Refining Artfinder's Navigation: A Data-Driven IA Redesign

Refining Artfinder's Navigation: A Data-Driven IA Redesign

Refining Artfinder's Navigation: A Data-Driven IA Redesign

The Problem Statement

Artfinder's main navigation suffers from an unoptimized architecture, negatively impacting SEO rankings and hindering effective product discovery. This results in reduced organic traffic, lower conversion rates, and a diminished user experience.

Impact: This issue affects multiple stakeholders:

  • Customers: Struggle to find desired art, leading to frustration and lost sales.

  • Artists: Experience reduced visibility and sales opportunities.

  • Marketing & SEO: Face challenges with organic reach and increased reliance on paid advertising.

  • Sales: Encounter lower conversion rates, impacting revenue targets.

  • Content: Efforts are undermined by poor content categorization.

This leads to:

  • Reduced organic traffic and brand visibility.

  • Lower conversion rates and lost sales.

  • Increased customer acquisition costs.

  • Potential loss of market share to competitors.

Discovery Focus: Comprehensive research will be conducted to identify root causes, including:

  • SEO audits to analyze architecture and internal linking.

  • UX analysis to assess usability and user flow.

  • Website analytics review to identify user behavior patterns.

  • Keyword research to optimize navigation for search.

  • Competitor analysis to benchmark best practices.

  • Information architecture analysis to evaluate content categorization.

  • Mobile navigation analysis.

Overview

I conducted a comprehensive discovery to create a better IA for the main navigation.

Outcomes

  • Refined taxonomy

  • Granular subjects improve search accuracy

  • Consistent terminology improves user clarity

  • Data-driven hierarcy

Discovery

Stakeholder Insights

Artfinder's current navigation and categorization systems require significant refinement to improve user experience, SEO, and sales. Key issues include:

  • Inaccurate and Inconsistent Taxonomy: The core taxonomy lacks adherence to art world conventions, particularly in style and subject categorization. A missing "Thematic/Trend-Based Style" category and confusion between high and low-value limited editions further complicate matters.

  • Limited Subject Granularity: The subject categorization is broad and lacks detail, hindering effective search and discovery. A hierarchical structure with expanded options is proposed.

  • Style Terminology Errors: Inaccurate and inconsistent style terminology creates confusion. A refined, accurate style list is needed for better navigation and merchandising automation.

  • Limited Edition Differentiation: Mixing high and low-value limited editions dilutes the value of premium offerings. Solutions include adopting a model similar to Saatchi Art, with clear separation, or implementing stricter quality controls.

SEO Agency Insights

During an analysis of Artfinder's main navigation, an SEO agency identified key areas for improvement that directly impacted both search engine visibility and user experience. The existing navigation suffered from redundancy, with many links that could be better handled as filters. This clutter not only hindered SEO by weakening internal linking to important subcategories but also created a less efficient user journey. By streamlining the navigation and strategically expanding subcategory options, we aimed to improve both the site's search performance and the overall user experience, making it easier for customers to discover and navigate the diverse art offerings.

Google Analytics Data

Artfinder's main navigation, while organized by art form (paintings, sculptures, etc.), needed refinement to better reflect user needs and drive conversions. Initial analysis of page visit data suggested a specific order for these categories, which we considered for navigation prioritization. However, the stakeholders and SEO team were keen to add subcategories to the main navigation, prompting a deeper dive into user behavior.

Categories and Styles

To inform our subcategory strategy, I began by examining the most frequently visited subjects within the "Paintings" category. Initially, I assumed that high visit counts would correlate with purchase intent. However, I realized that visits alone didn't tell the whole story. To uncover true purchase drivers, I shifted my focus to the subjects with the highest purchase rates.

This simple shift revealed a critical discrepancy. The order of most visited subjects differed significantly from the order of most purchased subjects. This "Eureka!" moment highlighted a fundamental disconnect between browsing behavior and actual buying decisions. Users were exploring certain subjects out of curiosity, while ultimately purchasing others.

Intrigued by this finding, I applied the same analysis to all main navigation categories (sculptures, printmaking, photography, drawings, and digital art). The results were consistently revealing. Each category exhibited a similar divergence between visit and purchase patterns. Some categories even presented entirely different subject lists when comparing visit and purchase data.

This analysis fundamentally changed my approach to navigation design. I realised that relying solely on visit data would lead to a navigation that prioritized exploration over conversion. Instead, I needed to align the navigation with actual purchase behavior.

Hotjar Recordings

The primary function of Artfinder's main navigation is to facilitate efficient artwork discovery, translating to increased task completion rates and reduced time on task. While we lack precise quantitative data from Google Analytics for these metrics, a qualitative analysis of 150 Hotjar recordings (out of 1105) over the past three months revealed critical usability patterns and pain points.


Artfinder users exhibit a unique browsing behavior. Unlike typical e-commerce, they dedicate significant time to exploring artworks, often ranging from 6 to 106 minutes. This indicates a desire for emotional connection and discovery. While we acknowledge this inherent exploratory nature, our goal is to streamline the initial navigation process, enabling users to quickly access relevant content.


1. Misaligned Expectations: Subject vs. Art Form (16% of Observed Users)

  • Observation:

    • A significant portion of users (estimated 16%) who selected a subject from the main navigation (e.g., "Landscapes") expected to see only paintings on the subsequent page.

    • Hotjar recordings revealed users scrolling through mixed-medium landscape listings (paintings, photography, drawings, etc.) and then backtracking to apply a "Paintings" filter.

  • User Impact:

    • Users experience frustration due to a mismatch between their mental model and the website's behavior.

    • They are forced to take additional steps (applying filters) to achieve their intended goal, increasing time on task.

    • This leads to a perceived failure in the initial navigation attempt.

  • UX Implications:

    • Task Completion Rate: Reduced as users fail to directly access desired content.

    • Time on Task: Increased by approximately 65 seconds due to backtracking and filtering.

    • User Satisfaction: Diminished due to unexpected results and increased effort.

    • The subject navigation should be more specific, or the landing page should be filtered to the most popular art form.


2. Inefficient Medium Filtering: Need for Granular Refinement (24% of Observed Users)

  • Observation:

    • Users who selected a medium from the main navigation (e.g., "Oil Paintings") frequently added subject or style filters on the product listing page (PLP), either immediately or after a short browsing period (20 seconds to 2 minutes).

    • This affected approximately 24% of observed users.

  • User Impact:

    • Users demonstrate a clear understanding of their desired artwork (e.g., "landscape oil painting"), but the navigation requires multiple steps to reach their target.

    • This increases time on task and the likelihood of misclicks and errors.

  • UX Implications:

    • Users are forced to perform redundant filtering, indicating an opportunity for more granular navigation options.

    • The navigation should allow the user to select multiple filters at the same time, or the landing page should have more specific filters.


3. Pricing Sensitivity:

  • Observation:

    • A significant portion of users (10-15%) who select "Paintings" from the main navigation immediately apply price filters or sort by price on the PLP.

  • User Impact:

    • This highlights the importance of pricing considerations in the art purchasing process.

  • UX Implications:

    • Pricing filters and sorting options should be readily accessible and prominently displayed.

    • It may be useful to test adding price filters to the main navigation.


4. Terminology Discrepancies and "Discover All Art" Usage:

  • Observation:

    • Users clicking "Discover All Art" often proceed to apply category filters (e.g., "Paintings") on the Product Listing Page (PLP), suggesting a confusion or avoidance of the "Medium" terminology used in the main navigation.

    • A critical inconsistency exists between the use of "Medium" in the main navigation and "Category" within the PLP's filter options.

    • Example: Main navigation uses "Medium" (Painting, Photography, Sculpture), the PLP's filter uses "Category" (Painting, Photography, Sculpture).

  • Potential Interpretations:

    • Navigation Obscurity: The navigation elements might not be immediately clear or intuitive.

    • Terminology Familiarity: Users might be unfamiliar with the art-specific term "Medium" and instead rely on the more common e-commerce term "Category."

    • Information Architecture: The nomenclature of navigation elements is a key part of information architecture. The user's mental model for online shopping categories does not align with Artfinder's use of "Medium".

  • Competitive Analysis:

    • Saatchi Art uses "Medium" correctly, aligning with art world conventions.

    • Rise Art's use of "Medium" is ambiguous.

    • Singulart follows a similar pattern to Artfinder.

  • Art World Context:

    • "Medium" refers to the material used to create an artwork.

  • UX Implications:

    • Users are bypassing the intended navigation flow, indicating potential confusion or frustration.

    • Terminology inconsistencies create a disjointed user experience.

    • The "Discover all art" button is being used as a catch all, which is not ideal.

    • The wording should be consistent throughout the website

Taxonomy

A comprehensive review of category terminology revealed a critical distinction between 'Prints' and 'Printmaking.' While both terms exist within Artfinder's navigation, they represent fundamentally different art forms. 'Printmaking' encompasses traditional techniques using physical mediums, while 'Prints' often refers to digital reproductions. Combining these categories creates confusion, as they cater to distinct customer segments with varying price expectations and material preferences. Separating these categories will enhance clarity, improve navigation, and better serve the diverse needs of Artfinder's audience.


A review of Artfinder's art style taxonomy revealed inconsistencies and inaccuracies in terminology. For example, 'Impressionistic' should be 'Impressionist,' and 'Surrealistic' should be 'Surrealist.' As an online art marketplace, it is crucial to use correct art world terminology to maintain credibility and enhance user understanding. Ensuring accurate style labels will improve navigation, search functionality, and overall user experience, fostering a more professional and trustworthy platform.

Market and Competitive Analysis

Once I've gathered all data above I did a competitve and market analysis on main nav to get an idea on what kind of implementation we can have.

Develop

Initial attempts to define Artfinder's information architecture (IA), particularly a stakeholder-driven version, failed to accurately reflect the insights gathered during extensive user research, SEO analysis, and stakeholder feedback. This highlighted a significant disconnect between the existing structure and user needs.

To address this, I embarked on an iterative IA development process, creating multiple versions and refining them based on continuous analysis and feedback.


Initial IA: The stakeholder-driven version, while valuable, lacked the user-centric focus required for optimal navigation and discovery.

  • Data-Driven Iterations: Subsequent iterations were informed by:

    • User behavior analysis from Hotjar recordings.

    • SEO agency recommendations.

    • Stakeholder feedback.

    • Analysis of user intent (visits vs. purchases).

    • Analysis of terminology accuracy.

Version 5: The Refined IA: Through repeated iterations, I arrived at Version 5, which successfully integrated all data points and stakeholder requirements. This version prioritized:

  • Clear and consistent terminology.

  • Accurate representation of art world conventions.

  • Alignment with user mental models.

  • Optimized navigation for both browsing and purchasing.

  • Granular filtering and category separation.

Deliver

Building upon the data-driven insights and iterative development process, Version 5 of the Information Architecture (IA) was finalized. To ensure its effectiveness, I conducted guerilla usability tests with representative users.

  • Positive User Feedback: The tests yielded positive feedback, confirming that the refined IA significantly improved navigation clarity, findability, and overall user experience. Users were able to easily locate desired artworks and navigate the site with greater efficiency.

  • Design System Integration: Following the successful usability tests, I transitioned the IA into a comprehensive design system for seamless development. This involved:

    • Providing detailed component specifications.

    • Creating interactive prototypes to demonstrate user flows.

    • Defining all necessary states for interactive elements.

    • Documenting design tokens for consistent visual implementation.

    • Providing all required assets.

  • Development Handoff: The complete design system package was then handed off to the development team, ensuring a clear and consistent implementation of the new IA.