Amazon Athena - Capacity reservations

Amazon Athena is an interactive query service that makes it easy to analyze data in Amazon S3 using standard SQL. Capacity reservations introduces a new billing model and feature that allow customers to reserve, assign, and manage dedicated query processing capacity for their workloads at any scale. Athena is ranked #24 of 200 most trafficked AWS service.

My team’s vision for this feature is to help our customers:

  • Improve performance in their workloads

  • Control cost with hourly pricing and have a more predictable monthly bill

Timeframe:

  • May 2022 - November 2022; Launched GA, April 2023

  • August 2023 - November 2023: Launched integration with Amazon CloudWatch metrics (Observability), November 2023

 

IMPACT

  • Crossed $1 million annual revenue run rate threshold within 6 months

  • Drove end-to-end user journeys for GA and future vision

  • Influenced engineering and product decisions

  • #1 key topic in customer meetings

How it works blog →

THE PROBLEM

  • Only have on-demand billing model

  • Charged based on amount of data scanned per query

  • Struggle to forecast monthly spending and predictable performance

 

THE SOLUTION

  • New pricing plan – billed at a flat, hourly rate

  • Performance-focused customers can ensure important workloads have capacity and can maintain a desired level of performance

  • Cost-focused customers can control cost with hourly pricing and have a more predictable monthly bill

 
 

MULTIPLE USER FLOWS BUILT

  • Create

  • Add Workgroups

  • Remove Workgroups

  • Add / Remove DPUs

  • Cancel

 

EXAMPLE SCREENS IN PRODUCTION

 
 

CHALLENGES

  • Learning technical constraints and negotiating with engineering team

  • Learning how to design for data visualization - it’s a different skillset

  • Using metaphors on how we designed our experience to get stakeholder buy-in

  • Working around technical limitations and exploring multiple solutions

 

IF I HAD MORE TIME…

  • Recommendation system using AI to analyze workloads

 

WHAT I LEARNED

  • Continue to question and learn more about why something may not work - especially on a technical level

  • Continue to push engineering to raise the bar on user experience

  • How to design and critique for data visualization graphs