Cloud Cost Optimization Startups have become ubiquitous and they have found a friendly ear among corporate clients looking to cut costs during the recession. But should younger startups scrutinize their cloud spending in the same way?
According to several cloud investors, startups should prioritize building over optimization unless it’s going to save them a lot of money.
Boldstart Ventures partner Shomik Ghosh summed it up succinctly: “In early product or go-to-market stages, optimizing cloud spend should be the last thing a founder thinks about, besides using as many cloud resource credits as possible.”
We’re broadening our lens and looking for more investors to participate in londonbusinessblog.com surveys, where we survey top professionals about challenges in their industry.
If you are an investor and would like to participate in future surveys, please fill in this shape.
While founders should not worry about cloud costs in the early stages, they should still carefully consider other expansive decisions, such as cloud marketplaces, before hitting the road. An entrepreneur herself, angel investor Anshu Sharma noted that using cloud marketplaces as a distribution channel has pros and cons, and perhaps shouldn’t be done from day one, as “it can turn your offerings into commodities.”
Astasia Myers, the founder of Quiet Capital, agreed, saying that startups should focus on finding a product-market fit first. “We encourage startups to consider cloud marketplaces as soon as they find the product-market fit, not before,” she said.
“To successfully leverage cloud marketplaces, a solution’s product marketing, value proposition, and return on investment must be clear while exhibiting rapid time-to-value, which happens after PMF.”
However, because of the speed at which things are moving, startups may be exploring marketplaces sooner than they could: “Historically, we’ve seen startups join cloud marketplaces on Series D+. Now we’re starting to see companies considering it after Series B.”
Founders should also remember that startups are destined to get bigger and therefore need to plan ahead. “It is always important to select a technology stack that is available in all major cloud providers and that is as elastic as possible to support those migrations as they are needed (using Kubernetes is a good example to enable that)” , Liran Grinberg, co-founder and managing partner at Team8 said.
To find out what cloud-related advice investors are giving startups today, we spoke with:
- Shomik Ghoshpartner, Boldstart companies
- Liran Grinbergco-founder and managing partner, Team8
- Tim Tullypartner, Menlo Ventures
- Astasia Myersco-founder, Quiet capital
- Anshu Sharma, angel investor and co-founder/CEO, Airflow
Shomik Ghosh, Partner, Boldstart Ventures
Founders are looking for cost savings amid the recession. How important is it for startups to optimize their cloud spend in the early days?
It depends on what is meant by ‘early days’. In early product or go-to-market (GTM) stages, optimizing cloud spend should be the last thing on a founder’s mind, beyond using as much cloud resource credit as possible. Finding a product-market fit, engaging users, and understanding the end-user’s workflow and how the product is essential to these users is the key area founders should focus on.
As the company starts to get a few million ARR, it starts to make sense to better manage cloud spend to improve gross margins and therefore bottom line (net cash burn or free cash flow).
Major cloud providers often lure startups with free credit, but later also charge for data going out. How consistent are early-stage decisions in choosing a cloud provider now that cost optimization is becoming a bigger consideration than ever?
I think choosing an early-stage cloud provider based on cost is missing the wood for the trees. I know some founders who changed cloud providers in the beginning to keep using free credits. This may be possible if there are only a few people on the team, but as the team grows, everyone has to learn and relearn documentation, APIs and user interfaces, which has greater hidden “costs” than the money saved.
Cost optimization is not only about the amount of the bill at the end of the month. It is also the speed of the team’s product development, the avoidance of downtime, the developer’s experience of making teams work faster, etc. All these points should be central when choosing a cloud provider in the early stages.
What are the pros and cons of using a multicloud configuration rather than building on top of a single public cloud?
As a company scales, teams become more focused on functional areas. Before, everyone does everything, but as the team scales, you not only have a back-end infra team, but within it a database team, a security team, an ML team, a QA team, etc. Multicloud can help reap the benefits of the leverage the best tooling of any cloud provider.
In the early stages of a startup’s life, going from zero to one is paramount. Astasia Myers, Co-Founder of Quiet Capital
For example, Google BigQuery may be better than Redshift or Azure Synapse for some use cases, while AWS may have the best infrastructure management tooling. The trade-off, of course, is to make all those tools interoperable across platforms, and the big cloud providers aren’t exactly incentivized to do this.
This is where startups come in, and by focusing on making one product the best they can work across platforms and integrate easily (i.e. Snowflake can be used by any major cloud provider).
When should a startup consider going on location, if at all? Would you advise AI/ML startups differently?
In terms of terminology, I think on-prem should also be called ‘modern on-prem’, which means: replicated conceived, as it tackles not only bare-metal self-managed servers, but also virtual private clouds.
The most common reason startups should consider modern on-premises is for handling sensitive data, which is especially common in regulated industries (healthcare, financial services, or pharmaceuticals). However, the scope of what is considered sensitive grows over time with regulation, so it’s something more startups need to be aware of.
A lot of ML tooling has to be implemented in any environment, as the large enterprises keep some of this data in strictly controlled environments. Ultimately, startups need to meet the customer where they are – if you’re designing cloud-first and dealing with customers who have sensitive data, you need to consider what your “any environment” implementation strategy would be, whether you’re using Replicated, your own customers or choose not to work with those customers.
Have cloud costs reached a plateau relative to the marginal cost of computing or storage?
I think this is a difficult prediction for everyone. People say Moore’s law is coming to an end, but then another law pops up. I don’t think human ingenuity has stopped and companies continue to cut costs on their platform with ASICs [application-specific integrated circuits] or ML to optimize workloads. For example, Snowflake continues to lower prices; so it’s hard for me to say cloud costs have reached a plateau.
What do you think of cloud marketplaces as a distribution channel?
They are great! Its most obvious benefit is that it is bundled into a customer’s overall billing obligation to that cloud provider. It speeds up the procurement cycle, allows the customer to consolidate billing and allows them to better take advantage of the huge forward contract that they have likely committed to the cloud provider for many years.
If that contract is not fully used at the end of the term, the customer ultimately pays for services not provided.
How big is the market for cloud providers to offer additional services beyond their core offerings?
I’m not funny when I say infinite. For proof, visit AWS and check out the product catalog for all the different services listed. It would take years to fully understand everything it offers.
And if we extend the terminology of “cloud providers” beyond the compute and storage layer, virtually every public and private company providing a cloud service has multiple product offerings at scale.