It’s Squarespace direct-listing day, and the SMB web hosting and design shop’s reference price has been set at $50 per share.

According to quick math from the IPO-watching group Renaissance Capital, Squarespace is worth $7.4 billion at that price, calculated using a fully diluted share count. The company’s new valuation is sharply under where Squarespace raised capital in March, when it added $300 million to its accounts at a $10 billion post-money valuation, according to Crunchbase data.

---

The Exchange explores startups, markets and money. 

Read it every morning on Extra Crunch or get The Exchange newsletter every Saturday.

---

The company’s reference price, however, is just that: a reference. It doesn’t mean that much. As we’ve seen from other notable direct listings, a company’s opening price does not necessarily align with its formal reference price. Until Squarespace opens, whether it will be valued at a discount to its final private price is unclear.

While the benefits of a direct listing are understood, the post-listing performance for well-known direct listings is less obvious. Indeed, Coinbase is currently under its reference price after starting its life as a public company at a far-richer figure, and Spotify’s share price is middling at best compared to its 2018-era direct-listing reference price.

This morning, we’re going over Squarespace’s recently disclosed Q2 and full-2021 guidance. Then we’ll ask how its expectations compare to its reference price-defined pre-trading valuation. Finally, we’ll set some stakes in the ground regarding historical direct-listing results and what we might expect from the company as it adds a third set of data to our quiver.

This will be lots of fun, so let’s get into the numbers!

Squarespace’s Q2

Per Squarespace’s own reporting, it expects revenues between $186 million and $189 million in Q2 2021, which it calculates as a growth rate of between 24% and 26%. That pace of growth at its scale is perfectly acceptable for a company going public.

For all of 2021, Squarespace expects revenues of $764 million to $776 million, which works out to a very similar 23% to 25% growth rate.

In profit terms, Squarespace only shared its “non-GAAP unlevered free cash flow,” which is a technical thing I have no time to explain. But what matters is that the company expects some non-GAAP unlevered free cash flow in Q2 2021 ($10 million to $13 million), and lots more in all of 2021 ($100 million to $115 million).

Automation is extending into every aspect of how organizations get work done, and today comes news of a startup that is building tools for one industry in particular: life sciences. Artificial, which has built a software platform for laboratories to assist with, or in some cases fully automate, research and development work, has raised $21.5 million.

It plans to use the funding to continue building out its software and its capabilities, to hire more people, and for business development, according to Artificial’s CEO and co-founder David Fuller. The company already has a number of customers including Thermo Fisher and Beam Therapeutics using its software directly and in partnership for their own customers. Sold as aLab Suite, Artificial’s technology can both orchestrate and manage robotic machines that labs might be using to handle some work; and help assist scientists when they are carrying out the work themselves.

“The basic premise of what we’re trying to do is accelerate the rate of discovery in labs,” Fuller said in an interview. He believes the process of bringing in more AI into labs to improve how they work is long overdue. “We need to have a digital revolution to change the way that labs have been operating for the last 20 years.”

The Series A is being led by Microsoft’s venture fund M12 — a financial and strategic investor — with Playground Global and AME Cloud Ventures also participating. Playground Global, the VC firm co-founded by ex-Google exec and Android co-creator Andy Rubin (who is no longer with the firm), has been focusing on robotics and life sciences and it led Artificial’s first and only other round. Artificial is not disclosing its valuation with this round.

Fuller hails from a background in robotics, specifically industrial robots and automation. Before founding Artificial in 2019, he was at Kuka, the German robotics maker, for a number of years, culminating in the role of CTO; prior to that, Fuller spent 20 years at National Instruments, the instrumentation, test equipment and industrial software giant. Meanwhile, Artificial’s co-founder, Nikhita Singh, has insight into how to bring the advances of robotics into environments that are quite analogue in culture. She previously worked on human-robot interaction research at the MIT Media Lab, and before that spent years at Palantir and working on robotics at Berkeley.

As Fuller describes it, he saw an interesting gap (and opportunity) in the market to apply automation, which he had seen help advance work in industrial settings, to the world of life sciences, both to help scientists track what they are doing better, and help them carry out some of the more repetitive work that they have to do day in, day out.

This gap is perhaps more in the spotlight today than ever before, given the fact that we are in the middle of a global health pandemic. This has hindered a lot of labs from being able to operate full in-person teams, and increased the reliance on systems that can crunch numbers and carry out work without as many people present. And, of course, the need for that work (whether it’s related directly to Covid-19 or not) has perhaps never appeared as urgent as it does right now.

There have been a lot of advances in robotics — specifically around hardware like robotic arms — to manage some of the precision needed to carry out some work, but up to now no real efforts made at building platforms to bring all of the work done by that hardware together (or in the words of automation specialists, “orchestrate” that work and data); nor link up the data from those robot-led efforts, with the work that human scientists still carry out. Artificial estimates that some $10 billion is spent annually on lab informatics and automation software, yet data models to unify that work, and platforms to reach across it all, remain absent. That has, in effect, served as a barrier to labs modernising as much as they could.

A lab, as he describes it, is essentially composed of high-end instrumentation for analytics, alongside then robotic systems for liquid handling. “You can really think of a lab, frankly, as a kitchen,” he said, “and the primary operation in that lab is mixing liquids.”

But it is also not unlike a factory, too. As those liquids are mixed, a robotic system typically moves around pipettes, liquids, in and out of plates and mixes. “There’s a key aspect of material flow through the lab, and the material flow part of it is much more like classic robotics,” he said. In other words, there is, as he says, “a combination of bespoke scientific equipment that includes automation, and then classic material flow, which is much more standard robotics,” and is what makes the lab ripe as an applied environment for automation software.

To note: the idea is not to remove humans altogether, but to provide assistance so that they can do their jobs better. He points out that even the automotive industry, which has been automated for 50 years, still has about 6% of all work done by humans. If that is a watermark, it sounds like there is a lot of movement left in labs: Fuller estimates that some 60% of all work in the lab is done by humans. And part of the reason for that is simply because it’s just too complex to replace scientists — who he described as “artists” — altogether (for now at least).

“Our solution augments the human activity and automates the standard activity,” he said. “We view that as a central thesis that differentiates us from classic automation.”

There have been a number of other startups emerging that are applying some of the learnings of artificial intelligence and big data analytics for enterprises to the world of science. They include the likes of Turing, which is applying this to helping automate lab work for CPG companies; and Paige, which is focusing on AI to help better understand cancer and other pathology.

The Microsoft connection is one that could well play out in how Artificial’s platform develops going forward, not just in how data is perhaps handled in the cloud, but also on the ground, specifically with augmented reality.

“We see massive technical synergy,” Fuller said. “When you are in a lab you already have to wear glasses… and we think this has the earmarks of a long-term use case.”

Fuller mentioned that one area it’s looking at would involve equipping scientists and other technicians with Microsoft’s HoloLens to help direct them around the labs, and to make sure people are carrying out work consistently by comparing what is happening in the physical world to a “digital twin” of a lab containing data about supplies, where they are located, and what needs to happen next.

It’s this and all of the other areas that have yet to be brought into our very AI-led enterprise future that interested Microsoft.

“Biology labs today are light- to semi-automated—the same state they were in when I started my academic research and biopharmaceutical career over 20 years ago. Most labs operate more like test kitchens rather than factories,” said Dr. Kouki Harasaki, an investor at M12, in a statement. “Artificial’s aLab Suite is especially exciting to us because it is uniquely positioned to automate the masses: it’s accessible, low code, easy to use, highly configurable, and interoperable with common lab hardware and software. Most importantly, it enables Biopharma and SynBio labs to achieve the crowning glory of workflow automation: flexibility at scale.”

Harasaki is joining Peter Barratt, a founder and general partner at Playground Global, on Artificial’s board with this round.

“It’s become even more clear as we continue to battle the pandemic that we need to take a scalable, reproducible approach to running our labs, rather than the artisanal, error-prone methods we employ today,” Barrett said in a statement. “The aLab Suite that Artificial has pioneered will allow us to accelerate the breakthrough treatments of tomorrow and ensure our best and brightest scientists are working on challenging problems, not manual labor.”

Without good data, it’s impossible to build an accurate predictive machine learning model. Explorium, a company that has been building a solution over the last several years to help data pros find the best data for a given model, announced a $75 million Series C today — just 10 months after announcing a $31 million Series B.

Insight Partners led today’s investment with participation from existing investors Zeev Ventures, Emerge, F2 Venture Capital, 01 Advisors and Dynamic Loop Capital. The company reports it has now raised a total of $127 million. George Mathew, managing partner at Insight, and former president and COO at Alteryx, will be joining the board, giving the company someone with solid operator experience to help guide them into the next phase.

Company co-founder and CEO Maor Shlomo, says that in spite of how horrible COVID has been from a human perspective, it has been a business accelerator for his company and he saw revenue quadruple last year (although he didn’t share specific numbers beyond that). “It’s related to the nature of our business. We’re helping enterprises and data practitioners find new data sources that can help them solve business challenges,” Sholmo explained.

He says that during the pandemic, a lot of companies had to find new data sources because the old data wasn’t especially helpful for predictive models. That meant that customers required new sources to give them visibility into the shifts and movements in the market to help them adjust and make decisions during pandemic. “And given that’s basically what our platform does in its essence, we’ve seen a lot of growth [over the past year],” he says.

With the revenue growth the company has been experiencing, it has been adding employees at rapid clip. When we spoke to Explorium last July, the company had 87 people. Today that number has grown to 130 with plans to get to 200 perhaps by the end of 2021 or early 2022, depending on how the business continues to grow.

The company has offices in Tel Aviv and San Mateo, California with plans to open a new office in New York City whenever it’s possible to do so. While Shlomo wants a flexible workplace, he’s not going fully remote with plans to allow people to work two days from home and three in the office as local rules allow.

Like “innovation,” machine learning and artificial intelligence are commonplace terms that provide very little context for what they actually signify. AI/ML spans dozens of different fields of research, covering all kinds of different problems and alternative and often incompatible ways to solve them.

One robust area of research here that has antecedents going back to the mid-20th century is what is known as stochastic optimization — decision-making under uncertainty where an entity wants to optimize for a particular objective. A classic problem is how to optimize an airline’s schedule to maximize profit. Airlines need to commit to schedules months in advance without knowing what the weather will be like or what the specific demand for a route will be (or, whether a pandemic will wipe out travel demand entirely). It’s a vibrant field, and these days, basically runs most of modern life.

Warren B. Powell has been exploring this problem for decades as a researcher at Princeton, where he has operated the Castle Lab. He has researched how to bring disparate areas of stochastic optimization together under one framework that he has dubbed “sequential decision analytics” to optimize problems where each decision in a series places constraints on future decisions. Such problems are common in areas like logistics, scheduling and other key areas of business.

The Castle Lab has long had industry partners, and it has raised tens of millions of dollars in grants from industry over its history. But after decades of research, Powell teamed up with his son, Daniel Powell, to spin out his collective body of research and productize it into a startup called Optimal Dynamics. Father Powell has now retired full-time from Princeton to become chief analytics officer, while son Powell became CEO.

The company raised $18.4 million in new funding last week from Bessemer led by Mike Droesch, who recently was promoted to partner earlier this year with the firm’s newest $3.3 billion fundraise. The company now has 25 employees and is centered in New York City.

So what does Optimal Dynamics actually do? CEO Powell said that it’s been a long road since the company’s founding in mid-2017 when it first raised a $450,000 pre-seed round. We were “drunkenly walking in finding product-market fit,” Powell said. This is “not an easy technology to get right.”

What the company ultimately zoomed in on was the trucking industry, which has precisely the kind of sequential decision-making that father Powell had been working on his entire career. “Within truckload, you have a whole series of uncertain variables,” CEO Powell described. “We are the first company that can learn and plan for an uncertain future.”

There’s been a lot of investment in logistics and trucking from VCs in recent years as more and more investors see the potential to completely disrupt the massive and fragmented market. Yet, rather than building a whole new trucking marketplace or approaching it as a vertically integrated solution, Optimal Dynamics decided to go with the much simpler enterprise SaaS route to offer better optimization to existing companies.

One early customer, which owned 120 power units, saved $4 million using the company’s software, according to Powell. That was a result of better utilization of equipment and more efficient operations. They “sold off about 20 vehicles that they didn’t need anymore due to the underlying efficiency,” he said. In addition, the company was able to reduce a team of 10 who used to manage trucking logistics down to one, and “they are just managing exceptions” to the normal course of business. As an example of an exception, Powell said that “a guy drove half way and then decided he wanted to quit,” leaving a load stranded. “Trying to train a computer on weird edge events [like that] is hard,” he said.

Better efficiency for equipment usage and then saving money on employee costs by automating their work are the two main ways Optimal Dynamics saves money for customers. Powell says most of the savings come in the former rather than the latter, since utilization is often where the most impact can be felt.

On the technical front, the key improvement the company has devised is how to rapidly solve the ultra-complex optimization problems that logistics companies face. The company does that through value function approximation, which is a field of study where instead of actually computing the full range of stochastic optimization solutions, the program approximates the outcomes of decisions to reduce compute time. We “take in this extraordinary amount of detail while handling it in a computationally efficient way,” Powell said. That’s where we have really “wedged ourselves as a company.”

Early signs of success with customers led to a $4 million seed round led by Homan Yuen of Fusion Fund, which invests in technically sophisticated startups (i.e. the kind of startups that take decades of optimization research at Princeton to get going). Powell said that raising the round was tough, transpiring during the first weeks of the pandemic last year. One corporate fund pulled out at the last minute, and it was “chaos ensuing with everyone,” he said. This Series A process meanwhile was the opposite. “This round was totally different — closed it in 17 days from round kickoff to closure,” he said.

With new capital in the bank, the company is looking to expand from 25 employees to 75 this year, who will be trickling back to the company’s office in the Flatiron neighborhood of Manhattan in the coming months. Optimal Dynamics targets customers with 75 trucks or more, either fleets for rent or private fleets owned by companies like Walmart who handle their own logistics.