CHAPTER 19

Unity, San Francisco, July-August 2020

For months, Chiou’s importance to the transaction had been growing. With his help, executives had begun to formulate a theory that could inoculate them from some of the risk entailed in disrupting the traditional IPO process. They believed that they could reduce their reliance on the largest institutional investors by finding a meaningful number of smaller investors to hold Unity’s shares for years.

The strategy was an outgrowth of a worry that Jabal and Buyer harbored, which was that some of the largest institutional investors might boycott the company’s IPO if Unity introduced too many new elements into the process. Many investors had sat out Google’s offering for similar reasons, and the lack of demand had contributed to the search giant’s decision to lower its IPO price. Bankers often said that when investors argued for a lower IPO price or against the introduction of new features, it suggested that they might eventually refuse to participate altogether.

Chiou, however, believed that there was an untapped supply of investors that Unity could draw from to ensure a successful offering. There were dozens of smaller funds ready to buy in if only they could get an allocation. The largest underwriters often ignored smaller funds because they weren’t big enough to buy millions of shares and weren’t large payers of commissions to Wall Street trading desks.

More than one investor had chosen to spend less energy evaluating IPO investments because he couldn’t get a big enough allocation through the investment banks to make it worth his time. Others joked ruefully that unless you were a client of Goldman Sachs or Morgan Stanley’s hedge fund business, you weren’t going to get a meaningful number of shares in any of their deals. If only companies he worked with chose to allocate more shares to smaller investors, Chiou argued, they would have a loyal base of shareholders.

Chiou had built his franchise on getting to know those investors. The informal contract he struck was that he would advocate for them to get bigger allocations in IPOs, on one condition: that they act in the company’s best interest. The moment the banker saw them shirking their duties as long-term shareholders, by flipping stock for a quick profit, Chiou would cease to put them forward to company management as loyal shareholders. In other words, the banker leveraged his investor contacts for the express benefit of the company.

If Unity’s auction-like structure alienated a few of the largest mutual funds, like Fidelity or T. Rowe Price—something that Buyer and Jabal were concerned about—Chiou contended that smaller investors would fill the gap.

In early July 2020, the company officially hired Chiou, offering William Blair a bookrunner role, the second tier below Goldman Sachs and Credit Suisse and one higher than the co-manager title that the boutique tended to get on high-profile IPOs. The bookrunner role conveyed bigger fees and thus more credit in the fee rankings watched closely by bankers and their clients, as well as more influence over the details of the transaction.

“Their banker had done a really good job and stayed engaged for a long time and knew John and the rest of the management team,” Davis recalled, referring to Chiou. “Here’s the fascinating thing about the investment banking world. It’s very much like a caste system. You’re kind of born into that position.” The only way to move up was by winning bigger mandates, one transaction at a time.

With his direct reports focused on the minutiae of the upcoming transaction, Riccitiello kept his mind on Unity’s nearly four thousand employees. The CEO had been intent on using the IPO to reward their loyalty.

Riccitiello had a keen understanding of the Silicon Valley talent market. Unity was only as good as the engineers it could hire and retain. Due to the fact that Unity was founded in 2004, it was also relatively long in the tooth by Silicon Valley standards. Many employees had been with the company for years. Riccitiello was obsessed with ensuring that they were able to cash in their shares at the moment the company entered the public markets, or soon after. It’s why he and Jabal initially favored the direct listing, since it gave all shareholders the option to sell any time they wanted. Even though Unity was no longer pursuing a direct listing, he believed that its IPO should have this feature.

For decades, underwriters prevented management teams and employees from selling stock sooner than 180 days after an IPO. The lockup was intended to give the shares time to season in the market. It artificially constrained the supply of stock available to trade, providing something of a floor under the price. Too much supply, the theory went, could depress the stock price.

Viewed from another angle, the lockup gave an advantage to institutional investors over insiders such as venture capitalists, management, and employees. The hypothetical price floor meant that the IPO investors had some assurance, though certainly no guarantee, that they’d have a supportive market in those first six months to exit their position. It gave them a chance to book early gains, while preventing employees and early investors from selling until the lockup expired. It was this dynamic that Hambrecht had pointed out to Templeton all those years ago.

Riccitiello was loath to disadvantage the people who had built Unity into a successful company. This was why he pushed his bankers to come up with a solution that would allow employees to sell early.

During one meeting attended by its bankers at Goldman Sachs, Unity executives floated the idea of allowing employees to sell some portion of their holdings before the customary 180 days. David Ludwig, Goldman’s equity capital markets banker, reluctantly agreed. He suggested lifting the lockup if Unity’s stock price rose by a certain amount. Other companies had begun experimenting with this type of modified lockup, and Ludwig told the executives that he thought something similar could work.

Of course, such a lockup would still provide protection to the institutional investors who had bought into the IPO. They would have a guaranteed return before having to worry about additional supply pushing down the price of their investment.

That wasn’t what Riccitiello had in mind. Was Ludwig suggesting that already well-heeled investors would get a chance to make even more money before his employees got a chance to cash out?

“Let me get this straight,” Riccitiello boomed. “So people on the Upper West Side can buy a bigger house. I have people that are living in four hundred-square-foot apartments in Copenhagen. They would like to be able to buy a big house too.”

The executives and Goldman’s bankers went back and forth. The bankers finally conceded to the idea of some percentage of shares being sold on the first day of trading but pushed to make it small. The two sides finally agreed on 15 percent on the first day and then another 15 percent after Unity reported its first quarterly earnings as a public company. Some of the executives had pushed for a higher percentage, only to face strong resistance from Goldman’s bankers. Riccitiello found 15 percent to be a reasonable compromise. Employees would be free to sell other holdings once 180 days had passed. The executives stopped short of allowing employees to sell in the actual IPO. Unity was aiming for a modest first-day gain and wanted employees to benefit from the higher price before they sold.

In August, Unity finally committed to doing its UPO: an IPO with auction-like properties. While Riccitiello had been on the fence about the modified structure, almost everyone agreed that the company should list the following month. The CEO was satisfied with what they had negotiated with their partners. And global markets had come roaring back from their initial pandemic dip. The timing seemed right to make a splash with an innovative IPO.

For months, Goldman’s bankers had postponed meetings regarding the technical challenges Unity now faced in taking itself public. If investors were going to confidentially submit orders for specific amounts of stock at specific prices, the company needed a new, secure order-entry system. Much of the IPO process was still conducted by telephone. Salespeople at the investment banks called their investor clients to tell them that the company was going public and to learn if they were interested in buying shares. As salespeople learned about investor interest, they could often share that information with other investors, creating a negative feedback loop that artificially depressed the share price. To prevent this, Unity wanted to create a technological upgrade to the sales process. They asked their bankers to create an online order-entry system that investors would use directly without engaging in dialogue with potentially leaky salespeople. This would put an end to the game of telephone that Unity felt would skew their IPO price.

Under the existing system, when salespeople received orders from investor clients, they electronically entered them into something called Equity Bookbuild, or EBB, a product from Ipreo Holdings LLC. EBB acted as a central repository for all the orders banks collected from investors for a particular IPO.

Credit Suisse theorized that Unity could offer an order-entry system that gave investors direct access to EBB. Using a onetime tokenized login could allow investors to place bids without first transmitting them to salespeople, effectively cutting out the middleman. The bankers felt so strongly about this that they persuaded the Ipreo folks to pitch Unity on what their front-end solution would look like.

On Friday, August 14, representatives from Unity, Goldman, Credit Suisse, and Ipreo met to hear Ipreo’s pitch for their product. As Buyer, Jabal, and Unity’s bankers listened in on the phone, Riccitiello showed his surprise that Ipreo’s solution was so rudimentary. He thought he could build a better product in-house, one that would reflect the company’s values and attention to high-end software design.

Another participant put it more baldly. “This is a company that creates 3D software.… You can’t have them throw up an ugly, kludgy-looking 1972 UI,” or user interface.

Unity couldn’t build the product themselves because of SEC regulations. But Riccitiello’s point, that a system for taking orders would be relatively easy to design, stood. As callers peppered Ipreo employees with questions about the potential for other features, they learned that it could take weeks to make such changes.

Goldman had considered the design of a new order-entry system for several months but had waffled on committing to it until Unity’s modified auction plan was settled. After the Ipreo call, Goldman engineers immediately sprang into action and put the finishing touches on what they had started.

On August 19, the Goldman bankers presented their solution. This order-entry system was built on top of a Goldman Sachs website that institutional investors used to access research reports and risk management services. A landing page contained two simple boxes. One asked investors to enter the number of shares they wanted to buy. The other asked them to submit a price. If either box was left empty, the system wouldn’t let investors submit their bids. Investors were encouraged to enter more than one indication at various price points.

Goldman’s engineers also built a charting feature that populated a bar graph with the total number of shares demanded at each price. The tool allowed a user to move a slider up and down to see the resulting decrease or increase in total shares that would be purchased. In that way, it was possible to see how much demand there was, and which investors would be lost at each successively higher price.

“As Unity decided that this was definitively the path they wanted to go down,” one participant in the process said, Goldman “accelerated the development to make sure it was all ready and appropriately tested and all that stuff ahead of its IPO.”

On Monday, August 24, Unity released its first public IPO prospectus. The cover page listed the multitude of banks who would be involved, with Goldman Sachs appearing in the coveted lead-left position, followed by Credit Suisse. The second line of underwriters, in smaller font on a black background, listed Bank of America, Barclays, and William Blair.

Bank of America’s involvement was somewhat unexpected. For months, the firm’s bankers hadn’t been intimately involved in Unity’s deliberations or done much to earn selection. But Johnny Williams, a senior technology banker and Bank of America vice chairman, was close to Unity board member Egon Durban. The two men vacationed with their families at summer homes near each other. Through emails and phone calls, Durban encouraged Unity to add Bank of America to the IPO group.

The document also disclosed Unity’s largest shareholders for the first time. There was little industry surprise that three venture capitalists—Sequoia, Silver Lake, and DFJ Growth—owned roughly 50 percent of the company. A holding company affiliated with cofounder Joachim Ante, Unity’s chief technology officer, owned 8.2 percent, while cofounder David Helgason owned 4.4 percent. Riccitiello came close behind, with his 8.3 million shares and options amounting to a 3.4 percent stake.

The document also showed the savvy of Unity’s lawyers at Cooley. The prospectus didn’t include any language directly pertaining to the company’s modified auction procedure.

Unity’s filing was quickly overshadowed that day by the public release of documents from four other high-profile software companies aiming to go public. Snowflake Inc., Sumo Logic Inc., and JFrog Ltd. disclosed their intention to sell shares in an IPO, while Asana Inc., a developer of project management software founded by Facebook cofounder Dustin Moskovitz, filed plans for a direct listing. Snowflake was a cloud-computing company, Sumo Logic was a software analytics firm, and JFrog made it easier to deploy software updates.

As if to add an exclamation point to the bevy of Monday’s filings, surveillance software developer Palantir filed for a direct listing on Tuesday. The disclosure made the company the third to announce an intention to use Spotify’s creation, following on the heels of Asana’s listing the day before.

The flurry of filings surprised the Unity executives. Months earlier, they had thought that the company would be in a class by itself when it came time to sell shares to the public. Instead, September was shaping up to be one of the busiest months for software IPOs in recent memory. Each of the six firms—Unity, Snowflake, Sumo Logic, JFrog, Asana, and Palantir—competed to win the affection of institutional investors interested in owning software companies.

Snowflake soon attracted more attention than the others. The eight-year-old company operated in the red-hot cloud-computing sector, the business of renting out expensive data centers and their attendant processing power for a fraction of the cost of owning them. Over the past decade, the business had become a cash cow for Amazon, Google, and Microsoft as more and more companies and consumers moved their digital lives into the virtual world. Snowflake didn’t compete with those giants as much as it complemented them, providing tools to companies to manage the data they may have stored across services or space provided by one or more of the other providers. In an industry dominated by an oligopoly of well-established players, Snowflake was a rare opportunity for investors to get in on the ground floor.

The company was run by Frank Slootman, a straight-talking CEO who was as close as the industry could get to an IPO expert. Slootman had already taken two companies public as a CEO—Data Domain Inc. in 2007 and ServiceNow in 2012—which put him in the rare breed of executives with repeat IPO experience. Riccitiello, by contrast, was on his second CEO stint at a public company but experiencing his first IPO with Unity.

Despite his previous experience, Slootman wasn’t an advocate for tearing up the IPO process. He had briefly considered a direct listing and quickly scrapped it because it wouldn’t have helped Snowflake raise the money it needed to fuel its growth plans. Slootman liked the traditional IPO process for the freedom it gave him to pick his shareholder base. Just as Riccitiello and Jabal had been curating a wish list of potential investors, Slootman had been lining up shareholders to buy into his vision for Snowflake’s future.

Slootman was turned off by the prevailing argument then dominating the conversation about first-day stock pops and whether they were a sign of a broken IPO market. Bill Gurley and others calculated the damage of the first-day pop by comparing the IPO price to the first trading day’s closing price. Gurley argued that the difference, multiplied by the number of shares sold in the IPO, was the money a company left on the table by underpricing its shares. Slootman thought that was a weak argument. In his opinion, the last trade represented a price that could be set by a single share trading hands, whereas the IPO involved millions of shares. Investors demanded a discount, reflected in the IPO price, for buying a large volume of shares, because the higher amount would be harder to sell. Slootman would later tell Forbes that the underpricing argument was “nonsense talk.”

He also ignored Gurley’s warnings by choosing Goldman Sachs and Morgan Stanley to lead a cohort of nearly two dozen investment banks underwriting the IPO. According to data compiled by the University of Florida’s Jay Ritter, Goldman Sachs and Morgan Stanley were the two worst underwriters when it came to underpricing IPOs.

Media reports suggested that Snowflake would seek a valuation of $15 billion to $20 billion, making it one of the most valuable members of its IPO cohort.

But first the company and its peers would have to navigate choppy markets. By Friday, August 28, the NASDAQ had fallen more than 10 percent. It was the worst performance for the tech-heavy exchange since the week of March 20, when the pandemic and its impact on the business climate was still uncertain. Nonetheless, the flurry of August filings began to build enthusiasm for what many bankers, investors, and company executives expected would be a busy IPO calendar.