Sophie Society Co.

Most Amazon sellers think they’re targeting a specific audience when actually Amazon is still showing their ads way more broadly than they realize. That’s why a lot of audience targeting feels weak. To get round this we built separate auto campaigns just for AMC audiences, kept the regular bids low and then cranked up audience bid adjustment so Amazon would only push harder once the shopper matched the audience we actually cared about. That changed performance really fast. Clickthrough rate came in 56% higher than the account average Conversion rate was 162% higher CPA dropped 46% ACoS came in 56% lower Then we pulled the search term reports from those campaigns and found converting ASINs the account hadn’t ever targeted before. We moved those into dedicated product targeting campaigns then layered the same audience on top and kept building up from there. Those followup campaigns got 23%, 28% and 30% ACoS vs the 35% account average. AMC works a lot better when you build the whole campaign around the audience from the start instead of just slapping the audience on top and hoping for the best.

Amazon just shipped something useful. I had to sit down. The Subscribe & Save dashboard finally got the glow-up we've been waiting for. Actual charts. Actual filters. Real, readable data. For anyone who's spent the last few years exporting CSVs and squinting at them like ancient scrolls, you can come out now. It's safe. This is a genuinely big deal for brands. Subscribe & Save drives serious recurring revenue, and now we can actually see what's happening with it. What are your thoughts?

Its now possible for sellers to easily do the kind of Amazon product research that used to take all day to finish or cost four figures to outsource. Its all possible now because of AI of course. You cant do this product research with basic chatbots though. This is one place where AI Agents are needed. AI Agents don't just answer one question. They can do way bigger jobs and come back with genuinely useful info like: where the market has gaps what product ideas make sense which features people actually want what pricing and margins could look like what customers keep complaining about which suppliers might be worth talking to That’s a massive change. Now instead of guessing what product to launch, you can spot where people want something and whether the current options for that thing kinda suck. Then you can find a product based on those insights. That’s where AI Agents start getting really powerful.

Amazon already tells you what hours your ads work best at… but the data on this is buried in a report so ugly that most sellers never actually do anything useful with it. That’s where day parting comes in. We actually have an SOP on this which you can get below. Anyway, doing day parting right is pretty simple. You just pull the Campaign by Hour reports, combine the last 28 days, run them all through AI, then turn the whole thing into a single clean dashboard Now you can actually see which hours are bringing in better conversions for you and which hours are just burning spend. Plus you can see which of your product groups perform best at different times of the day. This is important because a lot of ad changes sound smart in theory but you don’t really know if they’ve helped unless you can compare the before and after data. This SOP makes it way easier. Comment "Day Parting" and i will send you over the SOP... make sure to like this post as well.

Here’s part 3/4 of my story on how Sophie Society got created… My 20s were basically one long string of experiments. Some of these were smart and some were pretty ridiculous in retrospect. I spent a big part of my 20s building stuff like apps, products, ecom brands, random inventions, anything. Some worked. Most didn’t. So it goes. One of my bigger wins was a posture improvement ecom brand that went on to sell millions of units across 7 countries. That taught me something I still believe in hard - business isn’t about having one genius idea. Its about testing fast, learning fast, adjusting and not quitting too early. Eventually I started Judolaunch to help Amazon sellers rank products. We built it remote from day one, then grew the team internationally, got into 500 Startups, raised the money and then turned it into a real company. From the outside it probably looked like things were rolling. And they were at first…and then they very much were not. But that’s all part of the story, and Part 4/4 is up next where everything blows up and Sophie Society gets built the way it did.

This cinnamon brand went from making basically nothing to pulling in $11K profit in 2 months. Fun fact - we took them there WITHOUT throwing even more money into ads. We just fixed the stuff that obviously wasn’t working for them. When we partnered with the brand we saw their products were really solid. The problem was their page and the ad setup. They didn’t have enough people clicking or buying and the account wasn’t making real money. So we started off by making the listing itself better. We upgraded the main images, cleaned up other images, made the ingredients and product clearer, added in brand story and started putting better A+ content in place. Then we cleaned up the ads by cutting losing targets, moving budget into the campaigns that were getting results and pushing their products higher up in Amazon search. Here’s what happened from Dec to March: -Sales went from $33.7K to $98.7K -TACoS dropped from 35.23% to 24.97% -Clickthrough rate went from 0.88% to 1.66% -Conversion rate went from 19.22% to 25.69% They also broke into the top 7 in cinnamon and overtook brands like Whole Foods and McCormick. Sometimes the answer is to just fix the listing and the ad strategy so the whole thing finally works. More ad spend isn’t always the right solution. If this sounded good to you then check this - we’re giving sellers the frameworks, Claude skills and PPC systems that make this work inside The AI Amazon PPC Challenge we're running soon. ↳ join or get left behind - https://bit.ly/47v5xY7

A good Amazon video ad can pull in way more clicks and sales. But a lot of sellers mess up the results by screwing up the basics. Same video for every search. Run it on terms they haven’t proven yet. Waste the first few seconds. Test too many variables at the same time. Then they wonder why it flops. The better thing to do is make the video match what the shopper is actually looking for. Simple. Use your regular ad data first to figure out what search terms already work. Then get to the point fast. Also test one thing at a time so you know what actually made the difference. If you want the full SBV breakdown, comment "SBV" and i will send it over.

Amazon just low-key gave a bunch of sellers a new way to lose profit with a new 3.5% fuel and logistics fee for some third-party sellers using FBA. 3.5% is a lot bigger than it sounds when your margins are already tight. At first glance that might not sound like a huge deal. But if you sell heavier products or lower-priced products or anything where the margin is already pretty slim then that extra 3.5% can hit fast. Amazon says the fee is temporary and its only happening because of the higher fuel and shipping costs we’re seeing now. But whether it lasts a month or carries on longer, sellers still need to deal with it right now. The best thing to do now is go back through your products and check where this will hit hardest. Look at your pricing, pack sizes and which items still make sense to push with this fee in place. The reality is a lot of brands are only gonna notice how bad this is once their profit starts slipping.

Most Amazon sellers are doing keyword research the hard way, and they don't even realize it. They pull their Helium 10 data, spend a couple of hours manually filtering through thousands of keywords, copy-paste everything into a spreadsheet, and still walk away unsure what to actually prioritize. It's one of the biggest time sinks in any product launch, and the output is rarely clean enough to use directly. We fixed that by combining Helium 10 with Claude AI, and the difference in output quality is hard to overstate. Instead of a messy flat list of hundreds of terms, we get a fully segmented keyword spreadsheet organized by shopper intent, clean, prioritized, and ready to plug straight into listing copy, PPC campaigns, and content strategy. No extra cleanup, no second-guessing what to do with it. What used to take a few hours now takes under one. And the output is honestly better than anything we were producing manually. If you want the exact SOP we use to make this happen, comment "KEYWORD" below and I'll send it over. If this sounded good to you then check this - we’re giving sellers the frameworks, Claude skills and PPC systems that make this work inside The AI Amazon PPC Challenge we're running soon. ↳ join or get left behind - https://bit.ly/47v5xY7