I worked with Farfetch to explore building a global design system in the lead up to their IPO. My initial role was to explore value and feasibility. Once validated, I lead the strategy to define what it should be and laid the foundations to get it up and running.
This case study is broken into five sections:
You will also find some working photos and client feedback at the end. While working on this project I held the title “Design Consultant” as the responsibility was focused around research, strategy and advice leveraging my experience as a principal designer.
When I joined Farfetch they had been creating digital products for over ten years. During this period, teams multiplied in size exponentially and spread across the world. Naturally, this presented a more challenging environment for strong communication and close collaboration. As a result of this, products had become inconsistent, lower quality than they were capable of producing and built inefficiently.
- Product quality decreasing
- Cost per release cycle increasing
I was responsible for:
- Leading cause identification
- Defining strategy and solution
- Forming cross-function central team
Design System Site
The culmination of my work at Farfetch led to designing, building and launching the first version of a company wide design system.
The design system was composed of styles, components, guides and patterns that I worked closely with the principal designers and creative director on.
Before commitment was made to build a design system I had to deliver a robust valuation of what we stood to gain from doing so. I later went on to publish the methodology as an article for UX collective and a presentation which I gave at several conferences.
I delivered a comprehensive feasibility assessment, inclusive of risk profiles, based on a deep exploration of current practice cross-referenced with the journeys of many other organisation.
I delivered a suite of artefacts and presentations on the rationale for creating a design system. These were specifically tailored to address four key audiences; designers, developers, product managers and business stakeholders.
I delivered an inclusive strategy that aimed to consume and leverage work already completed in this space. This enabled faster delivery, garnered support from those already advocating and made adoption less complex.
A primary deliverable was to define what should and shouldn’t be included in the design system both initially and ultimately. For each item I also had to clearly define and agree the definition to limit miscommunication.
I delivered an adoption strategy that offered a flexibility and choice to the product teams with a view to achieve full alignment within 12 months.
I was responsible for defining the operational strategy, contribution process and management of the design system. This included building a dedicated team and building specific lenses on a core repository for different audiences.
I delivered an objective based roadmap for the design systems team to deliver against, setting clear goals to eventually reach a point where all internal teams had adopted and it was available to external teams.
After completing research, design, planning and validation I had to pitch the solution and approach to the senior leadership team before executing.
My final deliverable was a comprehensive write up of the entire process, plan and valuation to allow investors to better understand our operational strategy.
Analyse Current State
In assessing the current state of our product design and production there were five main questions I sought to answer:
- Were we making the best products we could possibly make?
- Were we delivering a consistent experience?
- Were we making products as efficiently as possible?
- Did we embody our company values throughout the design and production process?
- Did people in the product design, development and management community know of Farfetch and if so do they view us as a desirable place to work?
This research phase included:
- Design auditing
- Technical auditing
- Stakeholder mapping
Following an investigation into the current state, we formulated the objective to define and create a system, process or product that allows for the continual improvement of:
- Consistency of products and practices
- Efficiency of delivery
- Quality of output
- Impact of outcomes
- Perception both internally and externally
Once we had a clear objective I set about exploring how best to systemise an approach to the Farfetch product development lifecycle. After validating that a design system could indeed be appropriate in the current state analysis I looked to answer:
- Who would be a users of the design system?
- What would the users need to make this system useful?
- How would the design system affect or integrate with their current workflow?
- What is the current level of knowledge and awareness around design systems?
- Are there any existing efforts to systemise the product development process?
- What would have highest impact:effort ratio?
- How would we achieve buy in from all key stakeholders?
- How would we garner support and adoption?
- What are the risks and requisites?
This research phase included:
- Further shadowing analysis (from previous research)
- Affinity mapping
- Needs analysis
- Trend mapping
- Champion identification
Based on the problems identified in the current state analysis, the goals agreed in the objective response and the needs identified in discovery, I outlined what I believed would be core requirements in creating a successful solution e.g. Continually capture research, knowledge and improvements made from all role types across all products and platforms.
The approach gave abstracted requirements for creating a solution that exist independently from how it should be approached logistically. To give this meaningful value I then outline what needed to happen logistically to bring a solution to fruition e.g. Build a design system explorer in a publicly available location that is separate from our current digital estate and able to exist without dependency on it.
Considering the different information, artefacts and resources we discovered throughout our research, as well as who would typically want access to them, I proposed a content outline and minimum viable product.
The content outline went through several iterations, informed by tree testing, completed by people who will ultimately be using the design system. I also scrutinised the content outline from the perspectives of different user types through the use of content specific user needs statements and task scenarios. I did this in a workshop format, with those who will be using the design system, informed by the information we gathered in earlier interview and shadowing sessions.
To bring the design system to fruition, there are a number of people and processes that needed to be in place. It also required buy-in from a range of teams and awareness to be risen across those who it will affect. Firstly, we did a roadshow of the research and proposal. Following the success of this we defined and built a dedicated team using a blend of current employees and new hires.
Design and Build
Once we had the initial team in place we designed and built the first version of the global design system. I functioned as the lead designer until this point.
Prior to departure we put together a roadmap based on what we’d learned that was focused on goals and outcomes. The new product manager then owned translating this into a timeline, features and deliverables.
From current state analysis:
- There was clear opportunity to better embody the “todos juntos” value
- There was clear opportunity to do more to “amaze customers”
- There was a risk of not being able to attract the best talent
- Farfetch didn’t appear overtly scalable from an external perspective
- Performance against industry leaders was poor
- Potential waste due to inefficiency runs into the tens of millions (£GBP)
- A design system had high value and feasibility
- There were over 20 user types to get value from a design system
- There was a clear prioritisation to wants and needs focused around providing objective truth
- Support on the ground is more important than support from above
- For maximum efficiency you need design, development and product buy-in
- If you can’t get total buy-in you can still achieve localised efficiency e.g. Design only UI Kits
- Not everyone has the same idea of what a design system is or should be
- Not everyone values the same thing so strong product management is required
- It helps to have a dedicated team but isn’t essential initially
- A design system has most utility when a user specific lens is applied
- Given how embedded a design system becomes, it’s worth completing the upfront research
Efficiency. Quality. Satisfaction.
Better quality products, delivered faster by less frustrated employees.
- Actualised yearly saving over £11M
- Product quality up 30%
- Employee satisfaction up 22%
“Bryn is diligent and logical, bringing structure to everything he touches. When a project gets technical and complex he really comes into his own with an ability to distill abstract requirements into a tangible plan. His contribution to this project was invaluable.”