Avi Moskowitz

"We A/B tested it and there was no lift." Cool. How long did you run it? On which segment? At what time of day? With what AOV cohort? Most "failed tests" aren't failed tests. They're underpowered tests run on the wrong audience.

ICYMI- prettydamnquick.com/product-update…

The checkout is the only page on your site where the shopper has their credit card out. It's also the page most brands touch the least. Insane if you ask me.

Today, many of them hold this company up. A few months ago we did it again, 9 new grads across engineering and CS. Skill, curiosity, and grit beat formal experience every time.

2 years ago, while everyone chased senior hires, we bet on juniors with zero experience and built a bootcamp.

Every Shopify brand is still competing with Amazon. Not on ads. Not on products. On the checkout-to-delivery experience. Amazon set the standard: personalized, fast, transparent. Shopify brands got static checkouts and siloed delivery. The future of commerce is Amazonian experiences without Amazon.

@PrettyDamnTweet 4/4 Your checkout is leaving money on the table. Pick it up. Cashback (+14% repeat), Subscription Upsells (+6%), Multi-Tier Progress Bar (+22%), Cart Control, Dynamic Reviews in checkout. Plus 7 more. Read it before your competitors do: prettydamnquick.com/product-update…

@PrettyDamnTweet 3/4 Your checkout is guessing. Stop. MAXX → autonomous AI operator. Analyzes your store, runs experiments, decides what to roll out. Theme A/B Testing → 3.1x faster than manual. Strategy Library → 200+ tested plays, ranked for your catalog.

18 features. One quarter. Here's what we shipped at @prettydamntweet in Q1, ranked by how badly your checkout needs it 🧵

@EcomExposed These tests are all on different stores

5 recent checkout tweaks, and the real results: 1. Free shipping bar raised: ~$3.12M annualized 2. Free gift unlock at $X: ~$892K 3. Mid-tier shipping option added: ~$980K 4. Segmented upsell test: ~$742K 5. Trust content for first-timers: ~$502K The biggest mistake is not testing in checkout at all.

@sumxxt It is, but keep in mind these are for different stores, so volume varies. The coupon rejection recovery is making huge saves for a few brands who are still figuring out the best ways to use it, but overall, even the generic "That doesn't work, try this" has proven to be huge

@MoskowitzAvi awesome, super interesting to see "free" stuff doing better than upsells since checkout tweaks only happen on plus stores. how is the coupon recovery thing doing?

In a few hours I'll be talking with Casey Hill from DoWhatWorks about some interesting A/B tests we ran recently Join us >> luma.com/v1oqhfi7

Most "best practices" in eCommerce are just vibes. Between us we analyzed 11,000 A/B tests to find out what actually works. Join me + @CaseyHill on April 30, we're going some tests and data live. Free. 1 hour. No fluff. luma.com/v1oqhfi7

Shopify makes checkout easy to set up. That's the problem. Easy means every store ends up looking the same. A first-time buyer from a paid ad gets the same checkout as a fifth-time customer on their phone at 11pm. The brands actually growing don't run it that way. They segment. They test copy, offers, shipping logic. They keep rebuilding it to see what works best. Because default is expensive.

@navdeepsoni @sumxxt Ah yes, rules are set by merchant, can be specific segments/orders/codes

@sumxxt @MoskowitzAvi sorry I meant - is it open for all customers.. can be misused.. someone entering random code and will get a valid code?

Shopper types "SAVE20." Doesn't work. Tries again. Caps lock. No caps. Opens Google. Finds a dead Reddit thread. Closes the tab. Gone. 46% of shoppers bail when a code fails. We just shipped the fix.

@sumxxt Nice! Would have to check, but what's your app?

@MoskowitzAvi i built an app that does the same thing on the cart itself - how many of your customers use coupons on carts vs on checkout?

"We optimized checkout last quarter" vs. "our checkout optimizes itself every day." That gap is the difference between good and great in e-commerce. The future isn't manual. It's autonomous.