Our QA checklist: 80 things we test before every launch

If you take one thing away from this article, take this: our qa checklist: 80 things we test before every launch is not about chasing the latest trend — it is about applying a small set of principles consistently, with discipline, across every page of your site. We have shipped this exact playbook on dozens of client builds and it works whether you sell handmade soap or enterprise SaaS. The brands that compound month over month are not the ones with the most clever homepage; they are the ones who never let basics slip.

Most teams treat studio as something you do once and forget. That is the single biggest reason sites stagnate. You launch, you celebrate, the analytics dashboard sits unopened for six months, and by the time someone notices a problem the gap to your competitors is enormous. The teams who win are the ones who treat the website like a product, not a project — they ship, they measure, they iterate, and they refuse to call anything "done."

Here are the moves we make on every project in this category:

1. Cross-device sweep. This sounds obvious in a checklist, but in practice almost nobody does it well. The bar is low because most teams confuse activity with progress. Get this one right and you will be ahead of 80% of the sites you compete with. We always start with the easiest version of the change, ship it, measure the impact for two weeks, then escalate to the harder version only if the data justifies it.

2. Forms and emails. This sounds obvious in a checklist, but in practice almost nobody does it well. The bar is low because most teams confuse activity with progress. Get this one right and you will be ahead of 80% of the sites you compete with. We always start with the easiest version of the change, ship it, measure the impact for two weeks, then escalate to the harder version only if the data justifies it.

3. Speed budgets. This sounds obvious in a checklist, but in practice almost nobody does it well. The bar is low because most teams confuse activity with progress. Get this one right and you will be ahead of 80% of the sites you compete with. We always start with the easiest version of the change, ship it, measure the impact for two weeks, then escalate to the harder version only if the data justifies it.

4. Accessibility checks. This sounds obvious in a checklist, but in practice almost nobody does it well. The bar is low because most teams confuse activity with progress. Get this one right and you will be ahead of 80% of the sites you compete with. We always start with the easiest version of the change, ship it, measure the impact for two weeks, then escalate to the harder version only if the data justifies it.

5. Broken link scan. This sounds obvious in a checklist, but in practice almost nobody does it well. The bar is low because most teams confuse activity with progress. Get this one right and you will be ahead of 80% of the sites you compete with. We always start with the easiest version of the change, ship it, measure the impact for two weeks, then escalate to the harder version only if the data justifies it.

6. Analytics verification. This sounds obvious in a checklist, but in practice almost nobody does it well. The bar is low because most teams confuse activity with progress. Get this one right and you will be ahead of 80% of the sites you compete with. We always start with the easiest version of the change, ship it, measure the impact for two weeks, then escalate to the harder version only if the data justifies it.

On the engineering side, the implementation matters as much as the strategy. We default to TanStack Start with React 19 and Tailwind v4 for new builds because the SSR story is rock-solid, the type-safe routing prevents whole categories of bugs, and the design system tokens flow cleanly from design into code. For Shopify projects we lean on Dawn-derived themes with custom Liquid sections; for WordPress we standardize on a stripped-down Gutenberg base with ACF Pro for editor flexibility. Whatever your stack, the principle is the same: pick a small number of tools you know deeply, and resist the urge to chase every new framework that lights up Twitter.

Measurement is where most teams quit too early. You need a baseline (two weeks of clean data, minimum), a hypothesis written down in advance, and a guardrail metric so you know if you broke something else while improving the headline number. We log every test in a shared doc with the date, the change, the variant, the sample size, and the outcome. After a year that doc becomes the single most valuable asset in the company because it teaches the team what actually moves the needle for your specific audience — which is almost never what generic playbooks say.

The pitfalls are usually self-inflicted. Teams over-engineer when they should ship, ship when they should think, and think when they should be talking to customers. The single most common mistake we see is changing five things at once and then not being able to attribute the result to any of them. Move slowly. Change one thing. Measure. Repeat. The compound effect over twelve months will leave the "move fast and break things" crowd in the dust.

If you want a hand applying this to your site, that is exactly what we do at Weblystic — fixed-scope, fast turnaround, conversion-focused builds and ongoing optimization. Use the calculator to get a price in two minutes, or send us a note and we will reply within one business day with concrete next steps. We would rather help you ship one thing this week than sell you a six-month strategy deck.