns is tracking impression share and impression rank. So you can get an idea of relative to other sellers in the space or advertisers in the space are what percentage of impressions are you actually capturing? If that's staying relatively steady, you know that it's going to be a change in volume overall. Search volume happening within those individual search queries and you know kind of relative to other competitors where you stack up. Super important for making sure that you're maintaining your competitiveness with those bids and placements and everything. Speaker 1: Yeah, so it could just be you reduced your bids and that caused you to lose a lot of impressions. But looking at those placements is also very key because The impressions metric on its own has a high ability to be a vanity metric because around 90% of all impressions on Amazon go to product pages because if you just look at a product page and scroll through all the sponsored listings, there's carousel after carousel. You know, there's hundreds of really low visibility, really low click-through rate impressions on product pages, whereas top of search, you're only going to get like two, three, maybe four top of search placements, but the click-through rate difference Between top of search and product pages is, I mean, generally top of search is at least 10 times higher click through rate, but I've seen it get as high as 50 times to 100 times higher click through rates on top of search, just because the visibility is so much better there compared to product pages. So you could see That your impressions greatly increased, you know, it's like, wow, our impressions are up 50% month over month, but it's like sales are down. And it's like, why is that? And it's very likely that most of those impressions were going to product pages, which were never being viewed anyways, because these aren't viewable impressions, meaning they're not actually loading on the screen. And so you'll probably see that those click through rates tanked. And so a lot of times people are reporting on impressions. They want the client to see like, oh, we're growing impressions or whatever. But if you're properly optimizing your placements, after doing a good round of placement optimizations, which is usually going to include shifting spend away from product pages into better placements with higher conversion rates like top of search, you're generally going to see your impressions actually decrease, but your sales go up. And that's just because your click-through rate is going to go through the roof as you're pulling away from all those vain metrics. So if you're really trying to analyze a sales change with impressions, it's probably better to look at the placement level. So look specifically at all of your top of search placements combined and try to find out what happened to impressions there as opposed to just impressions as a whole for the account because there are a lot of vanity metrics. So tying in now to click-through rates, like I was saying, those placements are going to be the biggest thing. Then Andrew, what about these other two bullets? Speaker 2: Yeah, so your click through rate is definitely going to be influenced by, as you were kind of alluding to, the quality of traffic or the type of traffic that you're driving. So you can have keywords that you're targeting that are more lower funnel, long tail conversion keywords, and you have more top of funnel keywords, which are your higher volume search terms, which may have slightly lower click through rates, lower conversion rates. So you have to kind of understand which area of the funnel you're actually going after and the potential shifts in conversion and click-through rates on those different stages of the funnel. And then another thing is your actual product listing itself. So usually when people are talking about optimizing click-through rate, they're talking about optimizing their main image, changes to the main image, enhancements to the main image to improve click-through rate. But you could also have competitors doing changes to their listings that's actually influencing your product's click-through rate. So their click-through rate could go up by a change in the main image, yours comes down. Because they're capturing more of that traffic. And then also, you know, things that are obvious like price and comparison pricing to competitors and all that type of stuff is going to directly influence how frequently people are clicking on your product over another. So if you're changing prices, you could see a change in click through rate in proportion to it. Speaker 1: Yep, absolutely. And, you know, if we were to look at why conversion rates might change, it's going to be pretty much those same reasons. We probably could have also included placements in here, but we replaced placements with seasonality because Hopefully that doesn't require any definition, but products go in season, out of season based on the weather, time of year, whatever. So that's going to impact conversion rates. Once again, those types of keywords. So top of funnel, middle of funnel, bottom of funnel. If you're selling tennis rackets for children and someone's typing in tennis rackets for children, that's going to be a very bottom funnel keyword that probably has a high click-through rate, high conversion rate. But if you're also bidding on outdoor sports equipment, that's a much more upper funnel. And so you're probably gonna have a lower click-through rate, lower conversion rate on that keyword in particular. And if you're pushing a lot of spend and getting a lot of impressions and clicks behind that keyword, then that's going to pull down the average for your account. So if you are saying that your CTR or your CVR is dropping on your account, Again, in that previous episode, we showed you how you can sort by the changes, sort by delta, sort by change in impressions, change in clicks, change in CTR, etc. You're trying to find those items that had a lot of volume that was below average, which is pulling down the total average for the account. Once again, product listing changes, competitors. Product listing changes, you're probably not making changes to your main image that regularly. Like, you know, you probably do it a few times a year, if that. You probably do a period where you do some big changes and then you leave it if things are good. But the types of keywords, the seasonality, the placement settings, the bit adjustments, those are things that are constantly changing as you're optimizing and adapting. And also as the search volume trends change for different keywords. So those are the ones that you really want to drill into and try to find out. There's just a lot more volatility there. So that's going to be where you want to try to assess. Were you going to add one more thing, Andrew? Speaker 2: I was all I was going to say was that the biggest influencer on conversion rate when it comes to the product listing is going to be price changes for things like deals. So, you know, people run deals pretty frequently on their products. Competitors run deals pretty regularly on their products. So that's going to be a you're going to usually see a pretty sharp increase in conversion rate on products where you're you're running deals that can have, you know, probably one of the bigger More regular changes to an actual listing that's kind of happening within your listing and the broader market as a whole. Speaker 1: Absolutely. So with that, we're going to dive into an actual example of walking through these process, troubleshooting the sales growth or decline. But before we do that, we're going to have a quick word from our sponsors. I'm optimizing my Amazon TPC campaigns, but I could use a little help with AI. All right, AI, now get to work and optimize! Unknown Speaker: Beep boop, beep boop, beep boop, beep boop. Speaker 1: Oh no, you made the ACoS go up. Why'd you do that? Unknown Speaker: Trust the four-week learning curve. Speaker 1: All right. Is it working now? Unknown Speaker: Uh, no. Speaker 1: AI sucks. Get better Amazon performance with AdLabs. Now, I don't know about you, Andrew, but I think that was a pretty good ad. Speaker 2: I'd say so. Not to toot my own horn. Speaker 1: Yeah, that was our our first ever AdLabs commercial. It's the first of three. A few weeks ago, we were telling you guys be on the lookout for these because we're going to show you how to make good video ads because If you're going to be investing a lot of money in spending on video placements, whether it's on Facebook, on Amazon, whatever. Whether or not your video is good is going to determine whether or not that campaign is profitable. So you could spend tens of thousands of dollars at a loss because you tried to cut the budget on your video or you could, you know, spend tens of thousands of dollars with a huge profit margin because you decided to invest in that video. So that's the weakest video of the three commercials that we did. Be on the lookout. The other two are in editing, but we're very excited for those. They get Significantly better and significantly more expensive as well. Speaker 2: That was super fun to put together. Everybody just leave a comment. Let us know your thoughts down in the in the comment section. Speaker 1: All right. So now we've got an account pulled up here. So you guys have probably seen this before, but with our accounts, we have obfuscated all of the campaign names and stuff to maintain client confidentiality. But this is a real client account. I was going Throughout the whole year trying to find a month with month over month sales loss. Unfortunately, we were unable to because this account has been growing every single month. So what we're going to be doing is just showing what impacts sales growth and then just understand to troubleshoot sales loss, you're just going to pretend the negatives are positives. So in this case, sales are up 10% month over month. So it's September compared to August. And what we want to do is we're trying to figure out where did that sales growth come from. And like we were saying on that deck previously, it's going to be from either the traffic or the conversion rates. So this is up 10%. So we'll look at the conversion rates and we'll see, well, those haven't really moved, right? It went from 8%, 8.1%. So that sales growth likely did not come from conversion rates. But look, the clicks are up 11%. And so Now we know exactly where that sales came from. It came from increased traffic. So if we were down 10%, then that would mean either the conversion rate was down 10% and clicks were flat, or clicks were down 10% and conversion rate was flat, or a combination of the two. In this case, sales are up. We figure out where that's from. It's because the clicks are up. So now if we want to, you know, further try to analyze why are clicks up, the question would be, did that come from click-through rate or impressions? So in this case, we can say, well, where did these extra clicks come from? Did our click-through rate go up? No, CTR was actually down slightly, but the impressions are up pretty big. So if you were to take these, now we were saying the change in impressions times the change in click-through rate equals change in clicks. What you'd have to do is you have to do like You can't just say 16% times negative 4.5% because that would obviously give you a negative number. You have to convert these to decimals so it would be like 1.167 to signify 16.7% growth and then to show that decline it would be 0.955 to signify 4.5% decline and then that's going to get you to 11% growth. Yeah, Andrew, any other... Speaker 2: There's probably some added context here that goes into understanding this performance here. It looks like your ACoS came up quite a bit. Your CPCs came up quite a bit. So perhaps increasing ACoS target from one of your clients wanting to be a little bit more aggressive. So seeing that increase in aggressiveness on the ads looks like conversion rate didn't change too much. So the keyword allocations and things like that probably didn't shift a ton. Just kind of increasing aggressiveness on existing targets and saw that requisite increasing in traffic and sales. So, yes, probably some added context there. Speaker 1: And that is exactly what happened. Yes, this client is profitable at 30 percent. They were running pretty conservative at 20 to 25 percent and they wanted to raise the target to 25 to 30 percent. And so on the last episode, we talked about troubleshooting that ACOS and trying to figure out where that would have come from. If an ACoS increase, that would have either come from revenue per click decreasing, which in this case it barely moved, or CPC increasing, in which case it did. So now you know exactly where the ACoS increase came from. It came from higher CPCs, but we're seeing that those higher CPCs resulted in more impressions, which led to more clicks, which led to more sales. Now you're hopefully beginning to see how easy this is, but it's also just you can't you can't confuse easy with mindless. So easy and mindless. You still have to be extremely mindful of these things. And then more specifically, now when you're trying to figure out, okay, so we know that these clicks did increase, and that's gonna be the primary driver of the increase in sales. There was a 10% increase in clicks. So now if you're trying to see where that came from, you can sort by the highest clicks is one way to do it. But sometimes you're gonna have items like, you're gonna have some Guys like this that are like pretty large clicks, you know, they're they're like the fourth and fifth highest clicked campaigns, but they actually had decreases in clicks. So you would want to sort by the Delta to try to find the items with the highest clicks. And so I mean, it just so happened that a lot of these campaigns were paused. And so now that they're enabled, you know, those clicks are up from zero. So, you know, it's probably not the Best example because of, actually maybe we can just exclude these campaigns real quick. Sorry guys, I haven't done this before yet, so I don't know what's going to show up if we exclude these campaigns. Exclude campaigns, but I'm curious to see just what the metrics would be on everything else without these campaigns. All right, so this one's pretty interesting. All right, so now we have sales decline. This is actually a good example. So sales are down 24%. So same question. Did that come from a drop in conversion rate or drop in traffic? Conversion rate was actually up. So obviously it came from a drop in traffic. Okay, why did that traffic decrease? Did it come from a change in impressions or click-through rate? Impressions barely moved. It was more the CTR. So now we know that the clicks decreased. And that led to the result in sales. We can also double check the CPCs. Those were mostly flat. RPC, ACOS, that was all like mostly flat. So, you know, now we can come here and sort by the Delta. Yeah, we already have that. Which one had the largest decrease in clicks? The reason why that's so important is because if you just sort by lowest clicks, you know, you're going to get everything with zero. So you really need to find what are the ones with the biggest decrease in clicks. And in total, we lost around 5,000 clicks month over month on these campaigns. And so we can start kind of mentally tallying to try to find around 5k or at least a big impact. So here's call that 2.5, 3k, roughly 4k clicks here. So just these these top five campaigns had the biggest decrease in traffic. You can then start looking at these click through rates in which of these, you know, perhaps had a larger decrease in Now, if we look at this one, this first one here, large decrease in clicks, not too bad of a change in click-through rate, but there is a pretty huge drop in impressions here. So this is going to be a pretty good indicator that something is going on. with this campaign because those impressions are down like 80%. So from here, it's pretty clear that there was that big decrease in impressions. So what we would want to do is try to figure out where did that change come from. Was it a change in volumes? We'd have to go to the SQP report, look up for this product, the search terms that it previously appeared on, see if there were any changes to that. Were there any changes to the product listing images? Did it go out of stock? Those types of things. So you just kind of go through that process. And you also just want to look at what were the bid changes, but you've now really distilled it down. So when you see, oh man, sales are down 23%, Well, you have one pretty large problematic campaign here. One other thing that you could do too is you could also sort by largest decrease in sales. So you could actually start with these ones that had the larger decrease in sales and try to drill into those and see, okay, did that happen at the CTR level? Sorry, the conversion rate level? No. Click level? Yes. Impressions are down on that one. So this could be a pretty big one to check out. And then also this other campaign that we were looking at previously with the largest decrease in impressions. If you're going to go to these campaigns within AdLabs, it's pretty easy. You'll just go to targeting and it's going to filter down to that campaign for these specific targets. And then you can, from here, just look at which ones had the largest decrease in clicks or impressions. You could do either one. In this case, it was really just this one keyword that I think just the ACoS was really high. I mean, it's obfuscated, so I don't remember which product or keyword this is, but I think this whole campaign was just the ACoS was really high. It's probably one of our newer products, and so we were just trying to reduce that ACoS, and that's going to be why we had this big cut. Actually, that CPC only came down a little bit. That's something else you have to be mindful of too. Sometimes a very small reduction in CPCs can move you from being first place in the auction to tied for 20th place. If there's just a ton of people, let's just say there were a ton of people bidding 50 cents, And you were bidding $0.51. Cool, you're the highest bidder. You drop it to $0.47 and now there were like 50 other people bidding $0.48, $0.49 and $0.50. And so sometimes just a small change can have a big impact there. And then the other thing we would want to do is check out the placements and see what's happening here. When you do see a huge change in the impressions, it's usually going to happen at product pages and rest of search just because there are far more Far more impressions there. Click-through rates across the board were kind of up, but the overall click-through rate, let's throw it up, yeah, the overall click-through rate increased actually. It was up by 26%. The reason for that was because we just shifted away to having, you know, fewer impressions on research and product pages where those CTRs are so much lower. It's actually really not a bad CTR for Product pages, it's like over 1%. I don't normally see that. But you know, 7% up at top of search. So the overall CTR improved as we were pulling away from impressions from these other placements. You know, you can look at the ACOS here. So in this case, what we probably should do is just, you know, top search seems to be doing rather well, so we could continue investing in there. But rest of search was really underperforming. And so at whatever point when I was making some optimizations here, that's pretty clear why that impression loss came. So we lost a lot of it, 76% of our impressions, 76% of our clicks, spend reduced by 76% and sales came down by 90%. But, you know, it was like over 140% ACOS. I think the target for this product, if it's the one I'm thinking of, is to be around like 40 to 50% short term. So, you know, realistically, we can optimize this, start spending more top of search. But now we know exactly why the sales are down. So just to kind of come back one more time, you're looking at a list of campaigns. So let's filter this again by this change in impressions. We're changing clicks even. So you're looking here, it's like, man, sales are down 24%. Why is that? You find a few problematic campaigns. You see exactly what's going on. Oh, it looks like we were reducing some spend on rest of certain product pages, which was serving a lot of impressions, but the ACOS was not strong enough to warrant continued spend on there. So we reduced that. That's why impressions dropped. That's why sales dropped. And now, you know, so it's just a few. It's a small handful of select campaigns and Really just this one because that was the one that was like high a cost. Everything else is fine. We could like increase bids there or whatever to get some impressions back. So there you had it. Speaker 2: Awesome. Very well done. I mean, that's such a powerful framework. Hopefully everybody watching got a ton of value and just seeing, you know, how you walk through a diagnose, understand the performance, what the metrics are telling you and how to know how to read them and drill into them and fully understand what's driving the change within your account and then what actions, what levers you have available to you to actually pull things in the right direction, which way you want to go. Speaker 1: Yeah, if you think we left anything out, don't forget to comment. Let us know your questions. We make a lot of our episodes based off of questions you guys are asking. And as always, like and subscribe and we'll see you next week on that Amazon ads podcast.