Contrary Research Rundown #98
The data duel between Databricks & Snowflake intensifies, plus new memos on Supabase, Thumbtack, and more
Contrary Research is growing our team, hiring a Research Analyst. Come help us build the best starting place to understand any private tech company. To apply, email research@contrary.com.
Research Rundown
In the rapidly evolving world of big data and cloud computing, two companies have emerged as rivals, vying for supremacy in the market of data analytics and AI — Snowflake and Databricks. We first covered this rivalry in a deep dive in a deep dive in September 2022. Since then, this rivalry has only intensified, with both companies pushing the boundaries of innovation and competing aggressively for market share.
Founded in the early 2010s, Snowflake and Databricks initially occupied separate niches in the data software landscape.
Snowflake was founded in 2012 by former Oracle architects Benoit Dageville and Thierry Cruanes. They set out to create a cloud-native data warehouse that could overcome the limitations of traditional on-premise solutions. Snowflake's platform has since evolved into a comprehensive "data cloud" enabling efficient data sharing, application development, and advanced analytics. It’s known for its user-friendly solutions, which make it particularly appealing to business analysts and data engineers.
Databricks, founded a year later in 2013 by a group of researchers from UC Berkeley, took a different path. Initially centered around Apache Spark, Databricks quickly expanded its platform to introduce the "data lakehouse" architecture — a hybrid of data lakes and warehouses. This platform is heavily favored by data scientists and machine learning engineers, offering advanced capabilities for data analysis and machine learning.
In the beginning, their relationship was even cooperative, with Databricks' marketing team using Snowflake for analytics.
However, as both companies expanded their offerings, they found themselves increasingly in direct competition. Snowflake's growth was meteoric, culminating in the largest-ever software IPO in 2021. Databricks, while still private, has become one of the world's highest-valued startups, with a highly anticipated IPO.
Databricks, once considered the "little brother," has taken on a more antagonistic stance with aggressive marketing and sales tactics. Their "SnowMelt" initiative aims to poach Snowflake's clients, especially in data warehousing. Databricks offers bonuses to salespeople who convert Snowflake clients and even helps pay off prospects' existing Snowflake contracts to encourage switching. When Snowflake held its 2023 user conference in Las Vegas, Databricks had billboards waiting at the airport and conference venue, claiming their product was nine times cheaper.
Databricks' pitch centers on being less expensive and offering a wider suite of advanced features, particularly for AI model building. They've equipped their sales team with tools to estimate potential savings for clients switching from Snowflake. Meanwhile, Snowflake maintains that its offering is more cost-effective. Christian Kleinerman, Snowflake's product chief, encourages customers to run their own benchmarks to compare costs.
The latest chapter in this fierce rivalry unfolded when Snowflake, looking to enhance its AI capabilities, entered exclusive negotiations to acquire Tabular for more than $600 million. Tabular was seen as a potentially meaningful asset to help Snowflake catch up in AI.
However, Databricks swooped in and ended up paying nearly $2 billion for Tabular. This price tag was a lofty valuation for a startup generating only about $1 million in annual recurring revenue.
As the battle continues, each company continues to enhance their offerings, all while facing the looming threat of Microsoft, which could leverage their vast resources to capture market share in this space.
The Snowflake / Databricks rivalry represents more than just a corporate competition — it is shaping the future of data analytics and cloud computing by driving innovation and expanding the capabilities of cloud data platforms. As both companies continue to blur the lines between data warehousing and data lake functionalities, the real winners may well be the customers who benefit from increasingly powerful and sophisticated data solutions.
Supabase's suite of backend services, including database management, authentication, file storage, and real-time updates, enables developers to build applications that are intended to be scalable and secure. To learn more, read our full memo here and check out some open roles below:
Site Reliability Engineer: Postgres - Remote (anywhere)
Support Engineer - Remote (EMEA timezones)
Thumbtack is an online platform that connects local service providers with users. To learn more, read our full memo here and check out some open roles below:
Senior Data Scientist, Product - Remote (US)
Director of Applied Science, Growth Initiatives - Remote (US)
Check out some standout roles from this week.
Lightyear | Remote - Strategic Finance Lead
Railway | Remote - Senior Infrastructure Engineer, Senior Full-Stack Engineer - Product, Product Designer
Wiz | Tel Aviv - Backend Engineer, DevOps Engineer, Frontend Engineer
Apple is forcing Patreon to switch to its in-app purchasing system by November 2025 or risk removal from the App Store.
Epic has opened up an API that allows patients to connect their electronic health records with apps outside of the Epic system, such as health coaches and exercise apps.
ARR can be a deceptive metric in early-stage investments. VCs are increasingly seeing AI startups jump from $0 to $2 million in ARR in just a few months, but often this revenue is concentrated in just one or two large consulting firm customers who are just testing the product.
ICONIQ Capital published The R&D Reporting Guide which provides key metrics and frameworks for technology companies to track and leverage best-in-class reporting.
Luma Dream Machine, a new AI video generator, has surged in traffic and is now ahead of many comparable web apps for AI video generation. Meanwhile, ChatGPT’s growth in worldwide traffic stalled in July. In the US, traffic continued to rise.
Tally, a debt consolidation startup, had raised $170 million but was forced to shut down, highlighting the tough challenges in the current market for consumer fintech companies to find sustainable business models.
Big tech companies like Google, Microsoft, and Amazon are avoiding outright acquisitions of AI startups, and instead are opting for more complicated deal structures that involve licensing the technology and hiring the top employees, effectively swallowing the startup without becoming the owner.
Meta and Google's existing apps already command significant user time, providing a captive audience for AI feature integration. The trifecta of engagement, retention and time spent, which Meta and Google already dominate, may provide the ingredients to win in consumer AI before ChatGPT.
Big Tech companies are trying to rewrite the rules on net zero emissions, suggesting they are actively shaping the narrative and policies around climate change and sustainability; not always for the better.
Apple is spending less than 10% of its operating cash flow on capital expenditures, while tech giants like Amazon, Meta, Microsoft, and Google are spending 40-60% on infrastructure buildout.
A hacking group leaked personal information, including Social Security Numbers, of potentially every American.
Forbes published its "Next Billion-Dollar Startups 2024" list, which features 25 venture-backed U.S. companies that Forbes and TrueBridge Capital Partners believe are most likely to reach a $1 billion valuation. Companies on the list include Hadrian, Linear, and Pinecone.
The U.S. Department of Justice is considering seeking a breakup of Google as one of the options after a court ruled that the company monopolized the online search market.
AI apps like Character.AI and Perplexity have very high weekly active user to monthly active user ratios (50-65%), showing how sticky and engaging these AI tools are for users.
Jessica Livingston interviewed the husband-wife founders of Replit, who had a vision they "couldn't let go of" despite working in multiple countries and employers.
Temporal, a durable execution platform for building and scaling invincible applications faster, surpassed 1,500 paying customers and 130 billion monthly actions on the platform and announced the appointment of Sahir Azam, Chief Product Officer of MongoDB, as its first independent board member, to help direct the company as it enters a new phase of hypergrowth.
One chief marketing officer found that Google's AI-powered advertising tools like Performance Max and Smart Bidding sometimes disrupted their advertising strategies or performed no better than human employees, leading them to stop using these tools.
A new paper, titled "The AI Technopanic and Its Effects”, argues that the recent public discourse around AI has been overly alarmist, and that a more measured approach to AI policy is required.
GenAI has shown strong code writing capability, which could help address the looming labor shortage and exponentially rising chip design costs.
California's controversial bill to regulate the AI industry, SB-1047, passed out of the Assembly Appropriations Committee. If it passes the full Senate by the end of the month, it will head to Gov. Gavin Newsom for his signature. The bill would require large AI models to have a "kill switch" that allows developers to stop them in case of danger, an interesting provision aimed at increasing safety and accountability.
At Contrary Research, our vision is to become the best starting place to understand private tech companies. We can't do it alone, nor would we want to. We focus on bringing together a variety of different perspectives.
That's why we're opening applications for our Research Fellowship. In the past, we've worked with software engineers, product managers, investors, and more. If you're interested in researching and writing about tech companies, apply here!