OVERVIEW
Art Market Insights Platform is a seven-day, self-initiated redesign concept. It builds on an existing art-market research product to explore one question: if art-market research no longer only served a small number of professionals, but let more people understand the market faster, what would the product become?
The original product centers on auction data and provides deep analysis through custom research reports. But after re-examining the product experience and existing content, I found the platform had already accumulated a large body of art-market information, including auction records, news trends, exhibition information, artist profiles, and research articles.
The problem isn't a lack of data. It's that this content is spread across different systems and pages, so users often have to search, organize, and compare it themselves to slowly build a complete understanding of the market.
So this redesign proposes a new direction: if the existing content were consolidated into a single research entry point, and the business model and data experience rethought at the same time, could the product serve more kinds of users and research situations? The exploration spans product positioning, the business model, a reworked information architecture, and concept validation (POC) through a high-fidelity interactive prototype.
DISCOVER
In art-market research, auction price is usually the most immediate and accessible information, so the existing product is built heavily around auction data. But after walking through the whole flow again, I noticed a few things worth rethinking.
The current access runs on subscriptions and credits. A user has to enter the database, submit a request, then wait for the analysis team to respond. The model suits deep research, but it limits those who just want to grasp the market direction quickly, or start exploring.
A sale price matters, but it is usually only part of the market outcome. A work's exhibition history, provenance, cultural context, and market attention all shape how its value is understood. And most of this already exists on the platform, just filed under different content and entry points.
Beyond auction data, the platform has also built up a lot of news, exhibition records, research articles, and artist data. But when a user wants to dig into a single topic, they still have to switch between pages and organize the information themselves.
The legacy funnel: paid search at the core, with conversion relying heavily on users requesting reports themselves.
Initial insight
It gradually made me realize that what is really missing might be the connections and context between data.
DEFINE
Problem statement
When I research beyond price, I want non-price signals too, so I can understand what really shapes a work's value.
HMW
How might we add non-price indicators to the existing analytics, and make them meaningful for users with different goals?
The goals and principles below set the scope: reorganize the existing content into a new research experience, and test whether that opens a new business model.
The legacy user flow: after searching, users fill out a form to request a report, which the internal team then handles manually.
Design goals
Design principles
DEVELOP
To help users quickly recognize where information comes from and what a feature does, the system extends two core colors into a complete set of semantic ramps. Data Blue handles all market and data content, AI Violet stands for the AI-assisted analysis features. Every step passes a contrast check against dark and light backgrounds, keeping reading consistent.
After the redesign, the Insights Card is split by perspective into Auction Insights, Gallery Spotlights, and an Overall Summary, so galleries, collectors, and market researchers can start from the angle that fits their own needs, instead of facing a wall of raw data first.
Placing the redesigned version alongside the old one: the new direction adopts a higher-contrast color scheme and refines the interaction logic of the filters, so power users can control each dimension more intuitively and smoothly.
The core users still include gallery operators, auction specialists, art advisors, and investment institutions. So the interface doesn't deliberately reduce the amount of information; instead it reorganizes the information hierarchy, the filter interactions, and how charts are read, so multi-dimensional information can be understood and compared in one view.
Price is only one outcome, so the redesign surfaces non-price signals too. The Attention Index reads momentum from press mentions, exhibitions, art fairs, and museum acquisitions; and AI insights sit on top of the charts users already read, where tapping Ask AI opens an assistant that answers from the same news and report sources behind the data.
DELIVER
The final output is a fully operable, high-fidelity interactive prototype, covering the Insights Card, the subscription-upgrade flow, the analytics dashboard, and the AI research interface, and it supports desktop, tablet, and mobile. The prototype mainly validates whether, once the existing content is reorganized, it lowers the barrier to research and creates a new way of using the data.
View the live prototype
Future metrics to watch and validate
Since this is still a proof-of-concept stage, if it moved into productization, these directions could be watched first:
REFLECTION
This project made me rethink the value a data product really provides. A lot of the time, the problem is that users lack a way to build understanding quickly. AI's value is in helping organize, search, and connect information, while the final judgment and decision still come from human experience and expertise.
Questions worth exploring next
These are also directions worth continuing to validate.