When thinking of suitable analogies in web development, the one I typically come back to the most is Electric Vehicles - this especially applies to Jamstack adoption. With EVs, you typically pay a high price up front, but the operational & upkeep costs are less than with an internal combustion engine vehicle. There are edge cases, or concerns, like 'can I actually take this thing on a road trip?'(ie range anxiety). Best of all, it's better for the environment ( but not always!).
Let's dig into how this all relates to technical debt. You see, with an internal combustion engine, it requires alot maintenance and upkeep. The engine, the brakes, everything. There are more moving parts and thus require more maintenance than an EV. Let's say you own the car for 5 years, and you have a routine maintenance scheduled twice a year (or every 5k miles). All those checkups, gas etc start to add up. A similar principal applies to websites.
Say you have an idea for an ecommerce site - you secure a domain and install a free shopify theme and away you go. A modest start for $29/mo plus the $10 domain (renew annually) and you're looking at $358 of operating costs in your first year. You start to get a loyal customer base and decide to "spruce up" your site and hire a developer. Depending on scope, this can be thousands of dollars and 1-2 months to complete. Hey, it's your brand and it's worth it! But say you are running into slower performance speeds, or you wanted to create a new feature or section on your site and aren't able to address either with the theme you purchased? That's an additional cost you have to eat. It's also important to note the cost applied to your users. Most people are going to be viewing on their smart phone. The slower the site, the more bandwidth it's eating up, and the more money you're costing them (and you).
According to Google, improving load time by 0.1s boosts conversion rates by 8%. An overhaul, or site migration, from a theme to headless ecommerce are typically marketed and best suited for Shopify Plus clients - or those with strong enough revenue numbers to justify the overhaul. But what if there are edge cases that haven't yet been considered?
For instance, I have two storefronts I'm currently operating from one Shopify backend. My operating costs are extremely limited - $9/mo for Shopify lite (I actually paid for 12months so it's 10% less than that number). Deployment and continuous integration is managed in the cloud via Netlify - which is free (their free tier is extremely generous - 300 build minutes per month).
One of the cooler things Netlify has offered, as I covered in my last post, is the click to install button. This brings headless jamstack adoption to the masses! All you have to is click the button, and if you dont have a github account or netlify, you can sign up through the link. Click it, sign up, and bam you have a performant headless ecommerce site which should blow your old Shopify theme out of the water. An additional advantage is now if you require a developer to add features or troubleshoot any bugs, you already have a Git repository with the source code ready to be accessed. Great developer experience too.
I've been extremely bullish on headless Shopify for some time now, and given the fact that Shopify doesn't even allow devs to submit themes to their platform anymore, I think they probably are too. While adoption is starting to see an uptick in enterprise Plus clients, I think there are edge cases for more mainstream adoption. It could be a mix of WaaS (Website as a service) type themes, where users can one click install a theme. This could cover $0-200 entry point. For more custom features, like different content, adjusting branding elements etc, this could cover the $500-4000 range as it would require manual implementation. $4k and up, generally speaking, would fit in the Shopify Plus/Enterprise level clientele.
TL/DR: Performance matters. Alot. If you have a DTC brand ready to launch with the intention of selling products and garnerning a sizeable marketshare, I would highly suggest looking into headless ecommerce.