How Fyxer sped up their landing page experimentation velocity with AI

Fyxer, an AI email assistant, is one of the fastest growing AI companies in Europe.

In 2025, they grew from $1 → $32m ARR, by building a fantastic product that wrote your emails for you, and continually experimenting to find efficiency gains in their conversion funnel.

As they grew, so did their ambitions and product lines, and to support that, they needed to keep increasing their understanding of what converts through experimentation, primarily of their landing pages.

Getting marketing to move like engineering

Fyxer’s marketing team and strategy were world-class to build up the initial phase to get to product-market-fit.

As we entered 2026 however, some new challenges presented themselves. The company had undergone a rebrand to a fresh new style. This gave the marketing team the foundation to build more of a consistent identity, and therefore create more experiments that kept in line with the brand.

Alongside the rebrand, the demands on the engineering team grew as they scaled. The engineering team were at capacity in building out functionality that customers needed, and turning their processes into more of an AI-native one.

The marketing team, and specifically the Google Ads team responsible for creating new landing page experiments, needed to keep up to not fall behind internally and externally.

This was their existing way of working, in the pre-AI era.

An 8 step process that took weeks to complete. In the age of AI, waiting weeks for a landing page to come out wasn’t optimal.

Speeding up the creation flow with AI

The team have more ideas than they could realistically ship, so instead of asking the engineers to deliver more, we flipped the question and asked, how could we deliver more to the engineers in the first place.

Enter Claude Code, and the first case of touching code for anyone in the marketing team.

Until this point, only Sakky had been the ‘non-engineer’ to have touched code within Fyxer. Sakky had helped the overall marketing team ship new case study pages, working closely with the frontend engineering team.

That addition of Sakky into the codebase sped up capacity and unlocked speed, a crucial (if not the most crucial) ingredient in the AI-age.

So we embarked on the journey to get the Google Ads team set up onto the codebase, so they could start to commit ‘draft PRs’ to engineering team, after prompting in natural language completely new landing pages.

This was the new flow:

This shift, by allowing other teams outside engineering, to begin to commit their ideas in code, sped up the engineering team by over 50%.

The idea was simple, by changing the medium in which the idea existed in, there was less abstraction (from code) and less questions that needed to be asked.

The tech stack required

The primary applications required to set up this new workflow already existed in Fyxer (and would exist inside most companies that build some form of software).

This reduces the amount of overhead and new cost centers, and the bulk of the time taken was in setting up and upskilling employees on best practices for code.

• GitHub for code (existing)
• Vercel for hosting and deploy (existing)
• Conductor for orchestrating agents (free)
• Claude subscription (existing)