In partnership with
Apple now lets app makers make a case for staying when a subscriber tries to cancel through the App Store. A reminder is worth testing first because a discount can cut the price for people who would have stayed anyway. Check whether the offer earns more when customers face the regular price again. YouTube’s product suggestions raise a separate question for retailers. How do creators find a product before the commission can help persuade them to tag it?
News
1️⃣ Apple lets app makers show a reminder or offer when App Store subscribers try to cancel
2️⃣ YouTube suggests products that eligible creators can save and tag for commission
3️⃣ Amazon lets invited U.S. shoppers ask Alexa for products and approve purchases by text
Insights
4️⃣ Meta’s Muse cited more sources per shopping answer than ChatGPT and gave shorter replies
5️⃣ UK and U.S. fashion buyers in a survey picked AI price alerts more often than outfit suggestions
6️⃣ In a survey, personal recommendations made people more likely to consider a brand than a top Google result
LET’S ENTER THE MAZE!
1️⃣ News
The Maze:
Since November 4, Apple has let app makers add a message or deal to its cancellation screen for subscriptions that renew through the App Store. Subscription teams get one more chance to explain why the service is worth keeping or offer a lower price.
Source: Apple Developer. Apple’s examples show the cancellation screen with an image or a discount offer. Both let the subscriber cancel.
Video: Apple Developer. Apple’s June 2026 developer session demonstrates how an app sets up the message and an offer in the dashboard, ahead of the November rollout.
The Details:
The subscriber still has a choice. After someone taps Cancel Subscription in Apple’s settings, the confirmation page can show a message from the app. They can continue to cancel, keep the plan or accept a deal if Apple shows one.
Apps set the offer rules. Teams assign one active message per plan in Apple’s dashboard. They can reserve a discount or free period for people who have paid for a minimum time. Apple reviews the text and displays the price separately to eligible subscribers. Choosing a different message for each person in real time requires separate access.
A save measures a decision at the screen. In June, Apple reported that subscriptions using these messages saw the share of people who stayed at confirmation rise by 1.4 percentage points on average. Results varied. Those figures predate November’s wider launch. The dashboard’s effect on paid renewals is still unknown.
MarketMaze Take:
Apple has made it easier to pitch a subscriber who is already on the way out. For subscription teams, a reminder about a benefit gives them something to compare a discount against. The discount has to retain enough extra paying customers to make up for the lower price given to those who would have stayed anyway. The measure to watch is what people pay after the offer ends.
2️⃣ News
The Maze: YouTube’s November 6 creator update revisited the product suggestions it said were rolling out on October 26. The list sits in YouTube Studio, where eligible Shopping affiliates manage their videos. It gives retailers another chance to get their products in front of creators.
Video: YouTube Creator Insider. YouTube’s October 26 announcement explains who gets product suggestions, why the picks vary and how creators can save products to tag later.
The Details:
YouTube makes the shortlist. It says it uses the channel’s past content, how viewers respond and what is popular in its category. Creators can compare estimated commissions and save products for a later upload. They still choose which items to feature and tag with a shopping link.
A tag can lead to a sale. A creator who reviews kitchen gear could save a suggested kettle, then feature and tag it in a video. In the retailer-site setup Google describes, a viewer can follow that link to the retailer. A qualifying purchase earns the creator a commission.
Eligibility sets the boundary. The list is rolling out to eligible YouTube Shopping affiliates. YouTube gave no country list or completion date. Creators can also search for items while uploading a video, so the list adds a way to discover products.
MarketMaze Take: A retailer can improve the payout a creator sees, but YouTube has yet to say whether paying more helps a product appear in the first place. That matters when setting commission budgets. A commission increase costs the retailer money on every order that qualifies. Once a product gets tagged, clicks and orders can help a retailer judge the offer. The rollout announcements gave no evidence of extra sales from the shelf itself.
3️⃣ News
The Maze:
Amazon has started letting a small group of invited U.S. shoppers text Alexa for Shopping, its AI shopping assistant. They can ask for products and approve purchases in the conversation. For sellers, Alexa’s reply becomes another place a shopper can discover an item.
Source: About Amazon. Amazon’s example shows Alexa replying to a gift request with a specific product and an option to buy it.
The Details:
Shopping moves into the message thread. The assistant already worked in Amazon’s app and on its website. Invited shoppers link their phone number and text a contact called Alexa Shopping. Amazon says millions more U.S. customers will get access in the coming weeks.
A request can become an order. Ask for a gift, send a photo or request a repeat purchase. Alexa can compare options, add a product to the cart and place the order once the shopper approves. Shoppers who set up automatic buying can let Alexa purchase an item when it hits their target price.
Ads can appear while shoppers choose. Digiday reports that an Alexa reply can include a Sponsored Product, Amazon’s product ad. Amazon is initially offering the text placement at no charge to advertisers. The report says purchases Alexa completes on its own stay ad-free.
MarketMaze Take:
A shopper who asks Alexa to find a gift gives brands a chance to appear in the answer. A shopper who asks for the same detergent as last time has already made much of that choice. If Amazon starts charging for text ads, sellers will need to see when their products appeared and whether the ads won a sale that would otherwise have gone elsewhere.
4️⃣ Insight
The Maze: AirOps, an AI content-workflow company, tested the same buying questions across four AI engines on October 23–27, 2026. Meta’s Muse drew on more sources but gave shorter answers.
A broader source pool: Muse averaged 25.1 citations—references to source material—per answer, versus ChatGPT’s 6.9. It cited 10,171 distinct website domains across the test, versus 2,405 for ChatGPT; those domain counts describe the whole test, not each reply.
More customer material: Review and user-contributed-content sites supplied 17.6% of Muse’s citations, versus 3.4% for ChatGPT. That is a share of citations, not shoppers or sales, suggesting brands’ own pages are only part of their potential representation.
Less conversation: Muse averaged 777 characters per answer, versus ChatGPT’s 5,192. Only 11.4% used a list or table and 0.2% asked a question back. Sample size, geography and settings were not disclosed; brief answers do not prove better purchases.
Why it matters: Muse combined wider sourcing with shorter answers. Brands may need to compete both to enter its evidence pool and to survive a brief recommendation.
5️⃣ Insight
The Maze: UK and US fashion shoppers value artificial intelligence for price help more than styling. Savings tools could also make retailers' prices harder to defend.
In ACI Worldwide's June 2026 online survey, fashion questions covered 3,194 UK/US clothing, footwear or accessory buyers aged 18–65. Shares finding each AI feature useful were 35% for price-drop alerts, 35% for retailer price comparisons, 18% for personalized products and 17% for outfits. People could select several; the figures are not pieces of one total.
Alerts help shoppers wait for a discount; comparisons help them find a cheaper seller. Both offer a visible saving, but could delay a retailer's full-price sale or send the order elsewhere. The survey measures stated usefulness, not actual buying, and does not show why styling attracted less interest.
Separately, 79% agreed that better sizing, product information and reviews would reduce unnecessary returns. That is a belief, not a measured return reduction or proof that AI is needed. Accurate product facts are the starting point for helping people buy something they will keep.
Why it matters: Test whether price tools earn profitable repeat sales and better information prevents returns. UK/US survey preferences establish neither outcome.
6️⃣ Insight
The Maze: An October 2026 NP Digital survey of 983 people separates getting found from getting trusted. Personal recommendations lead search in willingness to consider a brand.
Familiar people carry the strongest influence. A recommendation from someone personally known scores +83 in the survey, against +43 for a top Google result. These are net percentage points: the share more likely to consider a brand minus the share less likely, not sales growth. Geography and sampling details are not disclosed.
Other recommendations remain positive. YouTube creators score +28, Reddit or online communities +27, ChatGPT or another artificial intelligence assistant +17, and TikTok or Instagram creators +16. The published results do not provide uncertainty estimates, so the near ties do not establish clear winners.
Discovery may need a second source of reassurance. The 40-point gap between personal recommendations and search suggests brands should test what convinces their own customers, alongside where customers first find them. The survey measures attitudes; it does not prove a purchase effect or justify cutting search budgets.
Why it matters: Ask customers both how they found the brand and what made it worth considering. A search result may get the click while a trusted person supplies the confidence.
🧰 Resources
Useful stuff, handpicked by the team every day
🤖 AI prompts: 10 AI Prompts for Organic Social Media or 👉 All here
📅 Events: E-Commerce Connect Polska 2026 or 👉 All here
🔓 Get the full resource library with MarketMaze+
🗞️ Quick hits
Everything else you should know
🛒 Checkout and shopping tools
Shopify added five ways to pay at eligible stores, including Alipay and WeChat Pay. The cashier selects the method, then the shopper scans a code and pays on their phone. Availability varies by market and currency.
Firmly raised $15 million and launched Ask Agent, a shopping chat for retailers’ own sites. It recommends items based on current prices and stock, and stores can let shoppers pay without leaving the chat.
📱 Platform rules
The EU’s preliminary findings say TikTok can give viewing time more weight than a viewer’s request to see fewer similar videos. The Commission suggested dropping viewing time from For You recommendations.
Meta paused four business terms updates due November 20, including a proposal to let it read websites associated with a business’s products or services. Existing terms still apply, and Meta gave no new date.
🏬 Retail operations
ThredUp sold 17% more items in Q3 than a year earlier. Its CEO said buyers paid $18 per item from June, down from $21 earlier in 2026, while labor and packaging costs took a bigger share of revenue.
Australia’s David Jones kept trading after entering administration. Suppliers owed for earlier stock now have to decide whether to ship again on orders approved by the lender’s appointed managers.
📣 Retail advertising
AppLovin said Q3 revenue rose 55% year over year to $2.18 billion. Ad spending by its mostly ecommerce consumer group at least doubled year over year, and AppLovin gave no separate ecommerce revenue figure.
THAT’S IT FOR TODAY!
You’re the reason our team spends hundreds of hours every week researching and writing this email.
What do you think of this issue?
See you next time in the maze!
MarketMaze team
