Your discount has a past

FROM OUR SPONSOR:

Your AI is only as good as the Amazon data you give it

You can ask ChatGPT, Claude or Gemini to analyze your Amazon business in seconds.

But there’s a problem:

If your AI is working from incomplete, stale or poorly structured data it can give you a confident answer that’s completely wrong.

Amazon doesn’t hand you a clean “profit” number.

Your sales, fees, refunds, reimbursements, PPC spend, inventory, search terms and costs are scattered across different reports and APIs. So when an AI assistant has to pull raw Amazon data join it, reconcile it and calculate the answer itself… you’re burning tokens, adding extra steps and increasing the chance of bad decisions.

And when you’re making decisions on PPC, inventory, pricing or profitability yesterday’s numbers aren’t always good enough.

The better strategy:

Don’t make your AI figure out your Amazon data.

Give it a decision-ready data layer.

Give Nova MCP the question and get the Amazon numbers behind the answer.

It connects your AI assistant to 500+ Amazon data points across sales, profitability, PPC, search terms, inventory, fees, refunds, reimbursements, listing health and more - already structured and reconciled before your assistant asks the question.

Instead of digging through Seller Central you can ask things like:

“Which SKUs are losing margin after advertising and fees?”

“Where am I about to run out of stock?”

“Which search terms are spending money without converting?”

And get answers using your actual Amazon data.

The asset: Nova has opened its complete library of 500+ Amazon metrics so you can see exactly what your AI assistant can access before connecting anything.

→ Browse the 500+ Amazon metrics available through Nova MCP

500+ data points. 23 Amazon marketplaces. Multiple refreshes per day.

Give your AI every Amazon number it needs.

Your Fake Amazon Discount Just Developed a Paper Trail

Amazon’s latest pricing rules are turning every discount into a crime scene: witnesses, evidence, and one guy sweating aggressively in the corner.

For years sellers could perform one of e-commerce’s most beloved magic tricks. A product costs $39. Then, shortly before a big sale, it mysteriously becomes $59. Then the sale arrives and - SWEET MERCIFUL CAPITALISM, IT’S 30% OFF! Now only $41. Which is a discount in the same way that stealing someone’s wallet and returning their library card is technically “giving back to the community”.

The entire trick depends on the customer having no memory whatsoever. And fortunately nobody remembers what a garlic press cost in March. Most people barely remember March. March is just February with better lighting. But Amazon has expanded the price history it can show shoppers from roughly 90 days to a full year in supported markets. A YEAR. That’s not price history. That’s a Netflix true-crime documentary about your air fryer.

Suddenly January can testify. April has screenshots. June has hired an attorney. And that random Tuesday when Brandon from Growth dropped it to $34.99 after watching half a YouTube video about dynamic pricing? Brandon is now Exhibit C. That's why you need to get our Amazon Price Crime Scene Kit.

So now you try the classic move: $39 becomes $59, then “30% OFF” becomes $41. Except the shopper can look backward and say: “Hang on. This was $36.99 four months ago.” And suddenly your beautiful promotional strategy is a raccoon in a necktie standing before Congress saying: “Senator, I reject the premise that I was inside that dumpster”.

That’s the real change here. Amazon isn’t just judging reference prices behind the scenes anymore. It is increasingly giving shoppers enough evidence to conduct their own tiny financial investigations. Your customer is no longer merely shopping for a blender. They are now Detective Blender, Special Victims Unit: MSRP.

And this becomes especially ridiculous during major sale events, when Amazon’s homepage looks like a casino suffered a traumatic brain injury - DEAL! LIGHTNING DEAL! PRIME EXCLUSIVE! 43% OFF! ONLY 02:17:46 REMAINING! Sir, I came here for dishwasher tablets. Why does this feel like I’m cutting the blue wire?

The strategic implication is simple: your price now has lore. It has a childhood. It has ex-girlfriends. It has a regrettable semester abroad. Every coupon, promotion, desperate inventory clearance and late-night price change becomes part of a story that may eventually have to make sense.

So build a 365-day pricing map for every important SKU. Track the normal selling price, meaningful lows, promotional windows, coupons and major changes. Not because spreadsheets are exciting. Excel is what happens when a prison accountant discovers Minecraft. Build it because before approving another giant red “40% OFF” badge someone should ask: “If the customer saw the entire year would this still look like a deal?”

If yes, great. If no - congratulations: you’ve discovered a revolutionary new pricing strategy called “not lying with enormous red numbers.”

Use that map before Prime events, Black Friday, coupons and reference-price changes. Because the old promotional strategy was basically: raise the price, lower the price, add a red badge and hope the customer has the historical memory of a concussed hamster.

That hamster may now have twelve months of data.

Your product has a past. Amazon may show it to people. And the next time Brandon suggests raising a $39 product to $59 so you can heroically discount it to $41, remember: the customer may have receipts, Amazon definitely has receipts and Brandon should not be allowed near the bloody pricing dashboard.

I Built a Team of AI Employees to Run My Ecommerce Business

AI has officially graduated from “write me a product description” to “apparently, I work here now.”

Ecommerce businesses are building entire AI teams - analysts, SEO researchers, marketing assistants, strategists and even chiefs of staff. Which feels slightly excessive for technology that, not long ago, struggled to draw a human hand with the correct number of fingers.

These AI employees can analyse Google Analytics, monitor ads, research competitors, find SEO opportunities and automate the repetitive tasks slowly draining your will to live.

Some can even work together and report back to each other.

Fantastic. We’ve automated middle management.

But humans aren’t being deleted just yet. Judgment, negotiation, relationships and physical work still require actual people - because apparently telling an AI to “have a difficult conversation with our supplier” does not end well.

The smart approach? Start small. Give AI one boring, low-risk job, see if it behaves itself, then gradually hand over more.

Find out what an AI-powered ecommerce team actually looks like - before your spreadsheet asks for a performance review

Headline Sold Separately

eBay has quietly made a decision that is much more revealing than “we added livestreaming”. Starting October 15 it is basically telling its accounting system: whatever happens in the room where a grown man is screaming about a holographic Charizard under a ring light quarantine it from normal commerce.

eBay Live transactions are being removed from ordinary seller standards, service metrics and Top Rated Plus eligibility. Which is fascinating because that means eBay has looked at livestream shopping and concluded: “Oh. You’re not a feature. You’re a separate animal and apparently you need your own enclosure.”

And fair enough. Normal eBay is some guy in Wisconsin listing a carburetor at 11:43 p.m. and then going to bed. Live eBay is an auctioneer, game-show host, customer-service rep, nightclub MC, inventory manager and hostage negotiator compressed into one person who must keep shouting “LET’S GOOOO” because nobody has typed in the chat for eleven seconds.

The rules are already completely different. Live sales have their own fees. Auctions can last under two minutes. Winners can be charged immediately. Bids can’t be retracted. Sellers get livestream consoles, chat moderation, event analytics, video receipts, flash deals and a special shopping window before handling time even starts. At this point calling this “e-commerce with video” is like calling a casino “a bank where everyone is mysteriously sweating.”

And then you get to the rules for the actual broadcast. eBay has to care about what appears on camera, whether viewers are fake, whether sellers are running games of chance and whether a pack break remains visible on-screen. The marketplace is no longer just asking: “Did Todd mail the baseball card on Tuesday?” It now has to ask: “Did Todd manipulate his audience, fake engagement, accidentally operate a lottery or move the Pokémon cards outside the designated rectangle?”

Todd has become a regulatory sector. And that is the interesting part: live shopping is starting to require its own constitution.

Traditional e-commerce is built around a catalog. The product sits there silently. You search for it. Compare prices. Close the tab. Forget about it. Then three days later an ad follows you into a weather app like a private investigator hired by a refurbished bread maker.

Live commerce is almost the opposite. It is synchronous. The seller is part of the product. The audience is part of the product. The countdown is part of the product. And the suspicion that seventeen strangers with usernames like CardGoblin92 are about to steal your item is absolutely part of the product. Once those mechanics change the old seller scorecard starts looking absurd. A livestream seller can generate a tidal wave of transactions during one scheduled event, combine shipments afterward and operate under rules specifically designed around what happened on camera. Measuring that business exactly like someone selling three vintage lamps a week is like using the same performance review for a Formula One driver and a man backing a Honda Civic into a Costco parking space.

That is why eBay’s move matters more than another headline saying live commerce is “growing.” Mold grows. Toenails grow. The mysterious bag of spinach at the back of your refrigerator is probably establishing municipal government.

Growth does not make something a new industry.

Infrastructure does.

Once a marketplace gives a format its own fees, auction mechanics, transaction rules, moderation systems, restricted activities, seller access, analytics and now its own performance calculations that format is no longer sitting inside the old business model. It has become a jurisdiction.

eBay Live is basically a tiny country inside eBay now. It has its own tax code, customs desk, broadcasting authority, gambling-adjacent panic room and immigration policy for sellers. Its national anthem is one man screaming “SOLD!” while another opens foil packaging with the solemn concentration of a surgeon separating conjoined twins.

And that may be where live shopping is actually heading: not “shopping, but on video” but a different commercial medium entirely. One where the merchant is also the entertainment, the audience helps manufacture urgency, checkout becomes part of the show and marketplaces eventually have to admit that regulating all of this like an ordinary product listing makes roughly as much sense as regulating a circus with the employee handbook from Office Depot.

Organic, Until Further Notice

For years e-commerce teams were given a fairly simple proposition by Google. Build a good product feed. Keep your Merchant Center account healthy. Write accurate titles. Upload strong images. Maintain your prices, availability, GTINs and other product data. Follow Google's requirements and, if everything lined up, your products could earn valuable visibility without paying for every click. For merchants that visibility became part of the machinery of organic growth.

Then the shelf moved.

See how much of your “organic” traffic you actually own

How to Get More Views on TikTok: Here’s What Works in 2026

Have you ever uploaded a TikTok video that you thought was very fantastic, only to have it get 63 views before vanishing into thin air? You are definitely not alone. 

Amazon’s fight with Meta: Who owns the customer relationship when an agent does the buying?

Amazon blocked Meta’s Muse AI agent from its store Sunday night, and leaders across the tech industry have spent the past two days debating the bigger questions raised by the move.

Google opens native AI checkout to eligible Shopify merchants

Temu ranks first in four European ecommerce markets

Temu has broken into the top 10 online retailers in almost every European country. In four countries, the platform owned by China’s PDD Holdings even ranks first, according to ECDB, based on its 2025 data.