Personal blog of Yzmir Ramirez

Statsig is taking the A/B testing applications that drive Facebook’s growth and putting similar functionalities into the hands of any product team so that they, too, can make faster, data-informed decisions on building products customers want.

The Seattle-based company on Thursday announced $10.4 million in Series A funding, led by Sequoia Capital, with participation from Madrona Venture Group and a group of individual investors, including Robinhood CPO Aparna Chennapragada, Segment co-founder Calvin French-Owen, Figma CEO Dylan Field, Instacart CEO Fidji Simo, DoorDash exec Gokul Rajaram, Code.org CEO Hadi Partovi and a16z general partner Sriram Krishnan.

Co-founder and CEO Vijaye Raji started the company with seven other former Facebook colleagues in February, but the idea for the company started more than a year ago.

He told TechCrunch that while working at Facebook, A/B testing applications, like Gatekeeper, Quick Experiments and Deltoid, were successfully built internally. The Statsig team saw an opportunity to rebuild these features from scratch outside of Facebook so that other companies that have products to build — but no time to build their own quick testing capabilities — can be just as successful.

Statsig’s platform enables product developers to run quick product experiments and analyze how users respond to new features and functionalities. Tools like Pulse, Experiments+ and AutoTune allow for hundreds of experiments every week, while business metrics guide product teams to build and ship the right products to their customers.

Raji intends to use the new funding to hire folks in the area of design, product, data science, sales and marketing. The team is already up to 14 since February.

“We already have a set of customers asking for features, and that is a good problem, but now we want to scale and build them out,” he added.

Statsig has no subscription or upfront fees and is already serving millions of end-users every month for customers like Clutter, Common Room and Take App. The company will always offer a free tier so customers can try out features, but also offers a Pro tier for 5 cents per thousand events so that when the customer grows, so does Statsig.

Raji sees adoption of Statsig coming from a few different places: developers and engineers that are downloading it and using it to serve a few million people a month, and then through referrals. In fact, the adoption the company is getting is “bottom up,” which is what Statsig wants, he said. Now the company is talking to bigger customers.

There are plenty of competitors for this product, including incumbents in the market, according to Raji, but they mostly focus on features, while Statsig provides insights and ties metrics back to features. In addition, the company has automated analysis where other products require manual set up and analysis.

Sequoia partner Mike Vernal worked at Facebook prior to joining the venture capital firm and had worked with Raji, calling him “a top 1% engineer” that he was happy to work with.

Having sat on many company boards, he has found that many companies spend a long time talking about sales and marketing, but very little on product because there is not an easy way to get precise numbers for planning purposes, just a discussion about what they did and plan to do.

What Vernal said he likes about Statsig is that the company is bringing that measurement aspect to the table so that companies don’t have to hack together a poorer version.

“What Statsig can do, uniquely, is not only set up an experiment and tell if someone likes green or blue buttons, but to answer questions like what the impact this is of the experiment on new user growth, retention and monitorization,” he added. “That they can also answer holistic questions and understand the impact on any single feature on every metric is really novel and not possible before the maturation of the data stack.”

As companies increasingly rely on data to run their businesses, having accurate sources of data becomes paramount. Stemma, a new early-stage startup, has come up with a solution, a managed data catalogue that acts as an organization’s source of truth.

Today the company announced a $4.8 million seed investment led by Sequoia with assorted individual tech luminaries also participating. The product is also available for the first time today.

Company co-founder and CEO Mark Grover says the product is actually built on top of the open-source Amundsen data catalogue project that he helped launch at Lyft to manage its massive data requirements. The problem was that with so much data, employees had to kludge together systems to confirm the data validity. Ultimately manual processes like asking someone in Slack or even creating a Wiki failed under the weight of trying to keep up with the volume and velocity.

“I saw this problem firsthand at Lyft, which led me to create the open-source Amundsen project with a team of talented engineers,” Grover said. That project has 750 users at Lyft using it every week. Since it was open-sourced, 35 companies like Brex, Snap and Asana have been using it.

What Stemma offers is a managed version of Amundsen that adds functionality like using intelligence to show data that’s meaningful to the person who is searching in the catalogue. It also can add metadata automatically to data as it’s added to the catalogue, creating documentation about the data on the fly, among other features.

The company launched last fall when Grover and co-founder and CTO Dorian Johnson decided to join forces and create a commercial product on top of Amundsen. Grover points out that Lyft was supportive of the move.

Today the company has five employees, in addition to the founders, and has plans to add several more this year. As he does that, he is cognizant of diversity and inclusion in the hiring process. “I think it’s super important that we continue to invest in diversity, and the two ways that I think are the most meaningful for us right now is to have early employees that are from diverse groups, and that is the case within the first five,” he said. Beyond that, he says that as the company grows he wants to improve the ratio, while also looking at diversity in investors, board members and executives.

The company, which launched during COVID, is entirely remote right now and plans to remain that way for at least the short term. As the company grows, they will look at ways to build camaraderie, like organizing a regular cadence of employee offsite events.

As companies create machine learning models, the operations team needs to ensure the data used for the model is of sufficient quality, a process that can be time consuming. Bigeye (formerly Toro), an early stage startup is helping by automating data quality.

Today the company announced a $17 million Series A led Sequoia Capital with participation from existing investor Costanoa Ventures. That brings the total raised to $21 million with the $4 million seed, the startup raised last May.

When we spoke to Bigeye CEO and co-founder Kyle Kirwan last May, he said the seed round was going to be focussed on hiring a team — they are 11 now — and building more automation into the product, and he says they have achieved that goal.

“The product can now automatically tell users what data quality metrics they should collect from their data, so they can point us at a table in Snowflake or Amazon Redshift or whatever and we can analyze that table and recommend the metrics that they should collect from it to monitor the data quality — and we also automated the alerting,” Kirwan explained.

He says that the company is focusing on data operations issues when it comes to inputs to the model such as the table isn’t updating when it’s supposed to, it’s missing rows or there are duplicate entries. They can automate alerts to those kinds of issues and speed up the process of getting model data ready for training and production.

Bogomil Balkansky, the partner at Sequoia who is leading today’s investment sees the company attacking an important part of the machine learning pipeline. “Having spearheaded the data quality team at Uber, Kyle and Egor have a clear vision to provide always-on insight into the quality of data to all businesses,” Balkansky said in a statement.

As the founding team begins building the company, Kirwan says that building a diverse team is a key goal for them and something they are keenly aware of.

“It’s easy to hire a lot of other people that fit a certain mold, and we want to be really careful that we’re doing the extra work to [understand that just because] it’s easy to source people within our network, we need to push and make sure that we’re hiring a team that has different backgrounds and different viewpoints and different types of people on it because that’s how we’re going to build the strongest team,” he said.

Bigeye offers on prem and SaaS solutions, and while it’s working with paying customers like Instacart, Crux Informatics, and Lambda School, the product won’t be generally available until later in the year.

Aquarium, a startup from two former Cruise employees, wants to help companies refine their machine learning model data more easily and move the models into production faster. Today the company announced a $2.6 million seed led by Sequoia with participation from Y Combinator and a bunch of angel investors including Cruise co-founders Kyle Vogt and Dan Kan.

When the two co-founders CEO Peter Gao and head of engineering Quinn Johnson, were at Cruise they learned that finding areas of weakness in the model data was often the problem that prevented it from getting into production. Aquarium aims to solve this issue.

“Aquarium is a machine learning data management system that helps people improve model performance by improving the data that it’s trained on, which is usually the most important part of making the model work in production,” Gao told me.

He says that they are seeing a lot of different models being built across a variety of industries, but teams are getting stuck because iterating on the data set and continually finding relevant data is a hard problem to solve. That’s why Aquarium’s founders decided to focus on this.

“It turns out that most of the improvement to your model, and most of the work that it takes to get it into production is about deciding, ‘Here’s what I need to go and collect next. Here’s what I need to go label. Here’s what I need to go and retrain my model on and analyze it for errors and repeat that iteration cycle,” Gao explained.

The idea is to get a model into production that outperforms humans. One customer Sterblue offers a good example. They provide drone inspection services for wind turbines. Their customers used to send out humans to inspect the turbines for damage, but with a set of drone data, they were able to train a machine learning model to find issues. Using Aquarium, they refined their model and improved accuracy by 13%, while cutting the cost of human reviews in half, Gao said.

The Aquarium team. Image: Aquarium

Aquarium currently has 7 employees including the founders, of which three are women. Gao says that they are being diverse by design. He understands the issues of bias inherent in machine learning model creation, and creating a diverse team for this kind of tooling is one way to help mitigate that bias.

The company launched last February and spent part of the year participating in the Y Combinator Summer 2020 cohort. They worked on refining the product throughout 2020, and recently opened it up from beta to generally available.

Census, a startup that helps businesses sync their customer data from their data warehouses to their various business tools like Salesforce and Marketo, today announced that it has raised a $16 million Series A round led by Sequoia Capital. Other participants in this round include Andreessen Horowitz, which led the company’s $4.3 million seed round last year, as well as several notable angles, including Figma CEO Dylan Field, GitHub CTO Jason Warner, Notion COO Akshay Kothari and Rippling CEO Parker Conrad.

The company is part of a new crop of startups that are building on top of data warehouses. The general idea behind Census is to help businesses operationalize the data in their data warehouses, which was traditionally only used for analytics and reporting use cases. But as businesses realized that all the data they needed was already available in their data warehouses and that they could use that as a single source of truth without having to build additional integrations, an ecosystem of companies that operationalize this data started to form.

The company argues that the modern data stack, with data warehouses like Amazon Redshift, Google BigQuery and Snowflake at its core, offers all of the tools a business needs to extract and transform data (like Fivetran, dbt) and then visualize it (think Looker).

Tools like Census then essentially function as a new layer that sits between the data warehouse and the business tools that can help companies extract value from this data. With that, users can easily sync their product data into a marketing tool like Marketo or a CRM service like Salesforce, for example.

Image Credits: Census

“Three years ago, we were the first to ask, ‘Why are we relying on a clumsy tangle of wires connecting every app when everything we need is already in the warehouse? What if you could leverage your data team to drive operations?’ When the data warehouse is connected to the rest of the business, the possibilities are limitless,” Census explains in today’s announcement. “When we launched, our focus was enabling product-led companies like Figma, Canva, and Notion to drive better marketing, sales, and customer success. Along the way, our customers have pulled Census into more and more scenarios, like auto-prioritizing support tickets in Zendesk, automating invoices in Netsuite, or even integrating with HR systems.“

Census already integrates with dozens of different services and data tools and its customers include the likes of Clearbit, Figma, Fivetran, LogDNA, Loom and Notion.

Looking ahead, Census plans to use the new funding to launch new features like deeper data validation and a visual query experience. In addition, it also plans to launch code-based orchestration to make Census workflows versionable and make it easier to integrate them into an enterprise orchestration system.

Small enterprises remain one of the most underserved segments of the business market, but the growth of cloud-based services — easier to buy, easier to provision — has helped that change in recent years. Today, one of the more promising startups out of Europe building software to help SMEs run online businesses is announcing some funding to better tap into both the opportunity to build these services, and to meet a growing demand from the SME segment.

Xentral, a German startup that develops enterprise resource planning software covering a variety of back-office functions for the average online small business, has picked up a Series A of $20 million.

The company’s platform today covers services like order and warehouse management, packaging, fulfillment, accounting and sales management, and the majority of its 1,000 customers are in Germany — they include the likes of direct-to-consumer brands like YFood, KoRo, the Nu Company and Flyeralarm.

But Benedikt Sauter, the co-founder and CEO of Xentral, said the ambition is to expand into the rest of Europe, and eventually other geographies, and to fold in more services to its ERP platform, such as a more powerful API to allow customers to integrate more services — for example in cases where a business might be selling on their own site, but also Amazon, eBay, social platforms and more — to bring their businesses to a wider market.

Mainly, he said, the startup wants “to build a better ecosystem to help our customers run their own businesses better.”

The funding is being led by Sequoia Capital, with Visionaires Club (a B2B-focused VC out of Berlin) also participating.

The deal is notable for being the prolific, high-profile VC’s first investment in Europe since officially opening for business in the region. (Sequoia has backed a number of startups in Europe before this, including Graphcore, Klarna, Tessian, Unity, UiPath, n8n and Evervault — but all of those deals were done from afar.)

Augsburg-based Xentral has been around as a startup since 2018, and “as a startup” is the operative phrase here.

Sauter and his co-founder Claudia Sauter (who is also his co-founder in life: she is his wife) built the early prototype for the service originally for themselves.

The pair were running a business of their own — a hardware company they founded in 2008, selling not nails, hammers and wood, but circuit boards they designed, along with other hardware to build computers and other connected objects. Around 2013, as the business was starting to pick up steam, they decided that they really needed better tools to manage everything at the backend so that they would have more time to build their actual products.

But Bene Sauter quickly discovered a problem in the process: smaller businesses may have Shopify and its various competitors to help manage e-commerce at the front end, but when it came to the many parts of the process at the backend, there really wasn’t a single, easy solution (remember this was eight years ago, at a time before the Shopifys of the world were yet to expand into these kinds of tools). Being of a DIY and technical persuasion — Sauter had studied hardware engineering at university — he decided that he’d try to build the tools that he wanted to use.

The Sauters used those tools for years, until without much outbound effort, they started to get some inbound interest from other online businesses to use the software, too. That led to the Sauters balancing both their own hardware business and selling the software on the side, until around 2017/2018 when they decided to wind down the hardware operation and focus on the software full time. And from then, Xentral was born. It now has, in addition to 1,000 customers, some 65 employees working on developing the platform.

The focus with Xentral is to have a platform that is easy to implement and use, regardless of what kind of SME you might be as long as you are selling online. But even so, Sauter pointed out that the other common thread is that you need at least one person at the business who champions and understands the value of ERP. “It’s really a mindset,” he said.

The challenge with Xentral in that regard will be to see how and if they can bring more businesses to the table and tap into the kinds of tools that it provides, at the same time that a number of other players also eye up the same market. (Others in the same general category of building ERP for small businesses include online payments provider Sage, NetSuite and Acumatica.) ERP overall is forecast to become a $49.5 billion market by 2025.

Sequoia and its new partner in Europe, Luciana Lixandru — who is joining Xentral’s board along with Visionaries’ Robert Lacher — believe however that there remains a golden opportunity to build a new kind of provider from the ground up and out of Europe specifically to target the opportunity in that region.

“I see Xentral becoming the de facto platform for any SMEs to run their businesses online,” she said in an interview. “ERP sounds a bit scary especially because it makes one think of companies like SAP, long implementation cycles, and so on. But here it’s the opposite.” She describes Xentral as “very lean and easy to use because you an start with one module and then add more. For SMEs it has to be super simple. I see this becoming like the Shopify for ERP.”

Tecton.ai, the startup founded by three former Uber engineers who wanted to bring the machine learning feature store idea to the masses, announced a $35 million Series B today, just seven months after announcing their $20 million Series A.

When we spoke to the company in April, it was working with early customers in a beta version of the product, but today, in addition to the funding, they are also announcing the general availability of the platform.

As with their Series A, this round has Andreessen Horowitz and Sequoia Capital co-leading the investment. The company has now raised $60 million.

The reason these two firms are so committed to Tecton is the specific problem around machine learning the company is trying to solve. “We help organizations put machine learning into production. That’s the whole goal of our company, helping someone build an operational machine learning application, meaning an application that’s powering their fraud system or something real for them […] and making it easy for them to build and deploy and maintain,” company CEO and co-founder Mike Del Balso explained.

They do this by providing the concept of a feature store, an idea they came up with and which is becoming a machine learning category unto itself. Just last week, AWS announced the Sagemaker Feature store, which the company saw as major validation of their idea.

As Tecton defines it, a feature store is an end-to-end machine learning management system that includes the pipelines to transform the data into what are called feature values, then it stores and manages all of that feature data and finally it serves a consistent set of data.

Del Balso says this works hand-in-hand with the other layers of a machine learning stack. “When you build a machine learning application, you use a machine learning stack that could include a model training system, maybe a model serving system or an MLOps kind of layer that does all the model management, and then you have a feature management layer, a feature store which is us — and so we’re an end-to-end life cycle for the data pipelines,” he said.

With so much money behind the company it is growing fast, going from 17 employees to 26 since we spoke in April, with plans to more than double that number by the end of next year. Del Balso says he and his co-founders are committed to building a diverse and inclusive company, but he acknowledges it’s not easy to do.

“It’s actually something that we have a primary recruiting initiative on. It’s very hard, and it takes a lot of effort, it’s not something that you can just make like a second priority and not take it seriously,” he said. To that end, the company has sponsored and attended diversity hiring conferences and has focused its recruiting efforts on finding a diverse set of candidates, he said.

Unlike a lot of startups we’ve spoken to, Del Balso wants to return to an office setup as soon as it is feasible to do so, seeing it as a way to build more personal connections between employees.

Fishtown Analytics, the Philadelphia-based company behind the dbt open-source data engineering tool, today announced that it has raised a $29.5 million Series B round led by Sequoia Capital, with participation from previous investors Andreessen Horowitz and Amplify Partners.

The company is building a platform that allows data analysts to more easily create and disseminate organizational knowledge. Its focus is on data modeling, with its dbt tool allowing anybody who knows SQL to build data transformation workflows. Dbt also features support for automatically testing data quality and documenting changes, but maybe most importantly it uses standard software engineering techniques to help engineers collaborate on code and integrate changes continuously.

If this all sounds a bit familiar, it’s probably because you saw that Fishtown Analytics also announced a $12.9 million Series A round in April. It’s not often we see both a Series A and B round within half a year, but that goes to show how the market for Fishtown’s service is expanding as companies continue to grapple with how to best make use of their data — and how much investors want to be part of that. 

Image Credits: Fishtown

“This was a very productive thing for us,” Fishtown Analytics co-founder and CEO Tristan Handy told me when I asked him why he raised again so quickly. “It’s standard best practice to do quarterly catch-ups with investors and eventually you’ll be ready to fundraise. And Matt Miller from Sequoia showed up to one of these quarterly catch-ups and he shared the 40-page memo that he had written to the Sequoia partnership — and he came with the term sheet.”

Initially, Handy declined. “We’re very bullheaded people, I think, as many founders are. It took some real reflection and thinking about, ‘is this what we want to be doing right now?’ ”

In the end, though, the team decided to go ahead with this round — mostly because this round allowed the team to think long-term and provided stability and certainty.

One thing Handy has always been very clear about is that he did not found Fishtown to purely build the largest possible company but to solve its users’ problems, even as the market looked at companies like Databricks and Snowflake — and their financial success — as potential analogs. “My worry was that the financial markets were driving things that weren’t necessarily going to be good for our users,” Handy said.

Code is the lifeblood of the modern world, yet the tooling for some programming environments can be remarkably spartan. While developers have long had access to graphical programming environments (IDEs) and performance profilers and debuggers, advanced products to analyze and improve lines of code have been harder to find.

These days, the most typical tool in the kit is a linter, which scans through code pointing out flaws that might cause issues. For instance, there might be too many spaces on a line, or a particular line might have a well-known ambiguity that could cause bugs that are hard to diagnose and would best be avoided.

What if we could expand the power of linters to do a lot more though? What if programmers had an assistant that could analyze their code and actively point out new security issues, erroneous code, style problems and bad logic?

Static code analysis is a whole interesting branch of computer science, and some of those ideas have trickled into the real-world with tools like semgrep, which was developed at Facebook to add more robust code-checking tools to its developer workflow. Semgrep is an open-source project, and it’s being commercialized through r2c, a startup that wants to bring the power of this tool to the developer masses.

The whole project has found enough traction among developers that Satish Dharmaraj at Redpoint and Jim Goetz at Sequoia teamed up to pour $13 million into the company for its Series A round, and also backed the company in an earlier, unannounced seed round.

The company was founded by three MIT grads — CEO Isaac Evans and Drew Dennison were roommates in college, and they joined up with head of product Luke O’Malley. Across their various experiences, they have worked at Palantir, the intelligence community, and Fortune 500 companies, and when Evans and Dennison were EIRs at Redpoint, they explored ideas based on what they had seen in their wide-ranging coding experiences.

The r2c team, which I assume only writes bug-free code. Image by r2c.

“Facebook, Apple, and Amazon are so far ahead when it comes to what they do at the code level to bake security [into their products compared to] other companies, it’s really not even funny,” Evans explained. The big tech companies have massively scaled their coding infrastructure to ensure uniform coding standards, but few others have access to the talent or technology to be on an equal playing field. Through r2c and semgrep, the founders want to close the gap.

With r2c’s technology, developers can scan their codebases on-demand or enforce a regular code check through their continuous integration platform. The company provides its own template rulesets (“rule packs”) to check for issues like security holes, complicated errors and other potential bugs, and developers and companies can add their own custom rulesets to enforce their own standards. Currently, r2c supports eight programming languages, including JavaScript and Python, and a variety of frameworks, and it is actively working on more compatibility.

One unique focus for r2c has been getting developers onboard with the model. The core technology remains open-sourced. Evans said that “if you actually want something that’s going to get broad developer adoption, it has to be predominantly open source so that developers can actually mess with it and hack on it and see whether or not it’s valuable without having to worry about some kind of super restrictive license.”

Beyond its model, the key has been getting developers to actually use the tool. No one likes bugs, and no developer wants to find more bugs that they have to fix. With semgrep and r2c though, developers can get much more immediate and comprehensive feedback — helping them fix tricky errors before they move on and forget the context of what they were engineering.

“I think one of the coolest things for us is that none of the existing tools in the space have ever been adopted by developers, but for us, it’s about 50/50 developer teams who are getting excited about it versus security teams getting excited about it,” Evans said. Developers hate finding more bugs, but they also hate writing them in the first place. Evans notes that the company’s key metric is the number of bugs found that are actually fixed by developers, indicating that they are offering “good, actionable results” through the product. One area that r2c has explored is actively patching obvious bugs, saving developers time.

Breaches, errors and downtime are a bedrock of software, but it doesn’t have to be that way. With more than a dozen employees and a hefty pool of capital, r2c hopes to improve the reliability of all the experiences we enjoy — and save developers time in the process.

Today is the day for huge VC returns.

We talked a bit about Sequoia’s coming huge win with the IPO of game engine Unity this morning. Now, Sequoia might actually have the second largest return among companies filing to go public with the SEC today.

Snowflake filed its S-1 this afternoon, and it looks like Sutter Hill is going to make bank. The long-time VC firm, which invests heavily in the enterprise space and generally keeps a lower media profile, is the big winner across the board here, coming out with an aggregate 20.3% stake in the data management platform, which was last privately valued at $12.4 billion earlier this year. At its last valuation, Sutter Hill’s full stake is worth $2.5 billion. My colleagues Ron Miller and Alex Wilhelm looked a bit at the financials of the IPO filing.

Sutter Hill has been intimately connected to Snowflake’s early build-out and success, providing a $5 million Series A funding back in 2012, the year of the company’s founding, according to Crunchbase.

Now, there are some caveats on that number. Sutter Hill Ventures (aka “the fund”) owns roughly 55% of the firm’s total stake, with the balance owned by other entities owned by the firm’s management committee members. Michael Speiser, the firm’s partner who sits on Snowflake’s board, owns slightly more than 10% of Sutter Hill’s stake directly himself according to the SEC filing.

In addition to Sutter Hill, Sequoia also got a large slice of the data computing company: its growth fund is listed as having an 8.4% stake in the coming IPO. That makes for two Sequoia Growth IPOs today — a nice way to start the week this Monday afternoon.

Finally, Altimeter Capital, which did the Series C, owns 14.8%; ICONIQ owns 13.8%; and Redpoint, which did the Series B, owns 9.0%.

To see the breakdown in returns, let’s start by taking a look at the company’s share price and carrying values for each of its rounds of capital:

On top of that, what’s interesting is that Snowflake broke down the share purchases by firm for the last four rounds (D through G-1) the company fundraised:

That level of detail actually allows us to grossly compare the multiples on invested capital for these firms.

Sutter Hill, despite owning large sections of the company early on, continued to buy up shares all the way through the Series G, investing an additional $140 million in the later-stage rounds of the company. Adding in the entirety of its $5 million Series A round and a bit from the Series B assuming pro rata, the firm is looking on the order of a 16x return (assuming the IPO price is at least as good as the last round price).

Outside Sutter Hill, Redpoint has the best multiple return profile, given that it only invested $60 million in these later-stage rounds while still maintaining a 9.0% ownership stake. Both Sutter Hill and Redpoint purchased roughly 20% of their overall stakes in these later-stage rounds. Doing some roughly calculating, Redpoint is looking at a return of about 12-13x.

Sequoia’s multiple on investment is capped a bit given that it only invested in the most recent funding rounds. Its 8.4% stake was purchased for nearly $272 million, all of which came in these late-stage rounds. At Snowflake’s last round valuation of $12.4 billion, Sequoia’s stake is valued at $1.04 billion — a return of slightly less than 4x. That’s very good for mezzanine capital, but nothing like the multiple that Sutter Hill or Redpoint got for investing early.

Doing the same back-of-the-envelope math and Altimeter is looking at a better than 6x return, and ICONIQ got 7x. As before, if the stock zooms up, those returns will look all the better (and of course, if the stock crashes, well…)

One final note: The pattern for these last four funding rounds is unusual for venture capital: Snowflake appears to have “spread the love around,” having multiple firms build up stakes in the startup over several rounds rather than having one definitive lead.