Case Study

FYVO

Turning disconnected optimisation into a single system that actually improves performance.

Most ecommerce teams rely on a mix of tools, manual effort, and instinct to manage content, SEO, pricing, and growth.

A product-led platform designed to unify optimisation, remove guesswork, and create continuous improvement across the entire store.

Client

FYVO

Sector

Ecommerce / SaaS

Role

Product Director, UX, Strategy

Services

UX, Product Design, AI Strategy, Information Architecture, CRO, Technical Architecture

Project overview

Building a system that replaces fragmented optimisation with continuous improvement.

Most ecommerce stores operate across disconnected tools and . Product content sits in one place, in another, pricing decisions are made manually, and data is rarely connected back into decision-making.

The result is inefficiency. Teams spend time creating content, updating products, and making changes, but very little of it compounds. Improvements are one-off, inconsistent, and difficult to scale.

Existing AI tools focus on generation. They create content, but they do not improve outcomes. There is no memory, no loop, and no connection between decisions.

FYVO was created to solve that problem. Not as another tool, but as a that connects across the entire store and improves it over time.

What was happening

Optimisation existed everywhere, but worked nowhere as a system.

Product descriptions were written once and rarely revisited. was reactive, driven by individual keywords rather than a structured approach. Pricing decisions were based on instinct or isolated points.

Teams were stretched, meaning became inconsistent. Some products were well maintained, others were neglected. There was no standard, no , and no visibility into what was actually working.

AI tools added another layer, but did not solve the problem. They generated content quickly, but without , validation, or connection to . Outputs looked good, but did not necessarily improve results.

The core issue was not lack of effort. It was lack of structure.

was not designed as a . It was a collection of disconnected actions.

Approach

Design the system first. Then build the product around it.

The starting point was not , but .

How products are created, how they are optimised, how decisions are made, and how improvements are measured. Every stage was mapped out to understand where value was lost and where could genuinely help.

The was designed around a clear lifecycle.

Ingest product .

Analyse and gaps.

Generate improvements with .

Review through human validation.

Publish with control.

Feed results back into the .

This was not about removing humans from the . It was about supporting them with better inputs, clearer decisions, and faster execution.

A was introduced to ensure quality and control. Every is reviewed, editable, and explainable before being published.

Underneath that, a rule and were designed to ensure outputs are consistent, structured, and aligned with specific goals such as SEO performance, conversion, and clarity.

The focus was always the same. Make repeatable, measurable, and scalable.

Key decisions

Build for improvement, not generation.

A critical decision was to move away from the typical generate-and-forget used by most AI tools.

Instead, FYVO treats every output as part of a .

Each is tracked.

Each decision is logged.

Each result feeds back into future outputs.

This creates a loop where the improves over time, rather than resetting on every use.

Another key decision was to separate approval from publishing.

Nothing is automatically pushed live. Users review, edit, and approve changes, then choose when to publish. This maintains control while still enabling speed.

The was also designed to connect multiple areas together.

Content, , pricing, and social are not treated as separate . They influence each other, and the reflects that. Keyword strategy informs content. Content influences conversion. Pricing affects positioning.

Finally, the was built to scale across large catalogues.

It needed to handle thousands of products without losing , while still allowing individual control where needed.

Solution

A connected optimisation platform built around real workflows.

FYVO integrates directly with ecommerce to ingest product and structure it into a unified .

Each product is analysed against multiple dimensions, including content quality, opportunity, keyword , and pricing position.

The generates structured outputs, including product titles, descriptions, , keywords, pricing suggestions, and social content. Each output is designed with a specific purpose and aligned to performance goals.

Users review changes through a clear , comparing existing and proposed content, with the ability to edit and refine before approval.

Publishing is controlled and deliberate, with a clear audit trail of what was changed and why.

Behind the , a rules engine, builder, and validation layer ensure outputs remain consistent, accurate, and aligned to defined .

Over time, feeds back into the , improving future recommendations and outputs.

The result is not just a tool, but a that continuously improves how a store performs.

Outcomes

A scalable foundation for continuous optimisation.

FYVO transforms from a manual, inconsistent into a structured .

Content becomes consistent and aligned with .

becomes proactive rather than reactive.

Pricing decisions are supported by rather than guesswork.

Teams spend less time creating and more time improving.

The creates visibility into what is working and what is not, enabling better decisions across the board.

Most importantly, improvements compound.

Instead of isolated updates, every contributes to long-term , creating a measurable impact over time.

Reflection

Optimisation only works when it is treated as a system.

This project reinforced that the problem was never a lack of tools or effort. It was a lack of connection between decisions.

Content, , pricing, and were all being managed separately, which meant improvements never compounded.

By designing the first, and building the product around it, FYVO shifts from reactive work to continuous improvement.

The role of UX here goes beyond . It shapes how decisions are made, how work , and how value is created over time.

Get the right, and the outputs take care of themselves.