Online Concierge Generic Item Taxonomy for Order Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current online concierge systems require users to navigate through lengthy lists of items or provide multiple search queries to select specific items for orders, leading to increased time expenditure and decreased user interaction and order frequency, while also facing difficulties in importing generic item descriptions from third-party systems.
Innovation Solution
An online concierge system utilizes a taxonomy with different levels of specificity to associate generic item descriptions with specific items, applying machine learning models to predict user preferences and item availability, thereby simplifying the order creation process by displaying frequently purchased and available items prominently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If users navigate through long lists of items or provide multiple search queries to identify specific items, then users can find the items they need for orders, but users spend a considerable amount of time on the ordering process
Solution Approach 1:
The system performs preliminary actions by pre-processing item catalogs into hierarchical taxonomies and using machine learning models to predict user preferences before the user places an order. The system proactively generates personalized item recommendations and organizes items into logical hierarchies in advance, so when the user wants to order, the relevant items are already prepared and presented in an optimized manner, significantly reducing the time needed to find and select items.
2Adaptability or versatility
If the system displays all specific items under a generic item description, then users can select from complete item options, but users must navigate through lengthy lists to identify specific items
Solution Approach 1:
The system segments the item catalog into hierarchical taxonomies with multiple levels of granularity (e.g., category → subcategory → specific item). This segmentation allows users to navigate from general categories to specific items in a structured manner, presenting information in manageable portions rather than overwhelming lists. The hierarchical structure enables users to drill down only to the level of specificity they need, improving both navigation efficiency and item selection flexibility.
3Ease of manufacture
If third party systems use generic item descriptions like 'milk' or 'flour' in recipes, then recipes are simpler to create and maintain, but the system has difficulty identifying specific items to include in orders
Solution Approach 1:
The hierarchical taxonomy structure serves as an intermediary layer between generic item descriptions and specific items. When a third-party system provides a generic description like 'milk,' the system uses the taxonomy to expand this into specific available items (e.g., 'whole milk,' 'skim milk,' 'almond milk') while preserving the original generic intent. This intermediary taxonomy layer enables automatic resolution of generic descriptions to specific purchasable items without requiring changes to how third-party systems create recipes.
Data Source
AI summary
An online concierge system maintains a taxonomy associating one or more specific items offered by a warehouse with a generic item description. When the online concierge system receives a request to create an order from a user, the online concierge system selects a set of generic item descriptions from previously received orders and displays depictions of each generic item of the set to the user via an interface. In response to the user selecting a generic item description, the online concierge system identifies specific items associated with the selected generic item description from the taxonomy. Different identified specific items are displayed via the interface, for example as a scrollable list, allowing the user to select specific items for an order via the interface after selecting one or more generic item descriptions via the interface.