Ismo

I build fast, conversion-focused Shopify themes. Solo engineer, building in public. Store teardowns, speed tips, before/afters.

[Ismo @ismobuilds]

@lobanovskiy The one on the right seems a "little" AI'ic.

[Ismo @ismobuilds]

There has to be a way to stop AI content creation automation. "Authorities".

[David G Quaid (SEO) @DavidGQuaid]

As I predicted - Google is handing out Scaled Content Abuse penalties like Florida hands out water

[Ismo @ismobuilds]

@sytaylor Walmart's complaint about losing the customer relationship still applies tho. Living on the merchant site fixes where the checkout happens, not who owns the conversation before it.

[Ismo @ismobuilds]

@sytaylor The first thing bit merchants will feel is that the agent reads the catalogue, not the storefront. All the money that went into layout and design stops mattering at that step, and the messy metafields nobody cleaned up start doing the selling.

[Ismo @ismobuilds]

@arjunaditya_ The other quick one is thumb reach. Add to cart often sits above the fold on desktop but below three blocks of copy on mobile, so many people never reach it. Both are cheap to check before you touch the UI.

[Ismo @ismobuilds]

@arjunaditya_ Before redesigning anything, open your product page on a real phone on 4G and time how long it takes to tap the buy button. That number usually explains more than the layout does.

[Ismo @ismobuilds]

@soltwagner I would start adding apps from the bottom row as per mobile view.

[Ismo @ismobuilds]

@johntech778 Placement point is right. One thing from the build side, trust badges above the fold usually mean another image request before LCP fires. Inline SVG or plain system text keeps the trust and skips the weight.

[Ismo @ismobuilds]

@paddymedia Bundle builders live or die on variant state. Most I've pulled apart re-render the entire section on every selection. It feels laggy on mobile at exactly the moment AOV is on the line. Doing it through the Section Rendering API instead is a very noticeable difference.

[Ismo @ismobuilds]

@KaiCromwell Agreed. On Shopify the thing that actually moves transactional terms is template-level anyway. The collection pages that aren't thin, product schema that actually validates, and not shipping 400kb of app JS that wrecks LCP. All theme work,none of it DR. I am working on it heavily

[Ismo @ismobuilds]

@DanielSmidstrup Only fix I've found is posting the work in the customer's language instead of the builder's.

[Ismo @ismobuilds]

@DanielSmidstrup Ran into this exact split. My timeline is builders, but the people who'd actually buy my Shopify themes have never opened #buildinpublic in their life. Builders give you reach, merchants give you revenue.

[Ismo @ismobuilds]

@matankatzzzz https://t.co/BRC92Se6Pu

[Ismo @ismobuilds]

@MicroLaunchHQ ThemeStack. Shopify themes built by one person who also has to support them, so nothing ships that I can't explain at 2am.

Early days, and the lesson so far -> merchants don't buy the prettiest theme. They buy the one that looks like it won't break during a sale.

[Ismo @ismobuilds]

@pcshipp My take on this is the silence until someone can actually use the thing, then never again. The worst spot is loud and pre-product. You spend the attention before you have anything to spend it on.

[Ismo @ismobuilds]

@pcshipp Depends what you're buying with the noise. Building in public buys distribution and early feedback; silence buys the freedom to change your mind without an audience holding you to last month's version.

[Ismo @ismobuilds]

@Shopify Curious whether the reference implementation reads the theme's own structured data or goes straight to the Storefront API. That distinction decides how much work themes actually need to do here.

[Ismo @ismobuilds]

@Shopify The thing this changes for theme developers, an agent doesn't see your layout; it sees your data. Product metafields, variant naming, structured data... the stuff most stores treat as an afterthought quietly becomes the storefront surface.

[Ismo @ismobuilds]

@nickselman Cheaper doesn't move anyone here because the switching cost isn't the subscription. It's the deliverability history plus every flow you'd have to rebuild while live revenue runs through it.

[Ismo @ismobuilds]

@lukecodez Concurrency, data modelling, why the query got 100x slower. These are exactly what you don't bump into until it's production and it's yours. Best case is both, in that order.

[Ismo @ismobuilds]

@lukecodez Half agree. A side project teaches what a degree can't shipping, scoping, saying no, the fact that "done" is a decision rather than a state.
But it only teaches what you happen to bump into. The things that bite at 3am