HN Explorer

Can companies like SaaS companies like snowflake competele with cloud vendor?

1 points by lex2ross ↗ HN
SaaS companies like Snowflake, Databricks, Confluent, and Elastic offer high-quality solutions that work well. However, these services are hosted on different cloud vendors, which can be problematic. Cloud vendors have their own alternative solutions, such as Redshift, BigQuery, PubSub, Kinesis, and OpenSearch, which can make it difficult for SaaS companies to compete.

Typically, SaaS companies keep an 80% margin, while cloud vendors keep another 60-70% margin (approximately 65%). So, if you spend $100, $80 goes to the SaaS company, and $13 goes to the cloud vendor, which means only $7 is spent on actual hardware.

My hypothesis is that if a cloud vendor were to develop a similar service, such as AWS OpenSearch (an Elastic fork), and offer it at an 80% margin, the cloud vendor solution would cost only $7/.2 = $35, which is one-third the cost of the SaaS service. Additionally, the cloud vendor solution would be better integrated with other vendor services, reducing infrastructure management toil.

This creates a challenging situation for SaaS companies. To compete with cloud vendors, SaaS solutions would need to be at least 2x better, which is a daunting task given that both SaaS and cloud vendors often use different flavors of open source.

My questions for the community members are:

1. Do you agree with my hypothesis? 2. If my hypothesis is correct, do you think SaaS companies will be able to survive the tough competition from cloud vendors?

4 comments

[ 2.9 ms ] story [ 16.9 ms ] thread
atlgator
There are examples out there. DataStax is a managed Cassandra provider and AWS came out with their own managed Cassandra service.
lex2ross
Basically my question is whether in long run these companies will survive once the cloud start providing similar much cheaper alternatives.
atlgator
I understand. I’m suggesting DataStax is one to watch for this reason.
hodgesrm
I don't agree. Here are four reasons why public clouds don't automatically win.

1. Many companies operate multi-cloud and prefer vendors who cross platforms.

2. SaaS companies don't compete with cloud vendors in a simple, head to head way. Instead they operate symbiotically. A SaaS on for analytics AWS might compete with Redshift but have joint go-to-market with EC2 with deep discounting of cloud costs as the vendor grows. AWS makes money either way.

3. Your margin argument is wrong. Gross margins for independent SaaS vendors that are infra heavy are more like 50-60% all in (including payments to AWS, GCP, etc.) You can bring them up by managing the infrastructure running in your customers' cloud accounts.

You need to instead look at actual prices for competing SaaS services. They tend to be roughly comparable in the aggregate at least in the database markets I'm familiar with, because vendors charge what the market will bear. Instead, vendors compete on support, the billing model, and features. Public clouds don't have any particular advantage here. Just to give one example: Redshift offers deep discounts in return for 3 year prepayment. However, it can't handle the response time required by many customer-facing APIs.

4. The thing that screws up vendor finances is not cost of delivery (COGS) but customer acquisition costs (CAC). That's the principal reason why many of the big SaaS services are still underwater on profitability.

My company is a SaaS vendor for ClickHouse. We've experienced all of these. Our business is doing fine.