AI Personalization and AR Previews for Buying Art Online
🔶 Introduction
Artella is an AI-powered platform that helps users find the perfect artwork for their space. With the ArtMatchTool feature, users can scan their wall to get size, colour, and style-based art recommendations. AR previews let them visualize pieces at home before buying, making art selection easy and personalised.
My role
I contributed to the end-to-end design process, leading product scoping, user flows, wireframing, prototyping, and usability testing. Additionally, I conducted user research and synthesized insights to inform actionable design decisions.
Timeline
3 months
Tools
Figma | Google form
Maze | Figjam
Team
2 UX/UI Designers
Founder, Developer
🔶 Problem Discovery
People Find It Hard to See If Art Truly Fits Their Space.
During our early interviews and surveys, many users shared that buying art online feels overwhelming and impersonal. They struggle to imagine how a piece would actually look on their walls, whether it fits their space, or if the style truly complements their home. Browsing endless options without any guidance leaves them feeling unsure and frustrated. As a result, they often settle for generic art that doesn’t feel meaningful – or they give up on buying altogether.
🗨️How might we turn the process of art selection into a personalized experience that connects art seamlessly to a user’s space and style??
🔶 Our Solution | AI / AR
Art Match Tool, A personalized and immersive art-buying experience.
Discover, preview, and select the perfect artwork with confidence using AR and smart recommendations.
🔶 Our Solution | No AR device needed
Upload your space, choose your favourite art or get AI-curated picks, and preview before you buy.
Design Process
🔶 Market Research and Frame the Problem
Research highlighted how AI-driven personalization and AR previews reduce choice overload, enabling more confident and satisfying art purchase decisions online.
Applying design thinking, we started with literature and competitive research to uncover user needs and behavioural barriers in online art purchasing. These findings were later validated through user interviews to guide our solution direction.
🔶 Analysing Competitors
Market analysis showed competitors focus on generic listings, with few offering personalized AI recommendations or AR previews.
We hypothesized that integrating AI personalization and AR previews could differentiate Artella by making art selection easier, more interactive, and tailored to each user’s space.
.png)
.jpeg)
While the AR solution is helpful, better user experience during wall try-on is still lacking. The retry button supports users in finding the best fit in the end.
Artfully offers promising features such as in-room art previews on the desktop version and an 'AI Search' tool. However, features like 'More Like This' are not effective for finding personalized art, as they present too many options instead of narrowing down to the most relevant ones.
Focused on easy browsing with good preview and artist details, Society6 falls short in offering deeper personalization, customization, or recommendation tools.
Beyond Cost: Why AI & AR Still Struggle in Online Art Shops
while high implementation costs keep many retailers from adopting AI and AR, existing cases also fall short. Most tools act as novelties rather than decision-making aids leaving a clear gap for seamless, solutions that genuinely help users choose art.
🔶 Validating Our Assumption
Interviews with 20+ participants revealed the AR Alone Doesn’t Close the Deal in buying Art Online.
We discovered that users often lacked confidence when choosing art, feeling overwhelmed by options and unsure about fit with their space and style. While AR previews helped them visualize artworks, they didn’t fully support the decision-making process.
We discovered that users often lacked confidence when choosing art, feeling overwhelmed by options and unsure about fit with their space and style. While AR previews helped them visualize artworks, they didn’t fully support the decision-making process.
🔶 Identifying the pain points
Most users felt overwhelmed and uncertain when choosing art, which led to decision fatigue and inaction.
People are Space-Centric
Users are primarily driven by how well the artwork fits their own space and personal aesthetic. Recommendations that don’t feel aligned with their taste are quickly dismissed.
People Needs to Resonate with Art
People need to see themselves in the art they choose, in color, in style, and in how it fits their space. Without that sense of resonance, art becomes just decoration, not expression.
No Confident, no Decision
Without confidence or visual context, users delay decisions, endless browsing and uncertainty often lead to no purchase.
🔶 Approaching a Solution
How can we support the ideal art-buying experience?
Leveraging our research, we found that the best solution would reduce uncertainty by linking art choices directly to a user’s own space and style. Our approach focused on visual relevance and personalization letting users upload or show their space, receive AI-curated suggestions, and instantly preview artworks in context.
Introducing an AI-powered art style guide.
A tool that starts with 5 quick question quiz gives artwork options based on space basic features like size and color with AI-powered.
The Art Match Tool lets users upload or scan their wall or interior to detect its size, color, and style, then receive AI-curated art suggestions that match their space. With AR previews, they can instantly see how each piece would look in context, making selection faster and more confident. Guide Adding some questions to understand user taste.
Guiding the process
AI Art Match tool 4 primary questions
🔶 System Architecture
🔶 Iterations from Usability Tests - 5 rounds
1-Relocating Size and Frame Options
Moved size and frame selection closer to the artwork preview to improve visibility and ease of use. Helped users notice and interact with customization options more naturally.
2-Layout Shift: Vertical Toolbar to Horizontal
Moved size and frame selection closer to the artwork preview to improve visibility and ease of use. Helped users notice and interact with customization options more naturally.
.jpg)
3-Add Gallery Wall Suggestions
When users preview their selected art in their own space, we now show curated matching pieces as a gallery wall option. This helps them visualise a fuller look and explore related artworks easily.
Web App Iteration: Needs to be Visually Appealing
Using dark colors draws focus to the artwork and supports better decision-making, while icons enhance quick understanding and faster user response. An inspiring quote from the artist adds a layer of meaning, emotion, and inspiration to the user’s experience.
Final Web Version
Tasks Pages From Choosing the Art to
Faceted navigation system -
Wall Gallery and Single Art
Landing page - product Detail
Ability to try on wall with AR
When They have their house picture to try their desire art on the wall in desktop - Suggested wall gallery is available.
🔶 Project Takeaway
Design decisions become far more meaningful when grounded in real user context, not just abstract assumptions.
The project began with extensive exploratory research into personalization, AR technology, and online art-buying behaviors. While some early hypotheses didn’t align with MVP needs, user interviews and contextual insights proved essential in shaping valuable, focused ideas.
Understanding how people actually engage with choosing art their space, uncertainty, and emotional connection was what truly informed the direction of our design.