|
|
Shopify doesn't back up your store. Rewind does, free to try → |
|
|
56,016 founders and marketers are getting this newsletter today.
Welcome to the 851 new operators who joined us this week! 🤯 |
|
|
| |
| |
|
|
Hey Fastlane Insiders! 👋 |
Welcome to Q4. It's October 1, Black Friday is eight weeks away, and Shopify shipped the biggest change to store design this morning since themes existed. Canvas lets Sidekick write and edit your actual theme files, every page at once, from a chat window. Custom design just stopped being expensive. |
Then a 460,000-answer study landed the same week and made the flip side hard to ignore. When ChatGPT and Google's AI choose between a big retailer and a small one, they pick the big one nine times out of ten. AI will build your store in an afternoon. A different AI decides whether anyone gets sent there, and it isn't judging your design. |
That's the thread through everything this week. The cheap parts keep getting cheaper: design, A/B tests, even checkout. What's getting more expensive is the part AI can't do for you. Being known, being specific, keeping your product data clean, and having the judgment to leave your store alone in the last stretch before Black Friday. One practical note on that: Sidekick editing theme files makes mistakes faster, which is why this edition's partner, Rewind, is the first thing I'd set up before opening Canvas. |
Whether you're doing $40K a year or $40M, there's a call to make at your stage in every section below. |
Here's what's inside this week's edition: |
🛡️ This Week's Partner: Rewind, and the one Shopify assumption that costs merchants whole catalogs eight weeks before BFCM. |
🎧 The Podcast: Why a 10% welcome offer beats 15%, from hundreds of millions of popup impressions, with Taras Talimonchuk of Claspo. |
💡 Knowledge Drops: What Shopify Canvas makes expensive now, and where small stores still beat the chains in AI answers. |
🔥 Tool of the Week: Plytix, a PIM built for multi-store Shopify catalogs. |
📡 Industry Pulse: Adobe's record holiday forecast, agent checkout on Shopify, SMB dead stock doubling, Caraway's lawsuit playbook, and the one word that flips AI recommendations. |
Let's dive in! |
|
|
|
|
|
| |
|
| |
| |
One thing I'd check this week. If you open Shopify Canvas and let Sidekick loose on a duplicate theme, make sure you back up the version you're working on first, like seriously… |
Which is a good time to introduce this week's partner. 👇 |
|
|
BFCM is 8 weeks out. Your backup plan should already be in place. |
Shopify doesn't back up your store data. One bad CSV import, a bulk edit gone sideways, or an app conflict can wipe out weeks of product work, right before the biggest shopping weekend of the year.
Rewind protects 25,000+ businesses in 100+ countries with automated backups, and restores anything from a single product to your full catalog in a few clicks. Moomin recovered 11,000 products that were mistakenly overwritten in 30 minutes. |
Try it today for free. |
|
|
|
|
|
|
| |
|
| |
| |
🎧 New Podcast Episode! 🎧 |
|
|
Why a 10% Welcome Offer Beats 15% (And What to Fix First) |
Most Shopify merchants treat the email popup as a checkbox. Install the app, set 10% off, turn on exit intent, move on. When the opt-in rate stalls around 2 to 3%, the instinct is to sweeten the discount. The data says that's exactly the wrong move. |
I sat down with Taras Talimonchuk, CMO at Claspo, an onsite conversion platform built by the team behind Promodo, one of Ukraine's largest ecommerce agencies, and the email provider eSputnik. Claspo sits on hundreds of millions of widget impressions, and Taras did something rare for a CMO: he separated what the data proves from what he's still guessing at. |
Here's what we unpacked: |
The smaller offer wins. Offers in the 1 to 10% band convert at about 7%. Push to 11 to 15% and it drops. Taras has two competing theories for why, and revenue per visitor settles the argument. Beware the 17% spike. The 16 to 20% band jumps to 17% conversion. There's a catch hiding in that number, and it isn't a reason to chase bigger discounts. The segment is the variable, not the offer. Paid traffic on a product page, an organic blog reader, and an LLM referral arrive with three different intents. Greeting all three with the same banner is the most common failure Taras sees. Your promo codes may be leaking. Static codes get scraped and posted on coupon sites within days. Taras explains the check most merchants have never run, and what it usually shows. The on-air correction. I credited Claspo with a Shopify badge it doesn't carry. Taras explains why they chose not to, and the tradeoff behind that call.
|
If you're a $500K to $10M brand stuck at a 2 to 3% opt-in rate, this conversation is your testing roadmap. If you're still on your ESP's default form, start with the segment question before you shop for a new app. |
|
|
|
|
|
|
|
| |
|
| |
| |
💡 Knowledge Drops of the Week 💡 |
|
|
Shopify Canvas Made Custom Design Cheap. Here's What Gets Expensive Now. |
For ten years, the hard part of a custom Shopify store was the code. As of this morning, that's no longer true. Shopify's Canvas announcement shows Sidekick editing your actual theme files across every page at once, and Ben Sehl, Shopify's Director of Product, says a merchant can now build a fully custom store in about twenty minutes. |
Say you're doing $400K a year and you've been sitting on an agency quote of around $8,000 (illustrative, but not unusual) for a darker, more editorial homepage. Canvas can produce a credible version in an afternoon. That's the opportunity. It's also the trap. |
Four things to know before you open it: |
Write the brief before you open Canvas. Sidekick remembers your preferences, so a vague first session trains a vague design partner. A few adjectives and one clear job per template change the output completely. Redesign churn is the new premature complexity. When a redesign cost six weeks and five figures, the friction forced discipline. Remove the friction, and stores between $500K and $2M end up changing three variables a week and learning nothing. Mind the Black Friday clock. Canvas landed eight weeks before November 27. There's a specific date where I'd stop touching product and cart templates, and a rule for how long to hold each change before you judge it. Know the gaps. It's early access and desktop only. Themes edited in Canvas don't receive theme updates, third-party themes aren't supported yet, and app blocks and embeds won't show up in a Canvas-built theme at launch. For most stores on this list, that last one is the real blocker.
|
The honest trade-off is that Canvas will happily build a beautiful store around a brand that hasn't decided what it is. Agencies that mainly sell implementation hours also have a pricing conversation coming. |
|
|
|
|
|
AI Picks the Bigger Store 9 Times Out of 10. Here's Where You Still Win. |
Canvas has an uncomfortable flip side. AI can now build you a beautiful store, but a different set of AI tools decides whether anyone gets sent there. A study of roughly 460,000 AI shopping answers found that when ChatGPT and Google's AI chose between one big store and one small one, they picked the big one 90 to 94% of the time. |
The coverage led with that headline number. For Shopify brands, the finding that matters is deeper in the data: the tilt worsens as the shopper gets more specific. |
What the numbers say: |
Escape the specificity trap. Small shops get named about 1 in 3 times on broad prompts. When the shopper names a specific product, that drops to roughly one in ten. If a chain also stocks your product, that's the product you're most likely to lose. Cited is not picked. Small and large stores were cited as sources at similar rates. Chains still took the top recommendation about 2.5 times as often. The piece explains the gap, and the one-page element that closes it. Win on needs, not product names. Shopify reports that 75% of AI-attributed purchases come from outside its top 100 categories. The brands winning there write product pages a certain way, and it isn't the way most Shopify catalogs are written. Watch your retail partners. When a shopper asks for your product by name, AI often searches the chains it already knows first. There's a low-cost page worth building before BFCM, and most brands on this list don't have it.
|
One caveat worth knowing: Lightspeed, which sells to independent retailers and competes with Shopify, commissioned the study. The method is published, but it measured "where to buy" prompts in four cities, not online-only brands. |
|
|
|
|
|
|
| |
|
| |
| |
🔥 Tool of the Week 🔥 |
|
|
A reliable system of record for your Shopify catalog |
Managing 1,000s of SKUs across multiple channels? Plytix gives brands, manufacturers, retailers, and distributors a powerful PIM without the enterprise price tag, complexity, or long implementation. Use our Magic Importer to bring your entire Shopify catalog into Plytix in one click. From there, your team can manage, enrich, and optimize product content with AI and automations, safely review changes, and sync to all of your Shopify stores and markets whenever you’re ready. Launch products faster, expand into new markets sooner, and finally keep your Shopify catalog under control. |
|
|
|
|
|
|
| |
|
| |
| |
⚡ This Week’s Industry Pulse ⚡ |
AI traffic to retail sites is forecast to jump 130% this holiday season, while Adobe expects total online spend to grow only 6.7%. The gap between those two numbers is where your Q4 attribution gets messy, and Adobe names the week it peaks. |
Browser-based AI agents can now fill in and submit checkout on eligible Shopify stores. Three new checkout tools are rolling out to all eligible merchants, and the announcement doesn't say whether your store is opted in. |
Dead stock now makes up 24% of small businesses' excess inventory, double what it was two years ago. Only 7% of the companies surveyed hit all four resilience benchmarks, and the one most of them miss has nothing to do with tariffs. |
Caraway got sued by two of the world's biggest cookware companies over its PFAS claims. Instead of going quiet, it turned the lawsuit into a campaign that pulled in more than 50,000 petition signatures. |
One word in a shopper's prompt more than doubled how often ChatGPT recommended an independent store over a chain, across 460,000 answers. It isn't "local," and it isn't "small." |
|
|
|
|
|
| |
|
| |
| |
Day One of Q4 |
Line up this week's stories and the pattern is hard to miss. |
AI is collapsing the cost of the visible stuff: store design, A/B tests, even checkout. What it can't manufacture is what sits underneath. A brand you can describe in one sentence. Product data an agent can trust. The judgment to decide what's worth testing and what to leave alone. As Taras put it on the podcast, the routine work goes to the agents, and the creative direction is always the part worth paying for. |
If you're early, write that one-sentence brand description this week and rewrite your top ten product pages around it. If you're scaling, pick one template and one data problem to fix before mid-October, then leave everything else alone until January. And whatever stage you're at, before you let Sidekick near your theme, make sure there's a version of your store you can get back to. |
One ask this week: run the test from the second Knowledge Drop in your category, then hit reply and tell me which stores ChatGPT named. I read every one, and I'll share the patterns, anonymized, in a future edition. |
…and thank YOU! 🙏 |
More than 56,000 of you give me a few minutes every Thursday, and on the first day of the quarter that decides the year, that means a lot. |
Some of you run brands, some build the apps and agencies behind them, and some are watching this industry change faster than ever. If you only keep one thing from this edition, keep this: |
Design just got cheap. Being the store AI trusts enough to send shoppers to didn't. Spend Q4 on the second one. |
Thanks for spending part of your day with me. I'm grateful to be in the fastlane with you! |
Cheers,
Steve |
|
|
|
|
|
|
|
| |
|
|
