@ShaharAlbeck One quick tell: if the store itself pauses its Shopify plan too, it's seasonal. If the store stays live and only your app goes, that's churn wearing a holiday costume.
@rod_mallo An agent can only weight a review if it knows who wrote it: a 4.9 from students and a 4.9 from parents with a stroller look identical today. Do you think platforms start asking reviewers for that context, or do agents infer it from the review text?
@SammyIsseyegh Which number broke first: cost per install or install-to-paid? The first means more apps bidding on your keywords, the second means the ads still bring installs, just from merchants who were never going to pay.
Black Friday = 4.6x your average October day. Median of 2,200+ brands (Fairing).
Selling in EUR/GBP? Plan 3-4x. Cyber Monday beats BF at 3 in 10 brands.
Best forecast: your own 2025 multiple, not the median. #Shopify#BFCM
@hifloecom One catch on restoring from git: if an app deleted a variant, recreating it gives it a new ID, so reviews, ad feed mappings and old order links tied to the old ID don't come back. Do you diff IDs too, or just the field values?
@oost_marcel@Mastercard@OpenAI@MiebachMichael Verifiable Intent is the piece merchants should watch. If it records what the buyer actually authorized, does that shift liability in "my agent bought the wrong thing" disputes away from the merchant?
@gilgNYC Where does the brand identity come from: pulled from the store's theme, or set by the merchant somewhere? If it's pulled from the theme, a messy theme now hurts you inside agents too, not just on your own site.
@stacktraceee The leather example is hard on the merchant side too: in many stores material lives in the product description, not per variant, so there's often no variant-level evidence to check. Does BuyRightAgents return REVIEW then, or fall back to product-level claims?
@zmbnski If retention weighs that much, a promo-driven install spike could cost you the spot a month later when the trial churn lands. Have you seen a rank dip after a burst of installs from a single source?
@v_d_mw@ShopifyDevs Eight years means they kept your app through the switch from vintage themes to Online Store 2.0. That's not a customer anymore, that's a marriage.
@drewsanocki Did the letter break out habitual buyers? If that group grew while active buyers stayed flat, the frequency gain sits with Etsy's best customers, which is a very different story than broad loyalty.
@ecomchasedimond Your store's search bar is a second objection log: the searches-with-no-results report in Shopify analytics often shows things like "returns", "size chart" or "ship to Canada", which means the answer exists but shoppers can't find it.
@tamir_eden Do you know what actually moves ranking in Catalog right now: mostly structured data completeness like GTINs and variant availability, or do store-level signals like reviews and return rates feed in too?
$5.1M per minute. That was the Shopify BFCM peak in 2025 ($14.6B total).
2026 forecast (MobiLoud): $18.7B, +28%. Adobe forecasts AI traffic to US retail sites +130%.
Make your store AI-readable before Nov 27. #Shopify#BFCM
@yeramianm I'd narrow it: single-feature apps without their own data are dead. If Sidekick can rebuild your feature from what's already in the admin, it will; if the feature needs data only you collect across thousands of stores, that's a moat.
@iamitporwal Has anyone checked how Sidekick installs show up in the Partner Dashboard traffic sources? If they aren't counted as App Store search, part of the install drop could be the same installs under a different label.
@tobi The real test comes in January: when the bot buys my cousin the wrong size, who starts the return, the bot, my cousin or me? Gift returns were built for humans holding a receipt.
@zmbnski Of the 1 in 4 stores that sold no more than a normal day, how many still ran a sitewide discount? That group gave away margin for zero extra orders, which would be the scariest number in the thread.
@ayysoni How much of the $38.5k is usage-based billing vs flat plans? If it scales with your merchants' order volume, BFCM lifts you without new installs, but January hands that lift right back.
@drewsanocki One catch on the "same flows" step: an email on an agent order isn't consent. Shopify passes along whatever marketing choice the buyer made inside the agent, or your regional defaults if none, so check how many of those buyers your flows can actually reach.