Increasing Retention by Giving More Guidance
Role
Senior Product Designer — Datadog
Focus
Strategy, UX Research, Visual Design, Platform Design
People
Jamie Milstein - Product Manager
Joel Pfaff - Senior Software Engineer
Hamza Kadiri - Software Engineer
Julio Machado - Senior Software Engineer
Keith Kerr - Senior Product Designer
Ludovic Riffault - Senior Product Designer, Dataviz Team
Introduction
The Problem
Product Analytics was a new product suite at Datadog, and there was poor retention. In the private beta, customers were using some views, but there was no strong usage of the entire suite.
The Vision
From User Findings to Design Objectives
Product Analytics has quite a technical bend to it. Even our engineers could use a little more direction. It’s not very clear what to do after making a visualization.
— Product lead from Cvent (hospitality tech)
Product owners needed were lost when trying to build and read a Sankey Visualization. The configuration was confusing
Product owners often needed clear paths to key metrics like conversion rates
Product owners need a way to eaily share, save and export their findings
Simplifying the Visualization Builder Experience
Pain Points
1. The Sankey definition was confusing.
2. Users didn’t understand the “N+1” convention; they felt it was too technical.
3. Repeated view names caused confusion and complex visualizations.
Design Updates
1. A simplified Sankey definition was easier to comprehend with beta users.
2. Updated natural step titles like "1 step after" led to better understanding.
3. Smarter defaults for view depth and removing repeating steps led to simpler visualizations more often.
4. A context menu with guided next steps helped connect the platform.
Guiding Users with Contextual Next Steps
By Identifying where users tended to drop off, we were able to find the best place to provide guidance and connect the products of the Product Analytics suite.
When users want to learn more about a particular view, these menus give fast access to detailed info like Session Replays and Heatmaps. This helps users navigate the suite more easily and find more details about a view after making a Sankey Visualization.
Focusing on a Specific Path to View Conversion Rate
The Sankey Visualization is helpful for seeing popular routes people take, such as how they reached checkout or a particular product. However, once users identify a crtical the you choose a path to explore, you can create a funnel to track how many people complete specific steps, such as going from the homepage to checkout.
In under a week, we transformed a sketch of the “directed selection” interaction into a functional proof of concept. This rapid progress was facilitated by our collaborative efforts, leveraging a simple clickable prototype, an Excalidraw sketch, and a design document I created. The resulting interaction ensured users could only make valid selections between views with traffic flowing between them.
From a Sankey Vizualizaion, you can now easily select steps to see the conversion between two or more views.
Impact
The design changes resulted in a significant increase in beta user retention for the Product Analytics suite. Additionally, a high percentage of organizations continued using Product Analytics after the release of the Sankey Visualization builder.
2x Retention
We now have twice as many users within our beta cohort returning to use Product Analytics each month since we made these design changes.
92%
23/25 orgs from the private beta continued on to be customers or Product Analytics after the GA of the Sankey Visualization builder.