Tariffs are shifting. De minimis rules are under pressure. And new enforcement could change how retailers operate overnight. If you sell across borders—through eCommerce, marketplaces, or direct—these changes don’t just impact compliance. They impact your pricing, margins, and conversion rates.

Register for our Trade and Tariff webinar series for regular updates.

Target's marketplace just picked up Forever 21, Clarks, and JanSport. The brand names will get the headlines. The number that matters is buried further down. More than half of Target's Korean beauty assortment now comes from brands Target does not own.

Target Plus is invite-only. It launched in 2019 as the anti-Amazon. Every seller is hand-picked, which keeps out the gray-market resellers who flood the open platforms. Six years later that discipline has become the growth engine. Sarah Travis, Target's chief digital and revenue officer, calls the approach curation at scale. Those two words pull against each other, and Target is wagering it can hold both at once.

The K-beauty figure is why they need the channel. The category caught fire on social, Target watched demand build, and it answered by adding hundreds of items through Target Plus instead of its own shelves. Read that plainly. Target's owned buying cycle is too slow to catch a trend while the trend is still warm. The marketplace is the fast lane. The house brands are stuck in traffic.

A second move sits inside the same announcement. Target is shifting bigger boxes onto Target Plus. Laptops, televisions, printers. That pushes inventory risk off Target's balance sheet and frees floor space for categories with better store economics. It is a margin decision dressed as guest discovery. Shoppers get more choice online. Target gets less capital locked in slow-turning electronics. Both are true.

The part worth watching is at the edge of the page. Target is feeding its marketplace items into AI-powered search through deals with OpenAI, Google, and Microsoft. It is also using its own agents to screen who gets an invite. Chief Product Officer Prat Vemana told Retail Dive last fall that agents now review every applicant and hand the marketplace team a tidy summary. So machines screen who gets in, and the same AI layer increasingly decides who shoppers see. Human curation was the entire pitch in 2019. It is turning into a supervisory job.

Target has not solved the contradiction underneath all of this. A curated marketplace stays small on purpose. Amazon runs millions of sellers. Walmart runs hundreds of thousands. Target Plus, by design, is a rounding error next to either. Curation is the moat and the ceiling in the same motion. You cannot out-assort Amazon by hand-picking brands. You can only out-taste it, and taste has a hard cap on revenue.

Whether that taste converts into a number big enough to matter is the open question. Target's core business has been soft for a while now. A tightly held marketplace is a real lever with a short throw. It protects the brand better than an open floodgate would. It is far less obvious that it moves the top line.

The beauty stat is the one to sit with. When your fastest-growing category is mostly other companies' brands, the question that lingers is what your own shelves are still there to do.

Target Plus stopped being a side experiment. It is where Target's assortment growth is actually happening now, and increasingly where an algorithm decides who gets in and who gets seen. For brands, that makes it the rare invite worth chasing. No monthly seller fee, SKU exclusivity, and GMV that grew nearly 60% last quarter, all while Amazon and Walmart get more crowded by the month.

For Target, the harder read sits underneath. A hand-picked marketplace protects the brand but caps the upside, and a business this soft needs upside. The channel is winning. Target still has to prove the channel is big enough to move the company.

Klaviyo put out a release on June 30 saying Composer, its marketing agent, is now in public beta, plus a refresh of Customer Agent. The idea is that both agents run off the same customer profile and hand work back and forth. Customer Agent finishes a support conversation, drops what it learned onto the record, and Composer turns around and builds a campaign off it. They call it a flywheel.

There's a real argument buried in the marketing speak. Klaviyo is betting that keeping marketing and service on one data layer beats stapling AI onto a stack where those two things never talk. That bet holds up. It's also where Klaviyo actually has an edge, with 14 years of behavioral data across roughly 196,000 brands. A generic model can't fake that.

Now look at what's actually live. Composer is beta, not generally available. "More functionality coming through 2026" is covering for a product that isn't finished. And the customer quote about 113 flows tripping over each other openly admits the mess these agents clean up is a mess Klaviyo's platform made.

The revenue story is a promise. No lift numbers, no attribution.

The single-profile setup is the right call and it's genuinely hard for rivals to copy. But beta plus a 2026 roadmap means the flywheel is still mostly a plan. Wait for real revenue-lift data before you buy the story.