Webdam | Machine Learning
Webdam provides a Digital Asset Management platform to help marketers and creatives manage all of their digital media online in one easy-to-access place. Press release.
Task
Because Webdam customers often struggle with adding metadata to their content, it is difficult for other users to find the content they need. The task was to identify ways to make content more searchable for the end users.
Action
Spoke with users that search to learn more about their search behavior. Also spoke with users that are responsible for adding metadata to understand what works well and what they struggle with. Identified an ML technology that would provide keywords via an API based on content’s pixel data alone (provided by Shutterstock). Worked with design to prototype the solution integrated into our platform and presented multiple options to users to identify which was the most intuitive.
Execution
Updated Webdam’s user interface to provide suggested keywords via Shutterstock’s API with an option to easily accept all keywords at once or individually (~4 weeks).
Results
Increased amount of keywords added to content by 20% month over month.
Increased number of successful searches and downloads of content by 10% month over month.
Peterson's | eCommerce
Peterson’s is a college search, scholarship search, and test prep tool. For test prep, Peterson’s offers practice tests in over 160 subjects
Task
Identify areas or opportunities to increase revenue from test prep sales.
Action
Conducted user research on usertesting.com and by conducting interviews with 50 students interested in test prep materials. Based on both qualitative and quantitative research, I identified the following areas of opportunity:
1. Adding a shopping cart with recommended add ons at checkout. 2. Adding search functionality with auto-complete. 3. Making usability improvements to the navigation.
Execution
To address the above findings, I: 1. Added shopping cart with recommendations (~2 weeks). 2. Worked with an SEO firm for evaluation and to implement various page structure and name recommendations for test prep pages (~2 weeks). 3. Worked with marketing to produce the unique content needed according to SEO recommendations (~3 weeks). 4. Worked with external UX/UI designer to redesign pages with simplicity in mind (~1 week). 5. Work with external Wordpress developer to implement the new designs (~2 weeks)
Results
Increased average number of products per order from 1 to 1.3 per transaction. Increased test prep sales by 800% in four months. Decreased bounce rate significantly on test prep pages by 20%. Traffic to test prep pages increased exponentially.
Blackray | AI for Product Managers
As a startup, Blackray was focused on solving the toughest challenges for Product Managers as it relates to reviewing and synthesizing customer and prospect feedback and delivering a strong strategy for their development efforts.
Task
Product Managers spend hours sifting through information on customers and prospects through disparate sources. This is time consuming and prone to error. The goal was to identify ways to save Product Managers times and provide key insights that they can use to inform their strategies and roadmaps. As a new business, our goal was to secure users and gather feedback.
Action
Spoke with Product Managers across a variety of industries to learn more about the various tools they use to gather customer and prospect insights. Additionally, explored the process Product Managers go through to organize and leverage insights in a way that supports their decisions.
Identified opportunities to integrate tools such as Salesforce and implement LLMs to synthesize large amounts of information into useful points quickly.
Execution
Integrated with Salesforce, Google Drive, and Atlassian Confluence to easily surface information in one place (~4 weeks). Leveraged Open AI and Llama to process text surfaced through integrations and deliver the following key insights: Feature Requests, Pain Points, Positive Mentions, and Competitor Mentions. (~4 weeks).
Results
Increased user base from 0 to 300 Product Managers in under 2 months. Built a community of Product Managers that provided useful feedback to help us build even more powerful tools to help them conduct their research.