Snowflake shows new risks to consumption-based pricing for cloud computing companies

  • The pay-as-you-go model of cloud computing has become an unexpectedly risky proposition.
  • Some of the largest cloud data providers, such as Snowflake, use consumption-based pricing.
  • Experts say the model has not been tested in a downturn and may be risky for companies that rely on it.

One of Snowflake’s biggest appeals is the pricing model: You pay for the computing you use. This simplicity helped propel the data storage giant — and a class of other cloud-based companies — to massive success.

Now, with the market crash, this model has suddenly become a huge liability. Companies trying to cut costs seek to restore their cloud operations by slowing down their own businesses or increasing demand for efficiency internally. This results in fewer uses – and fewer dollars for cloud providers.

Snowflake’s shares fell after it reported earnings on Wednesday. The economic downturn has already wiped out more than $70 billion of market value from Habibie, which held one of the largest initial public offerings of software in 2020. On its earnings call on Wednesday, eight of the 20 questions asked somehow more information about pricing based on Consumption.

While Snowflake said it has seen a spike in consumption in recent weeks, it also indicated that some of its largest customers have pulled out, raising fears that a very successful model would put the entire industry under additional pressure in the process.


Recession

.

“Everyone likes a recovery model of consumption,” said Derek Wood, managing director at investment bank Coin & Company. “I don’t think we’ve been tested as to what it looks like in a tightening environment. In an aggravating environment, contracts, where you have assessable revenue contracted into profit and loss, looks pretty good.”

Cloud players say consumption-based pricing with companies like Snowflake is an unproven model in a downturn

Instead of signing huge contracts with set usage amounts — typical of old-school cloud vendors — Snowflake customers have the flexibility to pay for as much computing as possible. Their spending can go up and down as per the demand of their business.

This has been a big plea for many modern cloud companies. Mission-critical coordination tools such as portfolios rely on consumption-based pricing. So do other modern data tools, such as dbt, which have been widely adopted by companies to ease the data-crunching process. MongoDB, another large public trading provider, also relies on consumption-based pricing.

Now, as companies look to cut costs, some may cut back on these data operations. This can manifest itself in increasing the efficiency of existing processes, such as daily updates through tools such as dbt, or cutting these processes completely. Every incremental efficiency improvement or query dropped is a dollar for a company looking to tighten its spending budget.

“We may assert that databases are less prone to stagnation than other consumer companies, particularly in the field of data analytics, where particular databases are very important for software development (i.e. an application cannot be built without a database), while data warehouses, data science platforms and Even monitoring solutions support all discretionary analytical projects,” RBC Capital Markets wrote in a note this week.

Given that consumption-based pricing is a recent development, there isn’t a lot of data on how it will perform. There will be a steep learning curve for all gamers, said Emil Ephrem, CEO of Neo4j, another database provider with a consumption-based model.

“We don’t have answers in the downturn,” he said. “We’ll all be learning a lot over the next several quarters, and I think we’ll all get out of it smarter. Amazon Web Services was around in 2008, so they might have some data on it, but that’s it — nobody else has anything.”

Snowflake offers an early glimpse into the responsibility of consumption-based pricing

While Snowflake has shown that its pricing model has been exposed to the forces of economic downturn, it also provides a glimpse into the pricing model across the industry. On the earnings call, Wall Street analysts pressed executives for more information about the impact of spending cuts to gauge the sensitivity of the model.

Snowflake defended its data warehousing model — even though it took a heavy blow on its main competitor, Databriks, which focuses heavily on machine learning and data science, saying that kind of spending could be considered discretionary.

CEO Frank Slotman said in the earnings report that the types of workloads for the company’s core data warehouses “are not going anywhere.”

“It’s not optional,” he said on the earnings call. “They didn’t like, what am I feeling today? That, by the way, there are workloads like that, and that’s a lot more on the data lake side, where you basically have a massive repository of files. You might have the data of scientists who are just kind of fishing files out The lake and they try and decide to do some interesting things for it. That kind of thing is very appreciative, but that’s not the focus of our work.”

However, Snowflake is still investing heavily in machine learning and data science by launching new features to support the Python language and spending $800 million on a small machine learning startup called Streamlit. It’s an area where Databriks has traditionally been stronger, while Snowflake has thrived thanks to data warehousing and analytics.

“Our data lake architecture covers the entire journey, from raw data to predictive modeling, meta-analytics, and automated decision making,” Databriks CEO Ali Qudsi told Insider. “We see some of the most valuable data use cases in the latter. These projects aren’t discretionary, but are very strategic for organizations. So we’re seeing continued growth in our business thanks to that.”

However, there may be room for optimism about Snowflake, as it may offset lower spending among some key customers through increased use by other customers, Wood said.

“I wouldn’t condemn the model,” said Brad Zelnick, managing director of equity research in software for Deutsche Bank. “I would say that the most innovative and future-oriented companies are the ones using consumption models, whether it’s AWS, MongoDB, or Snowflake.

“The nature of the model is that if at some point a customer decides to adjust their consumption in some way, there will be a real-time impact on the income statement.”

Leave a Comment