Behind the Scenes of Live Image Optimization

Behind the Scenes of Live Image Optimization Speaker 3: Hey, guys. Welcome back to Seller Sessions. We're back here again doing Main Image Monthly. A little bit different today. We are going to go through and break down the product. We're going to do the testing, etc. But what we're going to do, our stop the scroll merchant, Dorian, today will be doing a bit more background work. So you'll be able to see in real time some of the adjustments that he makes in how he builds up the catalogue of images. He'll also introduce you to some new concepts. And yeah, we're just trying a different experiment for the layout to make it more interesting. I don't think this kind of thing's been done on the podcast before. So, without further ado, I'm going to bring Sim into the conversation. Hey, Sim, how are you doing? Speaker 4: Good, mate. Thank you. How are you? Speaker 3: Yeah, look, you always pick products every month. Do you want to explain what we've got this month, where it comes from, concept-wise, etc.? Speaker 4: Yeah, sure. So, it's not always easy to pick a product because, you know, I've run out of friends and not everyone wants their, you know, ace in on YouTube. But luckily enough, I know one that's Andri runs an aggregator, essentially, and was very open to the idea. And actually, technically, Andri's boss, I think, for another brand. So that's pretty cool. Last week, we did the board game, Grab the Mic, which was, I think, one of the best-selling board games in the world over December. But I think there's no hyperbole there. That was actually the case. And this one we're doing today is the number one teeth whitening pen in the USA as per the browse notes. So it's got the best seller badge at the moment. The reason that we picked this one, one, because if we can lift conversion rate, if we can lift CTR, then obviously the numbers involved in this product are enormous just because of the scale of it. But also it, I think, Quite obviously, there are things that can be done. When talking with the owner of the brand, he said that they've tried things before and that obvious changes that normally would work on this product for whatever reason, when they went to manage your experiments, it performed quite poorly against this original one, which was a surprise to them and, well, frankly, us looking at it now. I think what we've got today is a very high-level podcast. We've got one, a product with a lot of scale, and two, I've just seen backstage what Dorian's about to produce, which is he's opening up his playbook. This is going to be a very unique podcast, I think, where we're seeing the in-depth The process of how a company run by Dorian would work and it should be very interesting and a lot of action points for people off the back of it. Speaker 3: Excellent. I'll bring Dorian in quickly just to say hi. Dorian, give us a couple of minutes breakdown of what we're doing and then I'll bring Andri in because we're going to look at the tests and stuff, pre-tests that were run first, aren't we? Speaker 1: Sounds good. I mean, the focus today is really to show everyone the process from A to Z so you understand what kind of elements come into place when designing main images and actually the possibilities of how much you can do and how quickly you can come up with hypotheses. A lot of people think that Changing main image is this drastic, massive thing that you have to do, like come up with this new main image. But what I will show you today is how to break it down into what we call atomic design steps, where you slowly change one thing at a time, you progress, and obviously, based on the data, well, we're kind of working with Andri here in Productpinion, we get a lot of data, and we can slowly iterate and develop Something unique and slowly at the same time. We're not jumping too far and that's quite important because sometimes you can find, you know, improvement in CTR very quickly and you don't have to change the whole concept drastically. So I'll show you a couple of ways how to do it. Speaker 3: I'm not sure if it's a technical term, but we call it breadcrumbing. So you're leaving the trail, right? So that you can go through all the different stages of iteration. I'm going to bring Andri in now. Welcome to the show, Andri. Speaker 5: Hey, thank you. Happy to be here. Speaker 3: Excellent. So let's get straight into the show today. There's less of us involved and we want to save a lot of time for the experiments and stuff with Dorian at the end. So let's bring your slides up and then we'll get into it. Speaker 5: Let's do it. So I'll explain the whole testing process the way we've done it. We're working on this with Dorian. So first stage, we try to start with qualitative testing, where we essentially have the most unfiltered feedback from people in the form of video. So let me bring up a video file next that we can show and listen to what people say. Speaker 2: The first one that stands out to me is this snow gold-coloured one. Because of the gold colour, most teeth whitening products like these have the bluish background. So I'm not really drawn to this one. I don't think I would buy it just because it looks like an unknown brand. I probably wouldn't bother with this one because it looks like you only get one tube, whereas all these others you get at least two, it seems. Maybe not this one. As many ratings and reviews as the one that caught my eye second. Also because of the gold colour, which is this Aura Glow. Speaker 5: The other ones that caught my eye were this one, the Snow. Speaker 2: I really like the packaging. Also, it looks like this is a two-pack package, which is really nice, so you're getting more bang for your buck. Colgate catches my eye, but it's really because of the colours, not necessarily because Colgate is a great brand or company. Most of these other ones are just knock-off brands. That's pretty common on Amazon these days. From the ones that I'm seeing, if not Colgate, I might go with the Smile Shark one, just because of the graphic. Show some teeth, the ones that don't have pictures of the teeth I'm not as drawn to. Most important features are hydrogen peroxide level and the before and after results for me. My most important benefits and features would be the amount of peroxide which I don't see listed easily on any of them except for this auric. What I like about this one is it says 9% hydrogen peroxide versus I didn't really see that listed on the other ones. The 9% hydrogen peroxide like I don't know if that's good or bad or what. The first one we have right here is auric glow extra strength teeth whitening pen 9% hydrogen peroxide so I know hydrogen peroxide is really good for your teeth. I really like this one too with hydrogen peroxide and I like the design of the box and the pen as well. What I would improve is maybe put some teeth on the main image. The third one That I would look at is this Lumino, Lumino one. It does stand out as well. It is blue, but it's got this kind of, it looks kind of premium. I think I would try the snow one first. Even though it's more expensive, I guess the marketing works because I know they're all the same thing. I would probably click this Colgate Optic White Overnight one because it's overnight and I feel like that's pretty easy use and you get results in just one week. This looks very tacky, I would say. They could have used an actual picture and this also looks just very computer generated and doesn't really sway me. However, I scroll down and I check the reviews because I want to see real people's results. And this is Already Good. For me, the most important features are they don't hurt my gums and my teeth. Speaker 5: They're good for sensitivity. Speaker 2: Here it says right on the box, enamel-safe, peroxide-free. I like that. It doesn't cause any sensitivity. And here, I like that it's formulated and it's clinically proven. So this is one that I would definitely use. I don't really see any information in terms of clinical studies or a formula that's been approved by dentists or orthodontists, which is important to me. Again, would like to know if this is clinically tested, proven. How did it get to these claims? Here it does say that it's not been evaluated by the FDA. It doesn't use any harmful ingredients and it's non-toxic. So would love to try this one out. I put them on these because the tubes look interesting. Packaging really can make a difference. I think that it can make a product look more interesting. This photo is nice. While it doesn't look like she's really using the product, it just looks like it's been photocopied. This one looks much better. It looks real. It looks like she's really holding it. I really like the look of this product and I'd want to get something. I'd probably pay attention to the flavour. Some of them mention mint, so that might be something I would look forward to. Speaker 5: Yeah, lots of great insights from that one. So essentially what we do next after qualitative testing is quantitative testing, so collecting the actual data. And the beginning of everything is seeing where do we actually stand right now. So the baseline test is this Amazon search simulation test where we are right here and we excluded all the big celebrity brand names specifically for this test so that we can see how do we stand amongst everybody else. And this is where we are. Baseline is 7.23%. Essentially, 7% of people looking at us and clicking on us. Okay. And quick question within the opinion GPT explained that most people still mostly rely first and foremost on reviews, price and rating. So then we spoke about this with Dorian, figured out, maybe we can try this crazy idea. What if we pretend all the reviews, all the ratings, all the same, and everybody offers the exact same two pen in terms of quantity. So two pen in one package. Okay. Could we learn something from this? Let's see. So I just named it the skewed baseline task. It's not the real, real situation on Amazon, but we still use the Amazon Search Simulation. So this is what it used to look like. Okay, pay attention. This is what it looks like now. So again, back to what it used to look like. Some of them have three pans. Some of them have one. And pricing is all over the place. Reviews are all over the place, right? But here, everybody's the exact same. OK, now let's see what happens. We exclude the previous participants because they're biased. The baseline previously 7.3%, new one where everyone's the same, just different image and different title, we're slightly more, 8.73. But what's interesting, now that I ask Pinion GPT to analyse all the responses, we realise that the criteria people have at this point are mostly whether it's dentist approved, whether it's safe, and how they perceive the quality from the packaging and the design. Okay, that's where we play. So what matters most again, dentist approved, they want to trust some authority, now safety assurance, NML safe, no sensitivity claims, and premium packaging. Okay. So next stage, we develop new concepts and we test them. OK, so again, this is the original. This is our starting point. This is the base image. And these are the initial concepts Dorian came up with based on all of that research. And now let me ask you, which one do you think won? Look at these. Speaker 3: Sim, you go first. Speaker 4: Yeah, sorry. Yeah, my eye is drawn to the one in the middle. Just the size of the mouth is making me look that way. These are quite significant changes, though. If you compare the baseline to these three already, I know we've got, you know, plenty more to come. So, yeah, the middle one for me. Speaker 3: I'll go the same as Sim B, obviously, because that's the biggest picture showing clean white teeth as well. So it kind of says what it does on the tin. But it's testing. Who knows? Let's see. Speaker 5: All right. Let's show the results. So you're right. The B wins because they like we show the teeth here right there in the middle in the center and gets your attention. Now, These are the actual results. So it wins by a lot. And obviously, people explain it with the way you explain it, they wanted to see the results, what they actually get for what they pay. And we did test it on the mobile version, because we figured based on the business reports, at least half of the purchases from this brand happen through mobile devices. And do you remember that SKU baseline test results? Same pricing, same reviews, rankings, all options are two-pan offers. When we looked at the top three, the top choices were looking more like makeup brand products, not necessarily teeth hygiene. So that led us to check this quick statistic, the brand analytics. So I got the brand manager of the brand to check this for me. And essentially most of the purchase, so like twice as many, are females. Which led us to developing a little more polished minimalist style packaging. Dorian will dive into the whole process here later, but this is what we're playing with next. Now, what are your thoughts? If we test these, which one's going to win? Speaker 3: Sim? Speaker 4: I think A, I think it reads slightly easier and it's a little bit more feminine without the black. Speaker 3: Yeah, I agree because we've seen this come up in tests before, haven't we, Andri? We spoke about it on the last show, is that often you see guys that like a black sleek box and ladies would react to that slightly different, said I'd like to see more colour options and stuff. And I think, again, if you look at the, like Sim said, it's a bit masked in where it says dentist on there. The background and the grey and that kind of gradient effect and stuff. It doesn't really stand out. I mean, saying that on the other one as well, it doesn't stand out hugely, but it's a glaring difference between the two. So I'm going to go with A as well. Speaker 5: Yeah, I think very fair judgment. And for these reasons and more, A wins. And it won by a lot, like 98% certainty. Females essentially pick the one that you explained they would like better. Speaker 3: Yeah. Speaker 5: Now, what do we have? We have two strong potentials, right? We have the minimalist style and the crazy one, like the insane option on the right has everything thrown at it. What if we make another battle and split as these against each other? Do you want to guess which one wins this time? Speaker 3: Before I do, what is the split of the male-female group? Speaker 5: The exact same, all females. Speaker 3: All female? Speaker 5: Yeah. Speaker 3: Not a mix. Okay, I'm going to go for A. Speaker 4: I'm going to go for A as well, but just based on colour. Like, that's a female face aimed at females in B, so that should be the one. But yeah, just based on the colour alone, I'll go with A. Speaker 5: Yeah, that was actually my assumption as well. But actually, neither. They're essentially 50-50. Okay, so 50% of people like the one, 50% the other. So undecided results. So at this stage, I'd love for Dorian to dive into his process up until this point, how we got to these results, and then we'll jump back and finish the testing process. Speaker 3: Dorian's up next. Let's bring up his screen. Speaker 1: Okay, cool. So what I want to show you guys today is an entire process of how I quickly come up with ideas that are data-driven. And how quickly you can iterate. Here I'm using a software called Figma. So I'm going to run you through every single iteration and the decision behind it. So this will shine some light on what's needed in order to develop concepts for new main images. So the first step I always do is collect main images from Amazon search page, right? The second thing I did was just a very quick Keplow report that showed me As Andri mentioned that most customers that leave reviews are actually female and they're looking for these five things. In those products, which is teeth whitening, state removal, bright smile, convenient application and effective results. So that gives me a quick idea of what I'm looking for in the main image. Then the next thing is myself and Andri here, we looked at the video tests. And then I like to very quickly, just using GPT, start summarising everything. So at this point, what I love doing is the first step, which is just choosing a colour palette. Very simple. So these are the current brands out there. And I'm just looking at the colors palette that they're using. So obviously, we can't deviate at this stage too much from the original packaging. So I kind of have to stick with the blue. So the first thing I do, and I'll show you guys how quickly it is, is just to set up a quick gradients. So by using Figma, what we can do, and I do it very quickly and easily, is just grab a frame and apply the next Welcome to the next iteration to your design. And by doing this, as Danny said, I'm leaving breadcrumbs and these breadcrumbs are really useful because I can go back to them and see if something worked for me previously. I can very quickly go back to that and reuse elements from different types of images that I've created in this test, right? So the first thing I did here is just created a point of focus, just a simple, simple circle. So it's really important to understand that the first In this version of a test, whenever we run a test in ProductPinea or manage experiments, we're not looking for details. The first thing we're looking is for an impact. If I can create an impact with the main image and I can see a spike in click rate, then I build the confidence in that direction and then I'll start tweaking smaller and smaller things and start working on the details. There's really no point at this stage to dive into details. So what I'm doing here is just very quickly I'm mocking up some points of focus. So I've decided to use a circle. So once I'm kind of set on the idea of using a circle, very simple point of focus, then I've started to develop some assets. And I use what I love doing, what I create assets with at this stage is just quickly using mid-journey, created some ice blue backgrounds and some circles. What I also did at this stage, just reshot the product. So just a very, what's really important, you want a very crisp and clear product image. In the original main image, that wasn't very, that wasn't great. So I just reshot that and quickly edited in Photoshop. And then very quickly, you can also pull some We're going to talk about box mockups from a place like Envato and just to see how the box... I usually download a few mockups and see how the box looks, how it fits. Can I create a composition with it? Great. So at this stage, I'm set. And I'm starting to kind of create first initial ideas of how it could look like. So that was my first initial idea is big focal point. And here we have the original next to it. So you kind of see the difference. So again, for me, that's I started to look like something that I would like to test. In here, what I've quickly put here is enamel safe formula. That was really important. Up to eight shades whiter. So something that's catching attention is sort of a clickbait. And then what I do is very quickly just drop it in between all the other ones just to see. I kind of like to zoom out. I call it a squint test. I like to zoom out as far as I can and see Is there an impact? Does it catch my attention? At this stage, I wasn't very happy with it. What was missing is a bit more contrast, a bit more colour because we are up against these red, beautiful main images. So what I did here very quickly, just added the dentist tested badge and this red stripe here saying two pen. So when I zoom out, That kind of stands out. I was happy at this stage. So we started testing, right? So as you saw, Andrea and myself started testing. Then we decided to start testing more images by showing some images, right? So again, very quickly, what I can do here is pull another Frame to the side and change the image very quickly. So by doing this, I'm very quickly just creating as many designs as possible, zooming out, comparing and say, okay, I like this one. So then I, again, quick mockup, send it to Andri, product opinion, do a test and then learn something new. And that's, that got us to this stage. So when we, the next stage is We're trying to go in a slightly different direction. So it's always best to have two type of approaches. One, which is, as Andri said, the crazy idea. And the second one, which is slightly in the opposite direction. So what we realised here, as Andri mentioned, is you see those top three? They look like beauty brands. Right. And it makes sense because obviously it's a female based kind of mostly female purchased product. So things like this will catch attention. So what we did then, again, pulling frames and very quickly creating different type of much simpler Much simpler compositions. Playing with text, playing with layout. And once I'm happy, sending it to product opinion for a test and keep going and keep iterating. So that's great about this process because you can take this, go to the next step and say, okay, now let's try a lighter background, right? So I can go here and I can change the background to, let's say, purple. So at this stage, obviously, we had these three main images that we've tested. And I'm just going to tell you a little bit more about this. But before I stop, I want to show you something that's also really important. What we have realized, it's not just about the outer packaging. This is a concept that I quickly developed with Andri. It's not just about the outer packaging. What's really important, as you can see here, in the product that was really catching the most amount of clicks, it wasn't just the outer packaging. It was also the design of the product itself. So what's really important is also when you're developing products and you have And you have the opportunity to improve the design of the product itself. You should take that opportunity and test that also using Product Opinion or other softwares. Because that, in my opinion, is why this main image is capturing so many clicks. Sorry, because it's not just about the outer package in the box, but also how the product itself looks. So here's a quick concept how it could look like if we decided to change the design of the product itself. And that's pretty much all from me. If you guys have any questions, let me know. Speaker 3: Cool. Sim, do you have any questions? Speaker 4: Yeah, so when you're iterating and you're going through each one, some of these changes are quite minor. What's a sample size that you trust? Is it 40 people? Are you doing 100 every time? How are you getting enough data while keeping the cost down? Speaker 1: Yeah, it's usually around 100, the sample size. It's usually around 100 people. I wouldn't go less than that. At the end of the day, it's worth the money, in my opinion, especially for products that have a lot of session. You can build a confidence in your main image with three to four tests. We're talking about, what, 300, 400 bucks? If that main image is going to generate 20% more clicks, 20% more sales, For me, it makes total sense. Speaker 4: It's not just the votes either, is it? It's the comments and the other insights you can get from that. I agree. Speaker 1: Exactly. That's the thing. It's not just about the image itself. As I was showing, the final concept, when we change the way the product looks, This when you develop in this main these main image concepts. You'll get a lot of data, a lot of data and a lot of ideas that will come to your head. So you'll start questioning things about your product design. I'm seeing that a lot of people actually want it to be doctor formulated. Can we reach out to a doctor and maybe improve the formula and get certification? FDA's approval is really important for everyone on product opinion. Wow, I didn't realise so many females were buying my products. Wow, I didn't realise that way they were drawn more to the design that imitates beauty brands. Can we do that with the next order? Can we actually We're going to change the, not just the outer packaging, but actually redesign the product, the way the product looks. Maybe we have to, maybe here's an idea for new branding direction. Maybe let's not, let's, maybe let's go more in the beauty direction visuals, right? Maybe we don't have to be such a kind of a brand that is very cold and looks like something you can get in a pharmacy. Maybe we can make it a bit more fun. So just by doing these iterations and tests, There's so much to learn, so much feedback. So don't think it's an image exercise only. I think it's an exercise in product design and branding also. Speaker 3: I've got some questions because it probably helped the audience as well because you know inherently when to go through these stages of the testing, right? But let's first talk what software availability is out there to do the atomic process that you've run through. So you're using, what is it, Figma? Speaker 1: Yeah, we use Figma. I think it's the best one out there. Speaker 3: Yeah. Speaker 1: I know a lot of people use Photoshop. The problem with Photoshop is you have a lot of files, you have a lot of layers. It's really hard to show. It's messy. How if, you know, if, you know, Sim, you are the client here, I wanted to quickly show you all the ideas and we can do it today on this call is what do you think of this? What do you think? You know, can we change this? It's I love Figma because it's a visual whiteboard. It's a digital whiteboard. The four of us, we can stand in front of it. And say, OK, how about this? I like this. I don't like this, right? With Photoshop, I'd have to open layers, close files. Speaker 3: It's a mess. Speaker 1: It'll probably crash 10 times. So Figma is the way to go. Speaker 5: Another quick one. I don't know if you tried it, Dorian. Some people use Miro, MiroBoard as well, M-I-R-O. Yeah, I think it's pretty similar to Figma. Speaker 3: And look, for listeners out there who's getting started with this process, you've got the free versions. You can do most of that in the free versions to kick off to see If it's something you want to do in terms of building out the process, the other thing is testing. So you go for a series of iterations, but you know inherently when to stop and when to start testing. If you were to give people listening now a framework to do this, How far do you get with the iterations, then test, and the next set of iterations, then test? Just as a benchmark so people know not to go past the point of winning with the iterations and testing too late into the process. And what are some of the questions that you ask? When you're going through the iteration process, because it's not the same as a complete finished image, because the questions you ask is important to get the answers back that you're looking for to improve, right? It's how you phrase it. Speaker 1: Yeah. Test early. That's what I would say. Test early. We usually test As quickly as we realised we have something here, like we created an impact. So what I love doing, as you guys saw, is just paste your image in between all the other images, zoom out, do a screen test and say, okay, do we have something here? Like, does this actually stand out? Does it have a chance? If I feel it has, let's say 70% of chance, I would say, let's go test it. So test often, test early. A lot of times you can be surprised. So there's no point early on to really paying attention to detail and being very, giving your main image too much TLC because you're going to spend a lot of time and you're going to learn that Actually, that was the wrong direction. Speaker 3: So build some confidence And just go upload it test it And then when you do that testing and you get the information back What are you look because it's not a complete main image What are you looking for in that stage two after the dates come back and the feedback of the first iteration? And is that different to the third and the fourth if you depend on them and you move through? Speaker 1: Yes. What I usually look for are patterns. So I look for patterns. If someone's mentioning a lot of the times that, you know, doctor formulated is important or that the amount of peroxide is, safety is important. And I see it time and time again, I will double down on it. Because the thing is, as you guys saw, we had this one crazy main image where everything seems to be important. It's eye-catchy. From the design perspective, it wasn't the prettiest one, right? Everything is important. It's always good to test these main images. But at one point, at some point, you have to decide what's really important. Because if everything is important, nothing really is. Some of these main images that were winning, as you guys saw, were very simple, because they understood what's the essence. What is it that one or two or three things that customers are looking for? And they left everything else out. So that's the point of iterations. First of all, I want to put everything out there and then slowly learn what's important and look for patterns. So if I see no one's really mentioning that, I'll put this keyword, clickbait, that information out of the main image. But someone's really mentioning this a lot of times, I'll double down. It could be a keyword. It could be information or a color or a style. I'll make that more prominent and prominent. At one point, it becomes an exercise of deduction. You keep deducting things and double down on these things that people really are drawn to. Speaker 3: No, that makes sense. Sim, any final words before we move back into Andri's final part of the presentation? Speaker 4: I guess we can get to this later, but the title of the product says award-winning on there. It's not mentioned in the secondary images or the main image currently. But again, you're trying to stand out here and show the efficacy of the product. That would be great to have on the product itself, on the main image. And then second is just a minor note that, well, Figma can do a load of stuff. You can actually make the edits in there and move things around. But for mood boarding, What we do is we create whiteboards on ClickUp. ClickUp have really upped their game. ClickUp 3 is an exceptional product and great value for money for any business, really. They're not paying me to say that, but I just think it's brilliant, and the whiteboard feature works very much like Figma. You just can't obviously edit the images. However, they do have a Figma plug-in, so you can sort of run the whole thing through ClickUp. I love ClickUp. I mentioned that because a lot of people use that anyway in their day-to-day business. Speaker 3: Indeed. All right, let's get back into the second part of the slides with Andri. Speaker 5: Yeah, great points from everybody. And I think everyone wants to see the final results or actually the final stage. So let me dive into it. The last step of this whole process that we were working on We just call it contextual testing baseline comparison. OK, so we compare the previous baseline to new ones because then we want to play within the same context, which is Amazon search results page. So back to where we left it. These two were like 50-50 undecisive results. All right. And this is the very beginning, the baseline test. This is where we stand and our baseline Market share or click share, I should say, was 7.23%. Okay. Now, what if we do the exact same test? So real life situation, no edits, same review, same pricing, same offers. But now we use a different image. So what we also want to do is make sure we exclude previous participants because they're biased. They've already seen these products. We want fresh eyes on this. And let's test this concept A, the minimalist one, the one that won amongst the minimalist options. All right. So this is us. Again, this is search simulation. People see this when they provide feedback. And this is where we stand originally 7.23%. Now this is where we got with this minimalist one. So slight increase 8.79% click share, a little bit of an improvement, right? But in my opinion, slight improvement, not conclusive. It's not decisive enough, not a huge difference. Now, why don't we test this concept B, the winning insanity option, the one that has everything thrown at it, right? The crazy one. Let's see how that one stands out amongst the search results options. And again, the original 7.23 percent. The exact same number of people, the exact same situation. The only difference is the image. Boom. The new baseline for the concept B is 11.28%. So let me go back to the original. That's where we started. That's where we are actually as the brand. And that's where we could be. What does it mean really? So this is the base 7.23%. The winning currently concept B is 11.28% click share. That's like 56.1% higher click through rate. And in real life numbers, that's 113 more people actually go into your listing based on 1000 impressions. So out of 1000 people, 113 more end up on your listing, potentially with this main image. Okay. Now, we kept reiterating because this is a never-ending process. You can't just stop. You're never done testing. It's like brushing your teeth. You have to do this all the time, right? So we thought, what if we marry the two? The minimalist version that a lot of females were drawn to and the one that actually gets more attention because it stands out, right? So the baby could look something like this, right? Option C on the far right here. So, again, we haven't really gone that far, but this is the whole process of how people in our space should approach this. Continuously testing and iterating so that we can get to slightly higher percentage. Essentially, the rule is always be testing because you're never done. That's pretty much it on my presentation here. Speaker 3: Cool. All right. So that's the second part done of Andri's presentation. Let's do something a little bit interesting now where we'll do live feedback. Dorian, I'll bring up your screen. And then what we'll do is we'll go through this together to get some ideas and feedback to Dorian as he's working in real time. Speaker 4: First point I had was we heard in the video feedback from Product Opinion people that hydrogen peroxide and the amount that's in there was quite important. Looking at the secondary images for the Venus Visage, it looks like you're using a different chemical. Is there a reason why? If it's better, we could talk about that. It's difficult to get on the main image, but that clearly would be something that would resonate. Speaker 1: That's a good question. So, Andri, do you know if this product includes peroxide? From the looks of it, it doesn't, right? Speaker 5: I believe it does. I'm not sure about the percentage. I remember one of the options had 9% prominently stated on the packaging, which I may be wrong, but my understanding is the lower, the less invasive it is, but at the same time, it might be less effective. So it's a tricky question, what do you want to show? But I think what it may help people with is some scientific trust and that kind of claims. If you edit, they trust you more because you obviously had a study or something like that. Speaker 4: Yeah, it looks like it's carbamide peroxide compared to, you know, hydrogen peroxide, which I think a lot of the others use. But yeah, if there's a reason why I would, I would definitely try and do something there. Speaker 5: Good point. Speaker 1: Got it. Got it. Let's look at some of these original images that we had. Anyone mentioned in peroxide here? No. I remember we had one, right? This is the one that was mentioned in 9%, right? 9% hydrogen peroxide. Okay. So I think sometimes we have to find a deeper reason why people wear I was drawn to that information. And the reason why I didn't put it here, because one, I wasn't really sure about the formula, but for me, it was more important to mention that it's enamel safe formula. So what I wanted to make sure is two things. One is that your teeth are definitely going to get whiter. But at the same time, it's a safe formula. So that was kind of a workaround of this issue without actually letting everyone know what the actual number is. But if we are certain what it is, we can try that. Speaker 4: Yeah, I guess the other way of doing the eight times whiter as well would be well, how long does that take? Am I going to get that on this one batch of two pens? Is it going to be done in a month? How long does that take? Because, you know, if this takes six months, people aren't going to be as keen. If it takes a month, it's, you know, people want things now, don't they? So if there's a time frame we can put on that, it could definitely help. Speaker 1: That's a great one. So that's something, again, we have to understand and be confident of because we don't want to make false claims, right? But at the same time, we can try something here. So up to HR, ShadesWriter, and we can say in just two weeks, right? It could say something like this, and we could potentially test that. This is how you come up with another hypothesis. As you guys can see, I simply just put this to the right hand side and we came up with the one hypothesis. Anything else that we could try? Speaker 3: Scroll stopping in terms of the wording as well because, you know, safe is also associated with being safe, but it's also boring as well because you've got bright white and then it's enamel safe formula. Is there something that would be, you know, you could run through ChatGPT to come up with something a bit more compelling because you've already got up to eight times shades whiter in just two weeks, right? But is there something that could be played on those words? Because the other words really stand out. Now, obviously, we discussed earlier on, everything can't be important, because then it becomes less important or not important at all, right? So there has to be an order. Is you trying to think of is there something that could be more compelling which isn't presenting the truth? Does that make sense? Speaker 5: I think one of the things we could play with potentially because there is an award for this product. I'm not sure which one but even if we say enamel safe or whatever we pronounce it, enamel safe comma award-winning without saying that the award is for being extra safe. Speaker 3: Or award-winning enamel safe formula. Speaker 5: Then if it is, that would be the best. Yeah, maybe that's what it takes. And then everyone's like, oh, everybody else is not even in comparison because they haven't won the award. Speaker 3: Because we're stating the obvious. Yeah, bright white. Yeah. But we're already seeing that the teeth are there. So we could maybe use that top line text that we've got, which is in quite a large font size to bring in the award winning thing. Of course, like, how many people's teeth are not going to be white? So we're stating the obvious. We'll drop that down. Speaker 1: We can have them side by side, right? Speaker 3: Yeah. Speaker 1: Great. Another hypothesis. Speaker 5: I had a crazy idea that Dorian turned down. It may not make sense, but what I thought could be cool to show is the before and after. So half of the teeth would be slightly more yellowish, which I know may not look as attractive, but could show the A to B journey of the customer in one simple image. Speaker 1: I had this image somewhere. Let me just find it. Okay. We had this here. I'm going to pull... See, that's why... Speaker 5: That looks scary. Speaker 1: Exactly. And that's why leaving the breadcrumbs is so important because you just said that and you came up with hypothesis and I can go, oh yeah, I tried this. Let me just copy this. I'm going to pull it next to the side and we're going to do a quick iteration. So this is the image that was We had before, so I'm going to pull this one, right? And Andri wants to see if we can try. Speaker 5: Dorian, before you do that, what if we use the exact same smile we have on the left? Speaker 3: Yeah, I was going to say, because it's not that those teeth are bad by any stretch, right? But I think aesthetically, the award-winning, shall we call it, teeth are more structured, right? And the mouth isn't kind of tilted either. It's like... Speaker 1: Gotcha. Speaker 3: And can you zoom out that image just slightly as well? So you've got more of the shape of the mouth rather than like superimposed on the lips. Speaker 1: Not this image. We'd have to look for a new one, which would take some time. Speaker 3: All I'm thinking is aesthetics, right? So you're presenting aesthetics and there's nothing wrong with that person's mouth or lips. But if we were to compare, I think aesthetically, the award-winning mouth And teeth and lips look better than that close-up shot and the structure of the teeth. Because remember, we're adding the yellow now. We don't want to put people off that are flicking fast. And then you remember, if we look at the Google test, they look from the top left-hand side and they zigzag down, right? So the first thing they're going to do at the corner of their eye as they're flicking through the scroll, if they're on mobile, is they're going to see that yellow. Do you know what I mean? And they might not look across far enough. And that might be something while you're showing an improvement, it can also put shoppers off as well. Speaker 4: I agree with Danny on this. I think it'd be a great first secondary image to show the before and after immediately. But yeah, I'm a bit concerned that this would look unappealing. Speaker 1: Yeah. And that's the reason I didn't go with this one is because for me, showing this beautiful white teeth I feel more drawn to it. I think a lot of people would find this maybe disturbing a little bit. Why are you showing me yellow teeth? That's not what I'm looking for. I get it. We want to show the process, but as Sim said, the place for this image will be much better in the listing itself. Speaker 5: Yeah, I'd probably agree with that. Speaker 1: Okay, any other changes, you guys? We can pull another one next to it. Let's try. Speaker 4: So the 8x up to 8x shades whiter in just two weeks, it is worthy. And we've got two numbers here. So you could do eight times whiter in two weeks. Dorian, what are your thoughts on that? If I use number eight, I want to use another two. That's sort of how I think here. Speaker 5: Maybe all we say is 2x whiter. And it's already incredible, right? Speaker 1: Big numbers usually sell, right? So maybe eight is a bit too much, obviously. But that's a really good feedback, Sim. You see, we started with this. I've put up to eight shades whiter. But what you have just done is then added another element to it. And then ask yourself, OK, what is it really important? There's too much going on. And we came with a solution which basically married the two of them. So I think this is a really good idea. Speaker 3: What about AXY2SMILE in only 14 days? Then you've got the numbers in there as well. Speaker 5: I like Daze. Daze sounds shorter. Speaker 4: Danny, can you ask ChatGPT to make the prompts, the outputs shorter? Quick one, Dorian. We went away from black quite early on and now black's back at the bottom of this. Do we pull that colour at the bottom? Is there a reason we've gone for black here? Speaker 1: Which image has won? It was this one, right, Andri? The one, this one had the most amount of clicks, right? Speaker 5: Yeah, that's right. The ultimate winner. Speaker 1: Yeah, that was the one. Listen, Sim, in a moment, we can change that. That's no problem. So right now, what we can do, Slightly different color, right? So we can try in blues, right? So again, I like the black. Speaker 4: But yeah, I like black. Speaker 3: So yeah, it depends on the audience, right? We stuck with gold because it's always hard to match gold. We turn the text to white, then it's going to be easier to play with background colors and see what pops out. Speaker 1: White, right? I like to usually choose something You know, that's already in the background, so we can try choosing something. Let me just do another one. We can try... Speaker 4: That's why they're doing that, Andri. The teeth whitening pens, it's very small. Is that just for Amazon's AI that we've included that or...? Speaker 5: I don't even see it. Where is it now? Oh, at the bottom. Speaker 4: Yeah, right at the bottom, yeah. Speaker 5: Yeah, from my experience, Amazon's AI identifies bigger words easier, higher confidence. But as long as we have it there, I think we're good. Speaker 1: Some ideas here. Again, very quickly, I just came up with three new concepts, right? And I can see them side by side. Which one's your favourite, guys? Speaker 4: I'll go with frame 63, personally. Unknown Speaker: Or 62, I don't know. Speaker 1: Okay, we have some tags here. So now we can do if we picked one, let's say we go with number 63, right? So now we can try some of these that GBT make both fast. I think that's a really, really strong, really strong keyword, right? Fast results. And Just, let's say, do we prefer two weeks or 14 days? Speaker 3: Andri made a good point, because 14 days sounds less, doesn't it? Speaker 1: Just two weeks. Okay. Because, see, side by side, I think also that's easier to read, right? 14 days. Fast results in just 14 days. Speaker 4: I'm starting to really like that dark blue now and the white text. It's a whitening product, so we should be using white, I guess. Speaker 5: Well, depends on your competition around you, right? It's all about the context. To me, fast results actually sound a little salesy. Speaker 1: Yeah. Speaker 5: You don't really promise anything specific, but you promise a dream, you know? Speaker 1: Okay. So we have For example, let's look at these two, right? Which of these, in your opinion, is reading better? Speaker 4: I prefer 64 myself, but again, I've got numbers brain kicking in. Speaker 3: And can we also bring up UNML Safe Formula just slightly the font size or bold it? Because it's getting a little bit lost under there as you zoom out when you did the squint test. Speaker 1: Yeah, no problem. Speaker 4: Oh, that's nice. Yeah, that's good. So how important do we think the brand name is here? Obviously, it's the first part of the title. So people are going to be aware, but it's very small now. I know on the original, we had some stale to it, and perhaps that's how it got this far. Speaker 5: Yeah, I think it generally becomes more important over time when you dominate the space. Speaker 4: But that's where you guys are at now, isn't it, with this one? Speaker 5: Getting there, yeah. So I would say mid-size to big size. Speaker 4: You've got that space at the top above the circle. Perhaps we can move it there. It feels like this brand is big enough to be important, and we need to make it quite obvious. Speaker 1: Okay, we can try that. Speaker 4: All the text is getting bigger now. It's starting to move around, so we need to be a bit careful. Speaker 3: Right, take it right down, Dorian, for the SWIM test. Shrink it right back. Let's have a look what it looks like. Speaker 1: So the way we do it right now, I'll show you. So let's say we pick this one, okay? So I'm going to quickly just export it because this is just one part of what we're doing, right? Because we have all the other elements also. But right now we were just working on packaging, right? So I'm just going to drop it here. And now what we're going to do is test the new one. Just a very quick, dirty test. That's what we want to do, right? I've dropped it into my mockup. It's not perfect, but we're not looking for perfect at this stage. I'm going to save it and we're going to do this Quint test and we're going to compare it. So this is what we had previously. And this is what we just came up with. That's another hypothesis that we can test on product opinion. OK, now what we can do is a quick screen test. So I'm going to just drop your image in here and we're going to zoom out. You guys are going to squint and you tell me, does that create any additional impact? Speaker 4: I don't know about additional, but our mouth is certainly the most prominent now, I'd say. Speaker 3: I think the lack now is the sharpness of the colours is probably a bit downtone in terms of when you look at Colgate and Some of the reds in there, I'm not saying that you've got control over that element. Speaker 5: Yeah, I was going to say that too. Speaker 3: You can still read a lot of the Colgate stuff on the packaging versus the blue is a bit tougher to read. Speaker 5: Another idea there, something we may try on a separate test, that circle that says award-winning could be in a completely different radical colour as if it's not even supposed to be on that package. Speaker 1: It could be something like this, right? Speaker 3: It goes with the blues. It's not awkward in terms of colour. It stands out because we've got a lot of pastel-ly kind of downturn colours. Speaker 1: Yeah, that's another hypothesis. Last one very quickly. Speaker 4: Is that the same red that's around the sash of the pens in the top right? Not that it matters for this, but I'm just saying like... Speaker 5: It should. Speaker 1: Okay. So now we have decided to try a red circle. I'm going to drop it here just to compare it to our original one. And now we're going to do another squint test. I'm just going to copy this. So we can compare side by side. So always leave track of what you're doing. Always leave track because you can quickly compare and see where you have made a progression or where you actually messed up something that you don't like or something that you could improve. So another quick screen test. I think that Catches a bit more attention right now, don't you think? Speaker 3: Yeah. Yeah. Speaker 5: Yeah, quite busy, but definitely stands out. Speaker 3: Yeah, it does look busy with the pens because they're translucent on one side. I mean, if you look at Colgate, there's nothing you can do about it, but there's just showing the packaging and their Colgate. Off the page, for me, they stand out the most. Speaker 1: You asked me this early on, Danny. How do I decide whether I want to test something? That's the moment where I have developed some sort of a confidence where I believe this is a good idea to test. So now it's the moment. Well, we test this. Also, because we've made quite a lot of changes, right? We added the new keywords. We made this red. Here's another new keyword award-winning. So there's enough changes here at this stage where we want to test and get some new feedback. So that's the moment. Speaker 3: Yeah, indeed. Sim, best way that people can reach you? Speaker 4: Yeah, LinkedIn is probably the best way at the moment. Speaker 3: And gentlemen, start me, Dorian. Speaker 1: Yeah, so you can find me on LinkedIn, Dorian Gorski. Or you can, if you want to look up our studio, it's www.unit6.studio. Speaker 3: Excellent. Andrey? Speaker 5: Yeah, same for me. Pretty active on LinkedIn. And you can also email me, andri, A-N-D-R-I, at productpinion.com. Speaker 3: I'll be back here again next week. Take care of yourself and your family. There's some big Seller Sessions Live news dropping next week also, so we'll fill you in there. All right. Until then, take care.