How to pick a Cloud Provider for your Startup?
Running a startup, scaling it and achieving viability in the business is a hard problem. There are many landmines along the way, where a suboptimal choice can lead to insurmountable difficulties later on. Choosing the right Cloud Provider for your business is one of those critical decisions, esp. for early-stage startups, where you have to weigh multiple technical and business factors before making a call.
Vishal Pathania (CTO, Evok Analytics), Sudheendra Chilappagari
(Co-Founder & Head of Product, Belong.io) and Arunkumar Jadhav (VP — Engineering, HackerRank) were the other distinguished panelists. The discussion was moderated by Janakiram MSV (Ex-Microsoft, Cloud Expert) The agenda for the discussion was making the right infrastructure decisions for a startup.
The panelists identified some key factors to consider before choosing your cloud provider —
- For startups or early stage ideas, price and ease of use are primary considerations.
- Nature of the Entry barrier — how difficult is it to take an MVP/prototype code from on-premise infra to a cloud infrastructure.
- Cost of Exit — Is exiting the provider going to take too much time or effort?
- Road to Adoption — What is the learning curve as you incrementally begin to make use of the services/features offered by the provider?
- Custom Features — What are the additional services offered by each provider and how do they align with your needs? (Db Back-up)
- Cost of Failure — What are the hidden costs associated with the provider when things go wrong? What happens when there is down-time, how good is the SLA?
There are some common pitfalls that startups need to avoid -—
- Do not pre-optimise for performance till necessary, build to exploit elasticity.
- Startups need to be Provider-Agnostic. Do not use services which we don’t really need.
In a free-wheeling discussion, the panelists also spoke about the need to strike a balance between experimenting with the latest tech for your stack and being conservative about mission critical elements of the business. The priorities of the business have to come first and you cannot experiment for experimentation’s sake. At Hasura, we started using Dockers and React, way ahead of the curve and had to deal with issues pertaining to versioning and insufficient community support. Another way to experiment would be to take an iterative approach, where you try out the latest tech in Hackathons and slowly transition your business to it after proof-of-value emerges.
New technology has to move the needle on some critical business requirements for adoption. If you feel compelled to experiment, the best way forward is to use dev tools or managed services providers that/who make life easy and offer support.
For more information on similar infrastructure questions, check out this and this. We strongly believe that remaining provider-agnostic and going with what’s optimal for your immediate business needs is the way to select your cloud infrastructure. We also love building products that leverage the strengths of latest frameworks and technologies from the open source communities. So we built Hasura, an infra-agnostic managed services platform. Do try it out and let us know what you think!