There's a number buried in Amazon's ad documentation that changes how you should think about your entire campaign structure. It's 30. That's how many conversions per month a campaign needs before Amazon's algorithm can optimize your bids. In highly segmented accounts, most campaigns never get there. The structure you built for control is quietly working against you.

In this episode, I break down why the exact-match-only playbook from 2019 is now causing data starvation across most accounts. How Amazon's Cosmo engine changed the way the algorithm evaluates and optimizes campaigns. And what your account actually looks like to Amazon when most of your campaigns are sitting below that 30-conversion threshold with nowhere near enough signal to work with.

I cover why the instinct to fix it by pausing or toggling campaigns actually makes things worse, the 30-day audit you can run right now to see how much of your budget is stuck in the problem zone, and the 2026 campaign architecture built around three buckets that gives the algorithm enough data to actually do its job.

I also get into why budget allocation needs to completely flip from where most sellers have it today, and how negative keywords let you stay in control while running wider targeting.

If your account was built on the old segmented playbook and performance has been declining, this is probably why.

In April 2024, Amazon introduced a fee for carrying too little inventory. They already charged you for carrying too much. The window in between, 28 to 90 days of supply per FNSKU, is the only place you don't get penalized. And as of January 2026, that fee is now assessed per individual variation, not per parent ASIN. That rule change exposed a huge number of SKUs that were previously safe.

In this episode, I break down the full mechanics of how the low inventory fee triggers, the 2026 fee rates ($0.47, $0.87, and $1.11 per unit depending on how far below threshold you fall), and a real example where a single SKU selling 50 units a day generates $1,300 a month in fees. I also show you how to check if you're already being charged right now.

I cover how this fee interacts with DD+7 receiving timelines to create a cash flow trap that gets worse the faster you grow. Two Amazon policies pulling in opposite directions, one penalizing you for having too little stock and another limiting how quickly you can restock. Plus the May 2025 restock limit changes that made it even tighter.

I walk through three practical fixes: decoupling purchase orders from DD+7 timing, using Amazon Warehousing and Distribution, and domestic 3PL buffering. Plus the hidden lever most sellers don't know about, because a partial restock is enough to stop the fee from triggering.

You don't solve this with more cash. You solve it with better timing.

Amazon's ad platform has fundamentally changed. The old playbook of raising bids, increasing budgets, and winning more impressions is now actively working against you. CPCs are up 48% cumulatively since 2019. The algorithm no longer simply rewards the highest bidder. And more budget doesn't scale what's working. It gives Amazon permission to find traffic that doesn't convert.

In this episode, I break down what the algorithm actually rewards in 2026 and why a 15% converter beats a 2x bid every time. I walk through the 3-week budget trap where week one looks promising, week two softens, and week three falls apart as Amazon floods you with low-intent traffic. I cover the reset mistake that sends your campaign back to Day 1, and why ACoS is the wrong metric to watch when TACoS tells you what's actually happening with your business.

I also lay out the full diagnostic sequence you should follow in order before touching your budget. Start with creative and listing problems. Then bids. Then TACoS. Only then do you consider raising budget. Most sellers do this backwards and wonder why scaling makes things worse.

If you're spending more on Amazon ads and getting worse results, it's not a budget problem. It's a conversion problem. Stop outspending. Start out-converting.

#AmazonPPC #AmazonAds #AmazonAdvertising #AmazonFBA #AmazonSeller #PPCStrategy #ACoS #TACoS #AmazonAlgorithm #PPCOptimization #AmazonCPC #EcommerceStrategy #AmazonFBATips #AdSpend #AmazonBids

Amazon Marketing Cloud has always held the data that explains why your ads are actually working or not. Attribution across DSP and Sponsored Ads. New-to-brand rates. Frequency curves. Full path to purchase. The problem was always access. You needed SQL expertise and hours of query development to get any of it. That barrier is gone.

In this episode, I cover what AMC via Amazon's MCP server actually unlocks for your advertising decisions now that it's in open beta. You can ask plain English questions inside Claude and get answers back directly from your AMC instance. Which campaigns drive first-touch engagement for new-to-brand customers. How many touchpoints happen before conversion. Which DSP campaigns are influencing purchases that Seller Central credits to something else entirely.

I also walk through audience building without a data team. Cart abandoners who added in the last 14 days versus three months ago are completely different targeting propositions, and AMC lets you slice that properly through natural language instead of SQL. Plus the strategic questions most brands have never thought to ask: the overlap between your paid audiences and organic shoppers, the correlation between ad spend and BSR movement, and where you're double-serving the same customer across channels.

I'm also direct about what this doesn't do. It removes the execution barrier. It doesn't replace the judgment that determines what to execute. The insight gap is still on you.

The data has been sitting there for years. Now you can actually use it.

#AmazonAds #AMC #AmazonMarketingCloud #AmazonDSP #AmazonPPC #AmazonAdvertising #AmazonMCP #AmazonFBA #AmazonSeller #PPCStrategy #DSPAdvertising #AmazonAttribution #EcommerceStrategy #AmazonData #DigitalAdvertising

Over three weeks, 11 different AI systems crawled our website 10,932 times. Google Analytics reported essentially none of it. This isn't a GA4 bug. It's architecture. AI bots don't execute JavaScript. They never trigger the tracking script. Every crawl from ChatGPT, ClaudeBot, Meta AI, Amazon Q is completely invisible to standard analytics.

In this episode, I walk through our actual BotSight report from February 24 to March 16, 2026. Real data from our own site. You'll see which AI systems are crawling, how often they return, which pages they're targeting, and what that tells you about where AI search is heading.

The numbers were surprising. 47 AI visits for every human click. ClaudeBot up 107%. Meta AI now the second most active crawler. SearchGPT up 833% in a single week. And the conversion data changes everything: AI referral traffic converts at 4 to 15 times the rate of organic search because those visitors arrive already convinced. They've been prequalified by the AI answer before they ever hit your site.

I also cover the difference between being crawled and being cited, why recall frequency matters more than raw hit count, and what your top crawled pages are really telling you about your AI content strategy.

This is server-side tracking, not GA4. A fundamentally different layer of data. Once you understand what you've been missing, you can't unsee it.

#AISearch #AIVisibility #GoogleAnalytics #GA4 #AIBots #ChatGPT #ClaudeBot #MetaAI #SearchGPT #ServerSideTracking #EcommerceMarketing #DigitalMarketing #AITraffic #BotSight #SEO

Google's Universal Commerce Protocol is live. But there's one product-level attribute that decides whether your products get a Buy button in AI Mode, and most brands have never heard of it.

In this episode, I break down exactly what the native_commerce attribute is, how the AI mode checkout flow works, and what happens when the attribute is missing from your product data. I walk through the fix using a supplemental feed, why you must not edit your primary feed to do this, and how to start with your top 20% of products to test safely.

I also cover the hidden dependency beyond Google that most sellers miss entirely. Your payment provider might be quietly blocking you from UCP checkout through something called the Agent Payment Protocol. There are three layers in the chain that all need to be connected before AI can complete a purchase for your product.

This matters now because 83% of ChatGPT shopping results come from Google's product data. OpenAI's native checkout failed after five months. The commerce layer that survived runs through Google. If you're selling on Google and you haven't set this up yet, your competitors who have are getting Buy buttons in AI answers and you're not.

Three action items by the end of the week. This episode gives you all three.

#GoogleAds #UCP #AICommerce #GoogleShopping #MerchantCenter #AISearch #EcommerceStrategy #GoogleAI #AIMode #DigitalMarketing #Ecommerce #ProductData #ShoppingAds #ClearAds #AICheckout

OpenAI launched Instant Checkout in September 2025. By early March 2026, it was gone. Five months and roughly a dozen live merchants. But the real story isn't that native checkout failed. It's what the 83% number tells you about who actually controls the AI commerce layer and what that means for where your products show up.

In this episode, I break down what Instant Checkout actually was, the operational layer OpenAI never built, why consumer demand wasn't there, and why they couldn't just fix it with $600B in commitments and an ad business to build. I also cover what the coverage got wrong, because the broader AI commerce protocol didn't die. The integrations that survived tell you exactly where this is heading.

I get into Google's moat of 35 billion products and 20 years of merchant data, the Amazon paradox where $50B in infrastructure investment still meets a blocked crawl, and the single bottleneck that determines whether AI can sell your products: product data quality.

The traffic urgency is real. Position 1 on Google is losing 58% of its clicks to AI answers. If you're not visible in the AI commerce layer, you're losing sales you'll never see in your analytics. This episode shows you where the opportunity actually sits and the new tool that finally lets you see whether AI is using your content at all.

#OpenAI #ChatGPT #AICommerce #GoogleShopping #AmazonAds #Ecommerce #AISearch #MerchantCenter #GenerativeAI #AIOverviews #UCP #EcommerceStrategy #DigitalMarketing #ClearAds #FutureOfCommerce

Microsoft launched AI Performance inside Bing Webmaster Tools on February 9th. It's the first tool from any major search engine that shows you exactly how often AI is citing your content. Google hasn't built this. Nobody else has.

In this episode, I walk through a real case study. A beauty brand published one long-form article on a skincare ingredient. Less than three weeks later, we pulled up the AI Performance dashboard live on a client call. That single page had 3,200 AI citations. The next highest page on the site? 334. Nobody knew until someone looked.

I cover what the dashboard actually measures and what it doesn't, the five metrics it shows you, and why grounding queries are the most important section most people scroll straight past. Grounding queries aren't human prompts. They're the sub-queries AI fires before building its answer. That distinction changes how you read every piece of data in this tool.

I also break down why 3,200 citations doesn't mean 3,200 visits (there's a visibility gap between machine retrieval and human clicks), which AI platforms the tool covers and which it misses, and the exact strategy we're running off the back of this data, including why the brand owner's first reaction was to add product links to the cited page before creating more content.

The tool is free. It takes two minutes to set up. Go to bing.com/webmasters/aiperformance and check your numbers.

#AIVisibility #BingWebmasterTools #AISearch #GEO #GenerativeEngineOptimization #ChatGPT #Copilot #AIcitations #EcommerceMarketing #DigitalMarketing #SEO #ContentStrategy #BrandVisibility #AISearchOptimization #EcommerceSEO