Unleash Your Business Potential with Low Code

Published: June 15, 2023

I published the first version of this post in June 2023. It was a tidy little explainer: here is Shopify, here is Webflow, here are three reasons you might still need a developer. Nothing in it was dishonest. Nothing in it was checked either, and the claim that aged worst was the one about getting your stuff out. I wrote that Shopify makes it hard to export your data and that Webflow will not let you export your CMS content. Both platforms export now. Shopify pushes products, orders, customers and inventory to CSV from the admin, and Webflow shipped CSV export for CMS Collections with an update page that says, in Webflow's own words, you can move content between sites or "if you must" between platforms.

What I should have written is the version I now give people on the phone: the export button exists and it is attached to a subscription. Stop paying and it stops working. That is a different problem from the one I described, and it is the one that actually costs owners money.

So I am rewriting it, and this time around the questions instead of the tools. I take calls from business owners who are somewhere in this decision: about to build, halfway built, or holding a site a previous developer left in a state they cannot describe. They do not ask "what is low code." They ask seven questions, almost always in the same order, and the honest answers are more useful than another platform comparison.

Every answer below comes from either a documented platform behaviour I checked this week or a project I worked on myself.

"Can I build this myself and skip developers entirely?"

For a marketing site, a blog, a portfolio or a straightforward product catalogue, yes. Build it yourself. A well configured Webflow site or a Shopify store on a current theme will beat what a cheap freelancer produces for two thousand euros, and you keep the ability to change a headline without filing a ticket.

The boundary sits at the point where your business logic stops matching the platform's assumptions. Booking with rules about overlapping slots, memberships with tiered access, an integration that has to keep two systems in sync, pricing that depends on customer history. Those are the moments where the app store gives you something that covers 70% of the case, and the remaining 30% is what your customers phone you about.

My advice on that first call is usually to build the thing yourself and delay hiring anyone until you can name the specific behaviour the platform will not do. Founders who hire before they can name it end up paying a developer to rebuild a Webflow site in React for no reason anyone can articulate later.

"If I build on Webflow or Shopify, am I locked in?"

Your data is portable while you are a paying customer. Three separate things are being sold here, and owners tend to discover that in the wrong order.

The content export exists and it sits on a paid tier. Shopify sends products, orders, customers, inventory and discounts to CSV from the admin. Large exports arrive by email instead of downloading straight to your machine, and imports cap at 15 MB per file, so a big catalogue moves in batches. There is no free Shopify plan to run that admin on: a 3 day trial, then Basic at $19 a month billed yearly at the time of writing. Webflow exports each CMS Collection as CSV from the Designer with every field type supported, and images come out as URLs pointing back at the original project, so you download the assets separately or they break when the source site goes away. Having a CMS worth exporting means the Premium Site plan at $25 a month billed yearly. The free Starter tier caps you at 50 CMS items on a webflow.io subdomain, and Basic at $15 a month has no CMS at all.

Getting the code out is a second subscription. Webflow's code export is a Workspace plan feature, not a Site plan feature, so it stacks on top of the hosting you already pay for: Core at $19 a month billed yearly, on top of the $25 Premium plan running the site. Webflow's own pricing page describes the feature as exporting "clean, semantic HTML and CSS files for your dev team," followed by the sentence that matters: "Dynamic content (CMS items/pages) can't be exported." So the paid export gives you a static shell where Collection lists render empty, Collection template pages were never generated, native forms stop submitting, site search dies and User Accounts pages become public.

The exit is only open while the invoice is paid. Shopify's own documentation is blunt about frozen stores: miss a payment and you cannot access your Shopify admin. The admin is where every export button lives. A business that decides to leave because money got tight is exactly the business that discovers the door locks from the outside.

So the honest framing for a business owner is this. Assume your rows and your pixels are retrievable, on a plan, while the card on file still works. Assume every piece of behaviour gets rebuilt: checkout, apps, forms, member logins, the CMS wiring. The cost of leaving scales with how much custom behaviour you bolted on, and the deadline for leaving is whenever you stop paying.

Practical version: export your content on a schedule now, while everything is working and cheap, and keep the CSVs somewhere you own. It costs you an hour a quarter and it converts a lock-in problem into a migration problem.

"Does Webflow or Shopify make me accessibility compliant?"

No, and this is the question that did not exist when I wrote the 2023 version.

The European Accessibility Act has applied since 28 June 2025, e-commerce is explicitly in scope, and the technical bar is EN 301 549, which points at WCAG AA. The exemption for microenterprises is narrower than most owners assume: fewer than 10 employees and turnover or balance sheet under EUR 2 million, services only, both conditions at once, with no grace period once you grow past the line.

Platforms give you accessible defaults. They cannot give you compliance, because the failures I find in audits are almost all merchant side decisions. Custom interactions that override focus outlines. Heading levels chosen because of how big the text looked. Brand colours at 3:1 against white. Alt text left empty across a whole product catalogue. An app that injects a modal nobody can close with a keyboard.

Automated scanners catch somewhere between a quarter and a third of WCAG issues, which is why a green Lighthouse score reassures people who then fail a manual audit. When I audited X.com for accessibility I was working on a site with a large engineering team and a design system, and the keyboard and screen reader failures were still there. A theme purchase will not save a store from that. I covered the fix list in detail in how I made CeHDI WCAG compliant.

"Why can't I change my checkout?"

This one comes up more than any other Shopify question, usually with real frustration behind it, because the checkout is where the money is and it is the least editable screen in the store.

Shopify finished deprecating checkout.liquid. Information, Shipping and Payment pages moved to Checkout Extensibility first, and the Thank you and Order status pages followed, with automatic upgrades in January 2026 that dropped customisations built on additional scripts and script tag apps. Stores woke up to a checkout that had quietly lost its upsell block.

On a standard plan you now get Shopify's checkout with the settings Shopify exposes. Custom content blocks, extra fields beyond the defaults and conditional logic are Plus features, delivered through Checkout UI Extensions and Shopify Functions. Shopify lists "fully customizable checkout" as a Plus line item, and Plus starts at $2,300 a month. That is a plan upgrade plus development work, not a theme edit.

The logic owners want at checkout is rarely exotic. Most of the requests I have handled around payments involve subscription terms, coupon rules that depend on what the customer already has and order forms that change shape based on earlier answers. On a custom build that is an afternoon of work. Inside a hosted checkout it is a pricing tier and an extension, and it is worth knowing that before you choose the platform rather than after your first promotion.

"The site is live but Google shows nothing. Is that the platform's fault?"

Usually not, and here is the part low code advocates undersell.

Webflow and Shopify serve rendered HTML. A crawler gets content on first request without executing anything. Last year I took over a Next.js site built by a previous developer where every page rendered client side only, so Googlebot had been receiving blank shells for months. The site looked completely finished in a browser. In search it did not exist.

That client would have been better served by Webflow, for an unglamorous reason: the platform makes the boring correct choice on your behalf. A developer who does not understand rendering will make the wrong one and hand you something that looks perfect in a browser. I wrote up the full teardown in I took over a project after a vibe coder.

When a Webflow or Shopify site genuinely does not rank, the cause is usually thin content, duplicate collection pages, a missing sitemap or twelve apps loading 900 KB of JavaScript. All fixable, none of it a reason to leave the platform.

"What does hiring a developer at this point actually cost?"

Less than the internet tells you, if the project is unfinished rather than broken.

The rescue I keep referring to belonged to a policy organisation that had already gone through two developers. The site looked done. Underneath it, every headline and team bio was hardcoded into JSX, a Sanity CMS subscription had been billing for months with nothing connected to it, the contact form displayed a success message and delivered nothing, and the whole thing rendered client side. Fixing it meant modelling six content types in Sanity, replacing hardcoded strings with real queries, moving pages to server rendering and rebuilding the SEO layer. Roughly 50 hours across three months, and it cost less than what they had already paid the second failed developer.

That is the shape of most of these jobs. Owners brace for a five figure rebuild quote and what they need is a diagnosis followed by a targeted piece of integration work. Ask for the diagnosis before you accept the rebuild.

"Should I use an AI app builder instead?"

This is the 2026 version of question one, and it is the one where I push back hardest.

A low code platform constrains you into a working system. You cannot break the checkout, the hosting, the SSL, the rendering or the database backups, because you never touch them. That constraint is the product. An AI builder removes the constraint and hands you a codebase, and the ownership of everything the platform used to handle transfers to you the moment you press deploy, whether or not you can read a line of it.

For a prototype, a landing page or something you would throw away in a month, use them. For the system your revenue passes through, be aware you have chosen the option with the most surface area for silent failure and the smallest chance you will notice. I went through the industry data on that in why vibe coded startups fail, and what lands on my desk matches it: sites that look finished, produce nothing and cost more to diagnose than they did to build.

What I would tell my 2023 self

Pick the platform that already does the boring parts correctly, build as much as you can inside its constraints and hire a developer at the specific point where you can name what the platform refuses to do. Price the export path before you have 400 products in there, then run an export while the subscription is current, and treat accessibility as your obligation, because no subscription tier includes it.

Did a previous developer leave your project broken, or is a low code build hitting a wall you cannot get past? I take over and rescue Next.js and headless CMS projects: integration, server rendering, SEO, accessibility, the whole thing. Book a free 30 minute assessment and I will tell you whether you need a rescue, a finish or nothing at all.

Sources

Building with AI beyond this article?

I run The Agentic Architect Lab, live builds, agent workflows, and a playbook for technical founders shipping solo. No toy demos.

Join the Lab

← Prev Post Next Post →