Democratizing machine learning experimentation and operationalization

​

See the full report here

​

Overview

This research project for Azure Machine Learning (AML) was done to identify strengths and weaknesses of its web-based product marketing, which serves as the entry point of the user acquisition funnel and is therefore a key interest area for user experience. In particular, it focused on insights from competitor marketing sites and points of comparison with our own site. Contrary to most of my empirical work on piecemeal parts of the AML experience past the paywall, this analysis broadly addresses the platform as a whole, and only concerns pre-paywall content.

​

Team

This analysis was a solo research project that was then shared with the AML marketing teams upon completion

​

Timeframe

1 week for defining the scope and gathering materials (screenshots) from sites, 1 week for in-depth annotations, 1 week for synthesis and recommendations

​

Limitations & Parameters

Because of time limits, the scope of the study was set to 5 key areas across 3 competitors and 2 influencers, and forms a good unit for a single study. Future studies may address further areas and competitors, or iterative work following changes made from this study's insights. 

​

Reflection

Focusing on a different area of the product, as mentioned above, using a different, desk-based research method, was a welcome change to my repertoire that allowed me to zoom out and see the bigger picture of user acquisitions that then lead to user experience with the actual product, which is my primary area of study, often through user interviews, usability testing, and surveys. Naturally, if there are issues at the beginning of the funnel, users are far less likely to make it further into the funnel; therefore, it's critical to study the experience for an uncommitted user browsing the site for the first time.

​

Sharing the findings with teams outside my normal purview was also a good experience that introduced me to other colleagues, and helped grow my profile in the AML network. The study is now being reviewed by various leadership team members and changes to the site are pending. I'm looking forward to future follow-up or expanded work in the area.

​

​