Cloud Costs Optimization

Exotel is a leading customer conversation platform enabling connected conversations between businesses and their customers, and leverages cloud significantly. Exotel is growing rapidly and has seen its cloud costs grow by up to 7X in the past 2 years (from 10s of thousands of $$ to 100s of thousands of $$ per month). Embedded in this growth have been different phases to cost management: 2020 the focus was to reduce costs, 2021 focus was top line/growth (speed and agility), come 2022/23 focus is now margins. Each of these phases meant different approaches to cloud costs. These varied phases resulted in our cloud costs as a % of revenue varying between 55% and 150% of the mean.

Through this talk, I intend to share the various views we have developed on cloud cost management with specific anecdotes. I will talk about:

1. Considerations behind identifying cost sweetspot for a given scenario
   a. Cost Vs X tradeoff
2. Cost implementation learnings
   a. Layered approach to cost optimization (Contracts, Operational, Architectural)
   b. How the approaches need to vary with size of the org and nature of the workload ( Elasticity of workload, Product/feature set volatility)
   c. Sustainability of optimisations (Creating self correcting systems)