Were you forwarded this email? Sign up here to get the next edition.

I've been on a Claude rampage this week. I even bumped up my plan again. This thing's gonna end up costing me a fortune. But it is making me very intolerant. I’m off to London next week for work and had to pack my own bag. Claude Dresser come at me.

Enjoy
Bushy

This week’s ecommerce news you should know


Three AI design tools dropped in one week 🎨

Claude Design launched this week. It makes prototypes, slides, and one-pagers, but for the moment, it is a separate site and experience from Claude. It reads your codebase and design files at setup, builds a brand system from them, and applies your colours, fonts, and components to everything afterwards. Powered by Claude Opus 4.7, research preview for Pro plans and above. I went down the rabbit hole, redesigning the entire Add To Cart website. Got very excited. Stayed up until midnight. Ran out of credits. Currently waiting until Tuesday to start again. Kind of like working with a real designer.

Within 24 hours the response came, Canva AI 2.0 arrived with the promise of full conversational design, agentic editing, brand memory, and six new intelligent workflows. Their biggest product evolution since 2013. Then Adobe flooded the zone with a wave of brand intelligence and customer experience orchestration announcements that I don't have the buzzword bandwidth to process. Designers, if it's all too much, just go back to your sketchpad and put on a Death Cab for Cutie CD.

Ad platforms are getting very helpful all of a sudden 📊

A lot happened in platform land this week. The blue flowery petal icon you've been seeing in Business Manager is Meta AI Business Assistant, now available to Australian advertisers inside Ads Manager and Business Suite. It can optimise campaigns, analyse performance, benchmark you against industry data, and resolve account issues. The real test is whether it can reinstate accounts that have been unfairly banned. If it mentions my opportunity score, there will be harsh words.

It seems everyone just wants to help. Amazon Ads launched an AI video generator in Australia for sponsored brand campaigns. Available now. LinkedIn is trialling CrossCheck in the US. It lets you trail AI models side by side to compare outputs. Interesting service. And AusPost and eBay launched print-in-store labels. Scan a QR code at the post office to generate a shipping label without a printer. Useful for anyone dispatching without a home office setup and keen to get an update on neighbourhood gossip.

Google killed DSAs. ChatGPT wants your ad budget. 🎯

Google's Dynamic Search Ads are done. From September, all DSA campaigns automatically upgrade to AI Max. Upgrade tools are available now if you want to migrate on your own terms before the forced switch. Google claims AI Max delivers 7% more conversions at similar CPA/ROAS. Let’s see.

Meanwhile, OpenAI has moved some ChatGPT ads to a CPC model.$3 to $5 a click, minimum commitment down from $250k to $50k, and a self-serve ads manager now open to global advertisers. They're targeting $2.5 billion in ad revenue this year. The platform is moving fast from brand play to performance marketing. Will GPT be the Bing of search campaigns? Time will tell.

Your refunds might be your best retention tool 💳

Blackhawk Network launched a Stored Value Platform this week. One unified consumer balance across gift cards, refunds, promotional credits, and loyalty points. Instead of managing separate systems for each with different rules, expiry dates, and limitations, customers get a single balance that's simple to track and easy to spend. In one early implementation, the platform retained tens of millions of dollars that would otherwise have left as refunds, with 90% of it spent within 30 days.

On a similar note, Uber Eats in the US launched customer returns this week. Book a courier to collect your return immediately, with the fee based on the courier's time and distance. Send it back coooooold.

Lowe's new AI feature is mulch improved 🌿

Last week I called Starbucks idiots for their ChatGPT app. This week, Lowe's launched a Mulch Calculator powered by Mylow, its AI assistant. You type "Mulch Me Now," describe your garden, and Mylow works out exactly how many bags you need, recommends the right product, and drops it in your cart. I don't care if it puts me firmly in old dad territory. I love it. This is what the world needs. Get off my lawn.

If you work in ecommerce, you don’t have to figure it out alone. Inside the Add To Cart Community, you’ll find like-minded professionals, expert insights and live webinars. All for free.

Not in the Add To Cart community?

NEW episode dropping in your feed today
🎧 Spotify | 🎧 Apple | 📺 YouTube

How to Fix Invisible Conversion Problems | ATC Playbook #619

Your abandoned cart sequence is for the customer who made it that far. But what about the customers that never make it that far?

Most ecommerce teams spend their energy on the same part of the funnel. The checkout, the free shipping trigger, the abandoned cart flow. All aimed at the customer who is already close to buying. But the most expensive drop-off happens before any of that, and it doesn't show up in your data because the customer never got there.

"So many people would never buy a bra online. They have no idea what size they are."

Chloe de Winter, Co-Founder @ Nala

Chloe de Winter built NALA in one of the most technically difficult categories in ecommerce. Bras. A product most people have historically needed to try on, and a customer underserved by almost every other brand in the market. Their fix was a fit guide featuring 100 photographed bodies. Find one like yours, click it, see two products. One investment that removes the practical barrier and the emotional one at the same time.

We unpack what those invisible conversion barriers can look like and how to solve them in this Playbook.

Examples from this Playbook

🧴 Chloe from NALA built the fit guide to answer the one question stopping her customers from buying. It also captures life-stage data and flows customers through Klaviyo as their needs change.

🔍 Andy from humii co-built a mystery shopping platform because standard analytics tell you where customers leave, not why. A US brand running paid acquisition in Australia had no idea half its shoppers were bouncing on a "we don't ship here" pop-up.

👗 Molly & Emma from JAM the Label built for the 15% of the global population with a disability that fashion has always excluded. When the experience earns the community's trust, the advocacy follows without a campaign.

🐦 Jane from birdsnest built a style profile as a deliberate first-party data foundation. Her honest take: "We had all this great data and now we finally have the tools to do something with it."

🎧 Spotify | 🎧 Apple | 📺 YouTube

Reply

Avatar

or to participate

Keep Reading