Gallery profiles are the core experience of the ‘gallery network’, which is artnet’s second largest product in terms of revenue. Galleries from all over the world can apply to become a member, and once approved they pay a monthly subscription fee to connect to artnet’s vast audience of the world’s top art collectors and art enthusiasts via artwork and general gallery inquiries. Despite being used by artnet’s top clientele the product had not been updated in years to meet the evolving needs of the target users (collectors and enthusiasts). Collectors were now nearly always online with their smartphones with unlimited access to information, and expect the ability to explore artworks without having to be physically present. In order to remain relevant and on top of mind for their collectors, galleries needed to build out their own online presence and invest more of their time and resources to engage and curate their profiles on platforms such as artnets. In return, artnet needed to provide a platform that was representative of a gallery’s program and encouraged extensive engagement with their content.
It’s rare to be able to be a part of a company that is one of the pioneers of the early internet era (like Ebay and Amazon). Artnet, launched at the same time as these legends (more of which can be read here), and has experienced many of the same challenges (just on a smaller scale). Artnet’s seniority has given it a great advantage over the past few decades in an industry where heritage is everything in the form of deep industry relationships, vast amounts of data and most importantly of all, respect and trust in the 60B+ art industry. However, in order to stay relevant in a digital age that’s exponentially evolving, artnet needed to shift their strategy, evaluating each product in terms of risk and opportunity, and create a roadmap that would address fixing the most valuable areas of the company that were at the most risk - and gallery profiles was the first milestone.
My role
I led the project as the product design lead and collaborated with a visual designer to design the component library and the visual look and feel of the new design system. I led all UX research with the Product Manager, and collaborated closely with the front-end engineering lead, members of both data science and content strategy. I also involved non-technical stakeholders throughout the project such as the gallery liaisons (customer service representatives) and key members of the sales and marketing team who’s feedback greatly contributed to the success of this project.
Identifying the problem
It had been years since there were major updates made to artnet’s gallery profiles and its age was showing. The initial value proposition of offering gallerys a presence online, where they could upload their inventory was no longer in sync with our members needs and goals. Shifts in collector browsing behavior ( i.e. discovery and mobile browsing), as well as expectations in respect to what it meant to visit a gallery space online, were all indicators that we needed to do more than adding features to the existing pages or updating its sections. We needed to understand what it meant to do business as a gallery in the 21st century, what tools and experiences they needed to be successful, and to grow their business through an online platform (like artnet).
In order to get a holistic view and understanding on our current product and more importantly, how galleries operated both online and offline in the shifting market dynamics, I created a dedicated research project which consisted of members of the product and design teams, data science as well as key members of both sales and customer service i.e. our gallery liaisons. Over the course of 4 weeks we gathered research materials through the following methods
- Quantitative research from user surveys, NPS scores, data analysis of the current experience i.e. funnel analysis on current profiles, profile activity overviews and data mapping
- Qualitative research from structured interviews we conducted with current gallery members, non-members who were not on the platform as well as lapsed customers (within the past 12 months). During these sessions methods like card sorting were utilized to understand which content types were of higher importance. These sessions were recorded on video and shared internally for future reference and internally alignment.
- Competitive analysis from galleries during interview sessions as well as my own internal research on what key competitors (direct and indirect) were doing
The old experiences focused primarily on artwork listings and less on the galleries main activities i.e. creating exhibitions, attending art fairs. We discovered that many galleries were frustrated around the limitations of their profiles in terms of profile discovery and presentation of their content, and were discouraged to upload more than just the few artworks they had for sale in any given time. The overall decline in content uploads and gallery memberships seemed to be linked with their inability to have a robust profile on artnet. Combined with earlier feedback from our sales and customer teams, these discoveries formed the basis for our initial hypothesis that guided the overall research process.
Hypothesis: The gallery profile represents a two sided marketplace. Therefore, If we design the gallery profiles to incorporate quality and dynamic content from a gallery’s program, galleries would feel compelled to share more on artnet’s platform, and users increase their engagement with that content.
For example: If I see a George Condo artwork I think is interesting and click through to that gallery’s profile, I’ll see that Skarstedt Gallery (who represents Condo) also represents artists like KAWs, attended Art Basel Miami in the last 3 years and held 2 major exhibitions by the artist in New York. I see they have an upcoming exhibition I’d like to attend so I browse the photos and inquire on one of the artworks.
After synthesizing the research, sharing our report and getting key stakeholders updated and aligned, we identified three key challenges that we needed to address and were in the following areas:
- Discovery: The gallery profiles information architecture was very limited and the old versions consisted mostly of only an about page, a tab of their artists list, artwork inventory and exhibitions, limiting user journey’s to just a few pages of exploration.
- Content: It wasn’t designed for the richness of content types galleries were now uploading to the platform or that were generated elsewhere in the artnet ecosystem i.e. art fair participation or mentions on artnet news, both of which increasingly became validation and credible sources of information that were very important to the status of a gallery amongst their peers.
- Design language: The design language, UI elements and heavy use of white space were limiting how information was displayed on the page.
Designing the new gallery profiles
Armed with insights from our research and a good understanding of what we needed to accomplish, I started the design process by translating the research into user journey maps and created early sketches on paper. I also explored the information architecture and content structures, which were based on the data mapping and card sorting exercise results we conducted in our research and reflected what was most important to galleries:
- The artists they represent and works they have available
- Exhibitions they held and curated at their physical location
- Art fairs they attended
- Inventory they have available for sale
After some initial explorations I moved into Sketch to create initial concepts to share with key team members for feedback, gaining alignment regarding our direction and more importantly to have engineering on board early, since the project was meant to be the foundation of our new tech stack.
As mentioned in the beginning, the gallery profile project was not only going to be the first project to be built on top of the new front-end tech stack, but would also be the first to be designed and built around a new design language that would represent artnet’s first design system based on a universal component library.
The reasoning for this approach was equally design and engineering driven. Artnet’s architecture was aging and building on top of a fragile system, which introduced significant risk into any project, and it was becoming increasingly expensive in terms of engineering time since any new feature introduced a correlated amount of breakage and bugs. Introducing both a component library built on a new tech stack and a new design system were of equal importance. That way components created for things like exhibition and artwork pages could be reused throughout the site, saving future engineering time as well as providing users with a cohesive and seamless experience when interacting with any part of the artnet.
One of the first things I focused on was to create a richer profile experience that would instantly familiarize the user with a gallery’s program and brand to capture interest and invite to explore and browse additional content. The sections themselves were organized according to the results of our card sorting exercises, data mapping findings, as well as to what we identified as most engaging to our collector and art enthusiast audience (content and discovery). A bigger and bolder type hierarchy, dynamic rows of content and a new grid system would bring the various content types to the foreground and make sure the information presented would not overwhelm but rather engage the audience and encourage vertical as well as horizontal navigation paths.
In a second step I needed to evaluate the various tab experiences which previously presented only a limited amount of content and were not frequently visited. In light of the overall goal to increase engagement and discovery I decided to enhance them and add additional functionality such as creating a new filter system to let collectors browse all artworks more easily or adding biographical details pulled from the main artist profiles to a galleries artist section in order to add additional context, to name the most important.
Lastly I carefully designed the overall experience with the mobile journey in mind. Mobile usage increased significantly, especially since galleries would share their artnet profile not only via their own email newsletters with their community but more importantly via their social channels as well. Ensuring complete responsiveness and optimizations on a component level meant close collaboration with our front-end engineers and on occasion tweaking some of the react.js code myself.
Throughout the design process I made sure we received a continued validation from our users by sharing links to prototypes we made with Invision.This enabled us to make quick changes and validate our decisions around content structure and component designs. This method has its trade-offs of course, particularly in respect to testing more high-fidelity interactions, but for the initial validation phase it enabled us to move with speed and boost our confidence.
After we (engineering and design) aligned on the structure and component library we wanted to move forward with, we built out high fidelity HTML/CSS/React prototypes with test data that was representative of the ideal finished product and we Beta launched to a select group of galleries who were representative of our various customer sets. This method ensured we would being using engineering’s time as effectively and efficiently as possible, letting them focus on the most robust and risky aspects of the project - building the new tech stack.
Design validation
Post-launch it was extremely important to measure the impact of the new designs by monitoring key metrics as well as collecting user feedback. The KPI’s were the following:
- Pageviews (+- maintain, don’t damage current SEO)
- Pages/Session (+ increase)
- Time spent on page (+ increase)
- General Engagement with new metrics (sharing, article clicks)
I collaborated closely with the gallery liaisons to collect feedback from their clients post launch that was then collected, categorized and shared in our retrospective.