IKEA Content Experimentation
Experimentation and optimisation of IKEA’s retail online experience through triangulation of data
May 2022 - October 2023
IKEA Marketing & Communication
UX Researcher
Project summary
The Content Experimentation and Optimisation team within Inter IKEA Data Analytics & Information Management was specifically created to empower and inform content teams by providing insights and data into their creative processes, to enable data informed decision making. The team worked on many different initiatives, not all of them connected directly to the content teams but rather to a broader range of stakeholders across various different entities within Inter IKEA.
The overall goal was to advocate for analytics and insights and educate how beneficial experimentation and testing is to the business by enabling measuring not just performance of content but overall return of investment for assignments and projects.
Challenge
IKEA is a complex organisation and from a global franchisor perspective this can be challenging when it comes to achieving a global, consistent and cohesive user experience.
Inter IKEA, as the franchisor is responsible for and produces the majority of content that is visible to consumers on all channels and markets.
Research and data maturity within the content production teams is fairly low which leads to ideation of new content ideas without proper research or understanding where consumers actually experience frictions today.
Role
As UX Researcher and designer within the team I planned, facilitated and prototyped/designed material for all qualitative research activities and A/B tests.
I worked on a day-to-day basis closely together with Data Analysts and CRO Specialists to ensure triangulation of data and a mixed method approach to all of our use cases.
A major responsibility was also stakeholder management and supporting co-workers and colleagues understand how to work with research insights and data informed decision making.
Process
Experimentation is a repetitive process which continuously enables action-making from a cycle of research, ideation, experimentation and analysis.
Every use case that we received was somewhat unique so we customised and adapted our methods depending on the needs to be able to answer both the what and why during our cases. Whenever an A/B test was feasible, no technical constraints and enough traffic volumes, we aimed to provide statistically significant results.
By combining quantitative analysis, qualitative research and heuristic exploration we ensured that our insights were holistic, actionable and as operational as possible for the stakeholders with as little interpretation needed as possible.
The below visualised 5 phases are the core process we’ve applied as a framework to all of our use cases.
Measuring the impact
The cases we worked with have produced in-depth insights and have contributed to an increased data maturity level for stakeholders we have collaborated with.
Our A/B tests results were:
43% significant uplifts
28,5% inconclusive
28,5% significant decreases
Independently of the result of an A/B test, through applying a mixed method research we were able to give recommendations and actionable next steps for the stakeholders in all of our use cases.