🚨 9 Days Until the Tariff Swap. Importers Delaying Shipments Could Pay More, Not Less.

This week's signal: Nine days. That's what separates us from the biggest tariff transition since IEEPA was struck down. Section 122 expires July 24. Section 301 is designed to land the same day. Meanwhile, back-to-school is an $85 billion season that's already underway, IEEPA refunds are hitting importers' bank accounts, AI shopping agents are creating legal questions nobody has answers to yet, and data readiness has become the single biggest differentiator between brands that show up in agentic commerce and brands that don't. Here's your briefing.

1. Section 122 Expires in 9 Days. Here Are the Three Scenarios Importers Should Model.

Section 122's 10% surcharge expires at 12:01 a.m. EDT on July 24. The president cannot extend it unilaterally. Congress hasn't acted. Rebuttal comments on the Section 301 proposal were due yesterday (July 14). USTR aims to finalize the new tariffs before the sunset. The architecture of the replacement is visible: 10% on 14 economies (Canada, Mexico, EU, UK, Taiwan, and 9 others), 12.5% on 46 economies (China, Japan, India, Korea, Vietnam, Brazil, Australia, and 39 others). These cover 99.4% of all US imports by value.

Three scenarios for July 25: Scenario A (most likely): Section 301 lands cleanly before the sunset. Country-specific rates go into effect. USMCA exemption status under Section 301 is legally untested, so Canadian and Mexican goods face new uncertainty. Scenario B: Section 301 is delayed past July 24. Section 122 lapses. MFN rates apply alone temporarily, creating a narrow import-acceleration window before Section 301 catches up. Scenario C: Congress extends Section 122 at 10-15% while Section 301 is finalized. Rates hold steady temporarily. Gibson Dunn noted the Section 301 duties would apply "in addition to" MFN rates, existing China Section 301 tariffs, AD/CVD, and Section 232. An importer delaying shipments to clear after July 24 could pay more, not less. Model all three scenarios against your H2 pipeline this week.

2. Back-to-School Is an $85 Billion Season. It Started Last Month. July Is the Peak Window.

The numbers from three major surveys landed this month and they paint a clear picture. EMARKETER pegs US back-to-school retail at $85.42 billion in 2026. Deloitte says K-12 spending will reach $30.9 billion at $570 per child (flat YoY). JLL puts it at $489 per child, up 11.7%. The season now spans five months (June through October), with 73% of shoppers placing orders in June and July. July is the peak window. If you sell anything adjacent to back-to-school — apparel, tech accessories, organizational products, dorm gear — your campaigns should be running right now.

The value-seeking signal is the loudest in years: 75% of parents will switch brands if their preferred brand is too expensive (up from 67% in 2024). 57% expect the economy to worsen in the next six months, the highest since 2020. Walmart surged to 77.3% as the top shopping destination. 46% of families plan to jump on mid-July promotional events. Upper-middle-income families ($100K-$199K) are spending 9% less. Higher-income families ($200K+) are spending 20% less. Lower-income families are spending 10% more, driven entirely by higher prices. For DTC brands: this isn't a season where aspiration wins. Value messaging, bundle pricing, and quality-for-the-price positioning will outperform premium positioning. September still accounts for 29% of planned spending, so don't pull campaigns after August.

3. IEEPA Refunds Are Hitting Bank Accounts. The Government Is Fighting to Limit Them.

CBP certified the first wave of IEEPA tariff refunds in May and Treasury is depositing them via ACH into importer accounts. Over 56,000 importers registered through the CAPE portal for $127 billion+ in potential refunds plus 6% interest. Clean declarations are processing in 45-90 days. If you imported under IEEPA tariffs between April 2025 and February 2026 and haven't filed through the CAPE portal, you're leaving money on the table with every day you wait.

But there's a fight brewing: The Trump Administration appealed a CIT injunction that would have required the government to pay refunds on "finally liquidated" entries — older entries that would normally be beyond the refund window. If the appeal succeeds, only unliquidated and recently liquidated entries get refunded. More complex scenarios (drawback, FTZ interactions, protest-pending claims) are deferred to later CAPE phases with no timeline. And on June 3, Trump signed an Executive Order titled "Strengthening Customs Enforcement," directing CBP to impose new identification requirements on foreign entities, sharply curtail penalty mitigation, and dispose of seized non-compliant goods. The refund door is open. It may not stay this wide. File now if you haven't.

4. AI Shopping Agents Are Buying on Behalf of Consumers. Nobody Knows Who's Liable.

A question that's moved from theoretical to urgent in the last 60 days: when an AI agent buys the wrong product, orders the wrong quantity, or commits a consumer to a purchase they didn't intend, who's responsible? The brand? The AI platform? The consumer who delegated authority to the agent? Perplexity's Buy Now is processing $2 billion in annualized GMV through agent-mediated transactions. Google's Universal Cart works across Search, YouTube, Gmail, and Gemini. Amazon's Rufus influences 25-35% of searches. The legal framework for agent-initiated commerce doesn't exist yet.

What DTC operators should do now: Salsify's 2026 Consumer Research found 22% of shoppers already use AI search tools to discover products, but only 14% trust AI recommendations enough to buy immediately. That gap is where the accountability problem lives. The brands that win in agentic commerce are the ones that treat product content as a connected ecosystem: structured data, accurate attributes, clean pricing, and rich Q&A that AI agents can parse. If your product data is messy, incomplete, or inconsistent across channels, AI agents will either skip you or misrepresent you. Both outcomes cost you money. Product content accuracy is no longer a merchandising task. It's a legal and commercial risk management function.

5. Data Readiness Became the Real Competitive Moat in H1 2026. Here's What That Means for H2.

If you zoom out on the last six months, one theme connects every major story we've covered. Google shipped Universal Cart and UCP. Amazon's Rufus drives a third of searches. Perplexity bypasses checkouts. Shoptalk Europe declared "agentic commerce" the defining phrase of 2026. Shopify shipped AI merchandising and composable checkout. In every case, the brands that benefited were the ones with clean, structured, consistent product data across every channel. The brands that didn't have it were invisible to agents, misrepresented in AI recommendations, or losing sales to competitors whose data was better.

The H2 playbook is clear: Audit your product data across Merchant Center, Amazon, Shopify, and every marketplace you sell on. Make sure titles, descriptions, specifications, pricing, and imagery are accurate and consistent. Implement structured attributes that AI agents can parse (not just human-readable marketing copy). Optimize for intent, not keywords — Rufus, Gemini, and ChatGPT respond to semantic product content, not keyword density. Join Google's UCP waitlist if you haven't. And if your product data management is still a spreadsheet someone updates quarterly, that's the operational bottleneck that determines whether AI sells your products or your competitor's. Digital Commerce 360's recent webinar titled "How to Optimize Data for Agentic Commerce" confirmed what operators are discovering: data infrastructure is no longer a back-office function. It's the front line.

The Next 9 Days

Most Read Articles This Week