The Exact "Revenue Impact Formula" to Prioritize Your Amazon SKUs [The PPC Den Podcast]

Host: Welcome back to the show. Mr. Mike Danford, chief strategy officer of Adverio. Adverio was born in 2014. We have had some banger episodes, Mike, over the years. In case anyone is unfamiliar with your work here on the show, I think one of the best episodes on placement testing and placement optimization we did together in January 2025. So it has been a while since you have been back on the show and got into deep placement testing. We also looked at seven overlooked Amazon Seller Central tools. That was an amazing episode too, getting more out of your Seller Central account. And we had an amazing episode on branded spend. Today we are going to continue that trend. And you brought something very interesting along today, the Revenue Impact Formula. And I cannot wait to dig into it. But for now, welcome back. How are you doing?

Mike: I am great. Excited. Thanks so much for having me.

Host: Tell me about some of these products behind you.

Mike: Yeah, so some of the brands we work with, we call it the Profit Perspective corner. Each one has kind of a unique strategy and we go through it. Things where you can do SKU replication, unique packaging, all kinds of things, just helping brands. We love creative brands, brands that have fun with their products and bring a little joy to the day. So a lot of concepts here that I reference.

Host: I have to ask about this one. What is Full of Sheet?

Mike: Oh, so that is from Shinesty. And that is a dryer sheet or dry detergent sheet. So instead of a bottle, it is a little sheet you drop in. They have a ton of play on words with their expressions and their apparel. And that is a phenomenal product by the way.

Host: The Profit Corner seems like something you go to each morning and light a candle and ask the heavens above for more profit.

Mike: Well, we have candles here if you want to do that. Profit Perspective. Our newsletter of sorts.

Host: So when people come on the show I always like to ask, what is something that has been really helpful for you? Right away you came up with the Revenue Impact Formula. There are so many numbers and metrics and corners of Amazon that it is helpful sometimes to develop these internal formulas to categorize, prioritize, and wrap your head around the whole situation. So tell me more about why this came up for you. What is your Revenue Impact Formula?

Mike: Yeah, so this came up. We work with very large SKU count catalogs. The average is somewhere between 100 and 500 main SKUs, and off of that you can have thousands of variations from size, color, scents, bundling packages, all the different attributes you can stack. And when a brand comes to us they say, hey, this works on our DTC or this works on Walmart or Target or Amazon, or how do we know which ones to keep in stock, which ones to keep, kill, or optimize? How do we know what to work on first? How do we know the current status of 6,000 products? So let us level set first. And then what do we do first? Because you cannot work on all 6,000 SKUs at the same time, not for a reasonable cost, nor can you keep full stock for 6,000 plus SKUs at all times.

Host: Nor would it make sense to dedicate the same amount of time toward every SKU.

Mike: That is the key part. How do you prioritize? You obviously cannot dedicate everything, and then AI has made it where you can do more. So now, here are your priority SKUs, your group. Can we sling some stuff at the wall over here at the same time for low value, low intent and see if more of those pop up with low effort? It is a really cool way to break that apart. AI has been really tricky for people because you can do a lot more stuff. The world is very focused on productivity, how to do something faster or more efficiently. But it is easy to do that. Oh, you want to write an email faster? Just ask ChatGPT. Boom, you have an email. But to be effective, to actually enact change that has an ROI, is a very different thing. That is where a lot of people get lost, because you can work at this rapid pace now, and if you are not prioritized on what the actual lever is that will change revenue and profit, it is very easy to get lost.

Host: Yeah. If you are natively and historically doing busy work and you pull AI in, you just end up doing more busy work unless you have that directionality and intentionality behind your prioritization on what to do, and measure the impact. Did it actually work?

Mike: Yeah. I may have used AI and it may have written a more beautiful email, but I spent the same amount of time and mental juice when I could have just written it myself. So for certain sets of messages I post now, I do not even pull AI. It allows me to get more return, be more connected, and it is more authentic. I realized I was just managing prompts instead of actually writing the right update. When it comes to Amazon, I see this all the time, this pressure to do everything all the time, right now, non-stop. The most successful people I know are really good at prioritization. So any conversation about how to prioritize is one every Amazon marketer needs to hear.

Host: Absolutely. So at the top level, how do we level set and have a rubric or an objective way to score each product, group of products, or platform, and then break down here is where you are and here is where you could be from a scoring perspective?

Mike: You roll it up and you get a dollar amount attached to that, and you say, hey, if you do this, generally speaking this is your return in dollars or percentage lift. That helps you prioritize every single time. The optimization cycle is, okay, I have a top seller, it is already a 9.9. Do I go full in to get a perfect 10? That is a lot of effort when I have 700 other SKUs I can focus on. I can do an analysis and say, I just need to put one extra image, this image works, why not put this infographic across all other 700 products? What happens when you roll that up? It is amazing. And it is fun to see. Our team internally, the fatigue and the pressure on performance, when we have an effective way to tell them what to do next, they can do more.

Host: And that is where they can shine and pull their value in, as opposed to using their energy for something that should just be an objective formula right out of the gate. One thing coming up for me is, there are all these analyze your Amazon product page tools, like Helium 10 has one, and it will score your product page based on do you have enough images, did you use enough characters. So it does a mechanical score. Then imagine you had a list of every product, pass or fail on images, then multiply by traffic potential or market potential, so you start tacking on who is going to have the best lift. Even if a product page improves 100 percent, if it gets no traffic it is still not going to move the needle.

Mike: Yeah. For us that is the listing quality score, the main thing. We have offer score, media score, content score, and two or three others that roll up into that, and it is purely objective. Can be a binary yes or no. It could be within the character limit or whatever it may be. Then there is a subjective layer on top of that, an intent layer on top of that as well. And then we have another set which is the agent add to cart rating, a readiness score. LQS is what the human is looking at, the flow, the images, the copy. Agent add to cart is what Rufus is looking at, a lot of the same things, quality images, copy, offer, but it also pulls in your inventory history, pricing history, return rates, all these things a shopper cannot see on the front end. One of the biggest things we understand for Rufus and Alexa: hey Alexa, order me a candle. It does it based on your persona, profile, and history. If it orders the candle that matches exactly, subscribe and save, boom. Now the feedback loop with the user is, man, it gets me, so I want to order again. But if it orders a candle that matches the scent and price point yet the product is commonly out of stock, so you have to re-say purchase me a new candle, that reliability goes away and the user blames Rufus and Alexa. That is the feedback loop we are seeing. So we pull all that in.

Host: That is a lot. So how do you organize this generally? I take it this is a spreadsheet of sorts with each column being some kind of score for different categories.

Mike: Yeah. We have a scoring system that goes through 180 some checkpoints in the actual listing. Then there is a subjective layer scanning your images for intent, looking at your personas, and pulling your history through buy box percentage, pricing deltas, coupons, historical return rates, and doing it against your cluster of competitors as well. If I were to sit down and build this myself, I would start with a spreadsheet and think of the columns: do I have enough images, does the image have a lifestyle picture, is it high res, an action shot, product in use. Each element earns points. And I love that you mentioned the not so apparent stuff like shipping time, buy box percentage, return rate. There are a million data points. You populate that and get a score rolled up. Then you also want ease, how easy each thing is to implement versus time, because you mentioned what if we add an infographic to 700 products.

Mike: So we have that in the back end as well, our lift from our efforts, time, cost, quality, speed. Can we do a spray and pray first? We know this generally works, one image blasted on seven. It is going to hurt some listings, not change others, and crush others, and that surfaces those up.

Host: Can I pause you right there? A specific example would be every product title that is low on characters. It is wasted opportunity. So, okay, somebody go through and make sure every product title has the right amount of characters, or images, or add an infographic. I would call that low cerebral work, because you either have enough characters or you do not. Start there.

Mike: Yep. You have 100 listings that have two images. That is very clear. Let us put brand images in there, cross-selling images, fill it up with something of value, then iterate. We have a catalog with 70,000 SKUs in one marketplace and we come up with formulas specifically for a title. We can scrape all your listings, one per parent, pull in your Cosmo and Rufus data, the 15 things in the back end plus three to seven Cosmo questions. Literally take these answers and put them in these bullets and this title, and answer it. You will see your visibility take off instantly without even putting advertising dollars on it. As those come up, here is another cohort that moved from your D tier to your C tier. Put 20 or 30 bucks a day behind them and they go from C tier to B tier. Now you know which ones work, and that did not cost you anything. It is a formula in a spreadsheet you run on the back end through your PIM or whatever software you are using.

Host: So if I were to encourage action from listeners, you do not need to start with 180 data points. Even a simple one, do we have enough images on every product, knock that out. Populate your stuff, fill in your product images, use all the characters. You do not need every metric right now, you will scaffold this over time.

Mike: And that is a really cool mechanism. It goes the same way with pricing. I just talked to a brand yesterday, a large brand. They had not touched pricing on a little over 250 total products, 50 parents, since 2023. So there is probably room there, okay we have to work on this. Throw a coupon on, take 10 cents off, add 10 cents, do something, show Amazon you are in there. When is the last time you updated images? It has been six or seven months. Okay, put a new image in. You are telling Amazon you care about this product, and it is going to reindex. That is how you refresh it and show Amazon this is not a stale SKU.

Host: So we have a couple easy examples of the spray and pray approach, getting everything up to table stakes. But the other part of the equation is, where do I get more surgical, where do I spend the mental glucose? What factors indicate where to really double down?

Mike: That is going to be on the qualitative side. We have checked all the boxes, now let us understand the quality. You look at your conversion funnel and traffic. Do I run ads, yes or no. When was the last time I ran ads. Have I updated my pricing, coupons, deals. You go through these simple things, and if you have not done them in a reasonable time, do at least one and see what happens. When is the last time you updated your title? Well, we do not, why should we, it is at the character limit. What people were shopping for 12 months ago is different from right now, and we have the Cosmo and Rufus data that tells us that. Do you answer these questions? These are literally the questions people are asking when they come to your listing. If you do not answer the question, Amazon will either say no or make an estimation and guess based on the information it has. You want to control the narrative. On your top products, are you a nine out of 10? What do you need to do to get to a 10, and can you even do it? Is it the video, how old is the video, can you do more than one, have you tried UGC versus branded. Does your copy, images, and video all say the same message so any robot or human knows exactly the intent. What are your competitors doing? Have you parented before, reparented, can you add a new variation? Parenting analysis is huge for us. In your category it might be better to stack on different attributes, put your single packs with the right colors and patterns, stack the quantities, and let them comparison shop toward the one that gives you more margin.

Host: It is very easy to forget all these things in the flurry of selling on Amazon. As we get deeper, this is half spreadsheet, half project management board, because a team would be leaving notes to each other, the last time we changed the image was this date, and that becomes another formula, days since, and that impacts scores too. Generally when I do product page optimization it is because something went wrong. What I like about this is it is much more proactive. We know where everything is, we know the status, so when it comes time we know where to focus and do not have to guess.

Mike: Yeah. I will give you an edge case. A brand started getting a frequently returned badge, because Amazon started putting that on there, and only for certain colors, and ironically not all sizes across all colors. We talked to the brand: have you changed anything, trained new suppliers, changed this? No, we have been doing this for three years, healthy consistent product. So what happened? What was a medium three, four, five, six, seven years ago is not the same medium now. People expect different sizes. And now Amazon does not even allow us to put size guides in the images, they control it. So you have to have that feedback loop. You are getting this badge, why? Is it anything you have done upstream or downstream, inventory, supply, measurements, cut, cloth? If none of that changed, you have to change the way you classify your sizes, or be very specific, loose fit, tight fit, long fit, short fit, or regroup. This one is commonly too short, so call it a short and go get another medium that is a little longer and call it the long, just like a tux for men, short, regular, long. That is a way to be creative when nothing has changed on your end except the fit people expect.

Host: That is a perfect example of needing to take action. Just to make it clear, how did the process of ranking and prioritization surface that pretty specific, not everyday issue?

Mike: Advertising is about the same, general spend, CPCs. Then you start seeing conversion drop and nothing is happening, your pricing is right. You find the date, okay it happened in April, what happened, the badge hit. A good badge or a bad badge can impact your sales equally. How do we get rid of the badge? It is your return rate over 30 and 90 days historical. But it is not going to go away unless you change something. Then it goes back to what is the lift. Come to find out you used a chart for your sizes, measurement across the chest, shoulders, waist. I just got fitted for a tux last week, same body weight for 10 or 12 years, but my measurements are different. When you put the same shirt on a six foot male, 185 pounds, and show how it hangs on them, people identify with that, and you stop getting returns. That surfaces from what has changed, then what can we control, then what is the lift for that control, and then you prioritize it across hundreds of products.

Host: The last piece is, now that we have surfaced information and have the spray and pray, what determines where the lift will be greatest, where the biggest leverage is? Say you had two products, same thing going on, conversion dropping. Which one do you pick first?

Mike: So we have a model. We have a little over 350,000 SKUs on Amazon US and we take all these numbers. We can look at what a score of a six or seven or nine or 10 is and understand the delta. Here is where they were at a six, we make these changes, they are at an eight, what was that lift percentage wise, and you roll that out to the anticipated lift for that product. That is how you get the dollar amount to understand what to prioritize, even at an individual SKU. There are three things you can fix on the SKU. That is the model, and it is dynamic thanks to the AI and the large models, constantly iterative, and it rechecks so we are not using old analysis and weights. It does shift over time. Click-through rate, everybody is about main image optimization. There was a huge lift 18 months ago because nobody was doing it. Now everybody is, so the lift is less. And does improving click-through actually increase conversion? It gets more people to your page, you will probably spend more on ads, but is it relevant, quality traffic, did you convert? Sometimes click-through improves but conversion does not, because the quality of traffic is not there. So all that has to be pulled in.

Host: Do you factor total search volume of keywords per product, or market share percentage per product? Because if you have a product with a tiny share of a niche and bad scores, fixing those could potentially double revenue because there is so much more share available.

Mike: Yeah, so internally we have QueryIQ, basically your search query intelligence companion. We can pull TAM and relative market share out relative to the individual search. Then we have clickiness and scrolliness, another factor, people click and buy and go back and forth from this product a lot before they buy, or they click three times and buy, high intent, boom, that is where we want to go. And how far will they scroll down the page. You do not have to be number one, being number 10 is okay because you are getting a pretty equal share at that point, so do not keep pushing and burning budget when it is too cost prohibitive. Then we compare your share versus the market share over time. Have you taken it, have you lost it, has the total market shrunk, and it is on a weekly basis on a search query performance roll up.

Host: Amazing. One watermark I have for every show is, did we hopefully inspire people to take action, and I think we hit the nail on the head. If someone is listening and is not excited to have a list of their products and begin to quantify and rank where to prioritize, I would worry they might be stuck in a productivity loop, doing more and more without seeing where the low hanging fruit and the big leverage activities are. So I think we did it, Mike.

Mike: Yeah, love it.

Host: We have links to Adverio in the description. Thanks so much for coming back on the show. When you are not optimizing, what is something personally you are excited about lately?

Mike: I just returned from a Euro trip, a week in Greece and then a few days in London on the way back. When I travel like that I like to go analog. I do not want to pay for the travel pass across three phones for 12 days. I want to learn and get immersed. We always pick a local spot, get close to the tube, go the same way, know my three or four intersections, and just walk. You get to really experience the culture. And being analog, I did not do any work for the 12 or 13 days. It also stress tests my business, my flow, did I give the team enough work through a sprint, how did it go. This is where the topic we talked about at the beginning came from. It is a nice reset. In this highly digital world and this AI ubiquity, get out in nature. Hopefully you do not trip and break your wrist, but get out there anyway.

Host: What was the best thing you ate on that trip?

Mike: I ate my body weight in gelato, it was everywhere. The Greek food was absolutely amazing, alfresco every night. I am a big sucker for grape leaves, they call them vine leaves there. Probably equally my body weight in tzatziki. And something new, we have done ouzo, so now it is at the house as a digestive after dinner. The biggest thing, I love to eat and I train intensely, I am a hybrid athlete. I did not train there, only ran a couple times, gave my body a full break, but we did 20,000 steps on a daily average, about 10 miles, and I did not gain any weight, and I ate way more than usual. That was a big epiphany for me.

Host: I am thinking of my Greek trip, sitting down eating tzatziki and pita, one of my favorite meals, so good. Well, Mike, thank you so much for coming on. There are links to Adverio in the description. Everyone else, go build your ASIN prioritization list, and we will have you back on the show in a couple months.

Mike: Awesome. Thanks so much.