Slack

The S-1 named Microsoft as its primary competitor and described the bundling playbook that would beat it; Microsoft ran that exact play, and Salesforce's $27.7 billion was the escape hatch.

The direct listing itself worked. On June 20, 2019, against a $26 NYSE reference price, WORK opened at $38.50 — a 48% premium valuing Slack around $21.4 billion — and traded above $41 in the first hours. Price discovery without underwriters, the mechanism the risk factors agonized over, proved to be the one thing that went smoothly; Slack's debut helped normalize the direct listing as a path for later companies.

What followed was the competitive script the filing had already written. Microsoft, which had launched Teams inside Office 365, reported 13 million daily active users in July 2019 and 20 million by November 2019 — blowing past the 'more than 10 million' DAU Slack had disclosed in the S-1, at a marginal price of zero to Office customers. Then the pandemic turned the collaboration market into a land grab that rewarded distribution over affection: Teams went from 32 million to 44 million daily users in a single week of March 2020, hit 75 million by April 2020, and 145 million by April 2021. Slack grew too — remote work lifted the whole category — but it was no longer setting the pace, and in July 2020 it filed an EU antitrust complaint accusing Microsoft of illegally tying Teams to Office. The stock, meanwhile, spent long stretches languishing while pandemic-era software peers soared.

On December 1, 2020, Salesforce announced it would acquire Slack for about $27.7 billion — $26.79 in cash plus 0.0776 Salesforce shares per Slack share, worth roughly $45.86 at announcement. The deal closed on July 21, 2021. Read one way, it was a win: 76.4% above the reference price in about two years. Read against the filing's own ambition, it was a surrender — the would-be 'most important software company in the world' selling itself, for roughly its own stated TAM, to one of the partners its S-1 had flagged as a potential competitor, at only about 19% above where its stock had opened on day one twenty-five months earlier. Stewart Butterfield stayed through the transition and left Salesforce in January 2023.

The epilogue vindicated the complaint but not the company. The European Commission opened a formal investigation into Microsoft's Teams bundling in July 2023; Microsoft began unbundling Teams from Office in August 2023; and in June 2024 the EU formally charged Microsoft with antitrust violations for the tying. By then Slack had been a Salesforce product line for three years. The S-1's central risk factor turned out to be neither boilerplate nor pessimism — it was a forecast, and regulators eventually agreed with it, years too late to matter for shareholders.

Slack pitched itself as a new category of business software — channel-based messaging that replaces internal email as the connective layer of the workplace — with consumer-grade engagement metrics no enterprise vendor could match: 10 million daily active users, 88,000 paid customers including 65 of the Fortune 100, $400.6 million in revenue growing 82%, and 143% net dollar retention. Rather than sell shares, it direct-listed on the NYSE, betting that organic word-of-mouth adoption, a developer platform of 1,500+ apps, and 'customer love' would compound into, in its own words, the most important software company in the world.

The best case ran on engagement compounding into inevitability. Users at paid customers spent nine hours a day connected and 90 minutes actively using the product; free teams converted themselves into paid customers through organic advocacy, which is why net dollar retention sat at 143% without a traditional enterprise sales machine. Shared channels between companies would layer a cross-organization network effect on top of the within-company one, the 1,500-app directory would make Slack the interface to every other business tool, and the self-serve funnel plus a young direct sales force would carry it upmarket into Enterprise Grid. If everyone whose working life is mediated by email was a potential user, the $28 billion TAM was a floor, not a ceiling — the shift to Slack, or something like it, was framed as simply inevitable, and Slack held the category lead.

With hindsight: The product thesis was largely right: channel-based collaboration did become a default layer of knowledge work, engagement stayed deep, and enterprises did adopt it at scale — the category's growth through 2020-2021 dwarfed the S-1's $28 billion TAM, with Microsoft Teams alone reaching 145 million daily users by April 2021. But the company thesis failed. Slack did not become the most important software company in the world; it became a feature of two suites — first Microsoft's, by imitation and bundling, then Salesforce's, by acquisition. Shareholders who bought at the $26 reference did fine (~76% in two years via the buyout); anyone who bought the opening print earned about 19% over 25 months while the incumbent it warned about took the market's growth.

• Slack aspires to be the defining company of a new software category — not a tool, but the successor to email.

  “We believe that whoever makes it easiest for teams to function with agility and cohesion in an ever more complex world will be the most important software company in the world. We aim to be that company.” source ↗

• Consumer-scale adoption inside business software: 10 million daily users across 600,000 organizations.

  “During the three months ended January 31, 2019, our daily active users exceeded 10 million. As of January 31, 2019, Slack had more than 600,000 organizations with three or more users” source ↗

• Engagement so deep the product functions as the workday itself.

  “on a typical workday, users at Paid Customers averaged nine hours connected to Slack through at least one device and spent more than 90 minutes actively using Slack.” source ↗

• Best-in-class expansion economics: customers grow their spend 43% year over year on their own.

  “Our Net Dollar Retention Rate was 143% as of January 31, 2019. We believe that our Net Dollar Retention Rate is a reflection of the rapid pace of adoption that often occurs as usage spreads within and across teams.” source ↗

• The category shift is framed as a certainty, with Slack as its leader.

  “because of the universal need for organizational agility and the impact to current and potential organizations on Slack of getting more out of their investment in software, we further believe that the shift to Slack, or services like Slack, is inevitable.” source ↗

1. Microsoft can bundle a competing product into Office at zero marginal price

   came true

   competition · structural

   The filing names Microsoft as its primary competitor and spells out the exact mechanism of defeat: an incumbent with a vastly larger installed base folding equivalent functionality into an existing suite, sold at zero or negative margin, so the buyer never has to make a separate purchase decision. Slack's counter is product quality and focus — it has no structural answer to distribution.

   “some of our larger competitors have substantially broader product offerings and leverage their relationships based on other products or incorporate functionality into existing products to gain business in a manner that discourages users from purchasing Slack, including through selling at zero or negative margins, product bundling, or closed technology platforms.” source ↗

   What happened: Exactly as written. Teams, bundled into Office 365 with a free tier since 2018, reported 13M DAU in July 2019 — weeks after Slack's listing — 20M by November 2019, 44M by March 2020, 75M by April 2020, and 145M by April 2021, while Slack's S-1 figure had been 'more than 10 million.' Slack filed an EU antitrust complaint over the tying in July 2020 and sold to Salesforce five months later. The EU opened a formal probe in July 2023, Microsoft unbundled Teams in August 2023, and the Commission charged Microsoft in June 2024 — posthumous vindication.

2. Substantially all revenue comes from a single product

   came true

   other · structural

   Slack is one product sold one way. A point solution facing suite vendors has no second act if demand for that product is captured, bundled away, or replicated — a concentration the filing itself connects to large organizations choosing to rely on tools they already own, such as email.

   “We derive, and expect to continue to derive, substantially all of our revenue from a single product – Slack. As such, the continued growth in market demand for and market acceptance of Slack is critical to our continued success.” source ↗

   What happened: The point-solution-versus-suite asymmetry defined the endgame. With one product against a bundle priced at zero, Slack's strategic options collapsed to litigation (the EU complaint) or shelter inside someone else's suite. It chose the suite: Salesforce bought it to anchor its own bundle against Microsoft.

3. Five straight years of nine-figure losses with no profitability timeline

   came true

   profitability · serious

   Net losses of $146.9M, $140.1M, and $138.9M in fiscal 2017-2019 against fast-growing revenue, with operating expenses set to rise further as the sales force scales. The filing offers no path or date to profitability — only the hope that revenue growth outruns spending.

   “We expect to continue to incur net losses for the foreseeable future and we may not achieve or maintain profitability in the future.” source ↗

   What happened: Slack never reported an annual profit as an independent public company; it was still operating at a loss when the Salesforce deal was struck in December 2020. The buyout, not the business model, is what delivered shareholder returns.

4. 110% growth decelerating to 82% may keep decelerating

   partly came true

   growth · serious

   The growth rate was already slowing at filing, and the list of things that must all keep working — new users, free-to-paid conversion, retention, international expansion, sales force buildout — is long. In a winner-take-most category, deceleration is not just a valuation problem; it is evidence about who is winning.

   “In future periods, we may not be able to sustain revenue growth consistent with recent history, or at all.” source ↗

   What happened: Growth stayed healthy by normal SaaS standards but continued to decelerate from the 82% at filing, and the pandemic's demand shock — the single greatest tailwind the category would ever get — accrued disproportionately to Teams, whose user counts multiplied roughly elevenfold from mid-2019 to April 2021. Slack grew; it just stopped being the fastest-growing thing in its own category.

5. The free funnel and net dollar retention can stall

   partly came true

   growth · serious

   The model depends on free organizations converting and paid ones expanding, but subscriptions are short, renewal is optional, and fair-billing credits automatically shrink revenue when seats go inactive. A competitor giving away an equivalent product attacks precisely this conversion engine.

   “organizations have no obligation to renew a subscription after the expiration of the term, and we cannot ensure that organizations will renew subscriptions with a similar contract period, with the same or greater number of users, or for the same subscription plan” source ↗

   What happened: Enterprise expansion continued and large customers kept landing, but the top of the funnel now competed with a rival that was literally free to every Office 365 subscriber — the precise pressure this risk described. Slack stopped emphasizing the daily-active-user race once Teams' disclosed figures ran past its own.

6. Slack's platform value depends on rivals' APIs and app stores

   partly came true

   platform dependence · serious

   The integration story that makes Slack a 'layer of the business technology stack' runs through APIs, protocols, and mobile app stores controlled by companies Slack competes with — Google, Microsoft, Apple. Any of them can degrade interoperability or favor their own products, and today's partners (including Salesforce and Zoom) are named as potential future competitors.

   “we currently directly compete with several large technology companies whose applications interface with Slack, including Google and Microsoft. As our respective products evolve, we expect this level of competition to increase.” source ↗

   What happened: The feared API sabotage never became the decisive front — integrations kept working. But the underlying asymmetry the risk gestured at (rivals controlling the surfaces Slack lived on) did the damage through a different door: Microsoft owned the desktop, the suite, and the enterprise agreement, and simply made its own alternative the default. One named partner-turned-threat, Salesforce, resolved the tension by buying the company.

7. A novel direct listing with no book-building or stabilization

   didn't happen

   market · serious

   Only the second major tech direct listing: no underwriters, no roadshow, no price discovery beyond a reference price, no stabilizing bids. The filing warns at length that the opening price may be set disproportionately by retail enthusiasm and that the stock could fall fast from wherever it opens.

   “upon listing on the NYSE, the public price of our Class A common stock may be more volatile than in an underwritten initial public offering and could decline significantly and rapidly” source ↗

   What happened: The mechanism itself performed: the stock opened cleanly at $38.50 against the $26 reference and the listing was orderly enough that the direct-listing model gained legitimacy for later companies. The stock's subsequent underperformance reflected competitive fundamentals, not listing mechanics.

8. No lock-ups: every insider can sell from day one

   didn't happen

   financing · serious

   Unlike a conventional IPO's 180-day lock-up, no stockholder is contractually restricted, and the listing itself triggers vesting and settlement of over 63 million RSUs whose tax withholding forces day-one selling. The float is whatever insiders decide it is.

   “sales of a substantial number of shares of our Class A common stock into the public market, particularly sales by our directors, executive officers and principal stockholders, or the perception that these sales might occur in large quantities, could cause the market price of our Class A common stock to decline” source ↗

   What happened: Day-one RSU-driven selling was absorbed without the feared disorderly glut, and no lock-up-free supply spiral materialized. The absence of lock-ups became a standard, accepted feature of subsequent direct listings rather than a cautionary tale.

9. A security breach in a product holding companies' entire communications

   didn't happen

   tech & security · serious

   Slack stores the full internal conversation history of hundreds of thousands of organizations, making it a high-value target for criminal and nation-state attackers — and it has already been breached once, in 2015. Shared channels can even transmit compromise from one customer to another.

   “for a period of approximately four days in March 2015, a security incident occurred in which unauthorized third parties had access to information maintained by us that included user names, email addresses, encrypted passwords” source ↗

   What happened: No breach on the scale of the disclosed 2015 incident was reported during Slack's roughly two years as a public company. The catastrophic scenario — an attacker inside the message archives of hundreds of thousands of organizations — stayed hypothetical through the acquisition.

10. Outages hit harder when the product is the workplace

    partly came true

    operations · serious

    The engagement stat cuts both ways: when entire organizations do substantially all of their work in Slack, downtime stops the customer's business, triggers service-level credits, and is publicly visible. The filing admits past incidents locked many organizations out entirely.

    “we have experienced intermittent connectivity issues and product issues in the past, including those that have prevented many organizations on Slack and their users from accessing Slack for a period of time” source ↗

    What happened: Outages recurred and made headlines precisely because remote-work dependence made them newsworthy, but none proved existential or triggered material customer flight. The risk was real, chronic, and survivable — a cost of being infrastructure rather than a cause of decline.

11. Total dependence on AWS, terminable on 30 days' notice

    didn't happen

    platform dependence · serious

    All of Slack runs on Amazon's cloud — at filing, from US data centers only — with no owned infrastructure and an agreement AWS can in some cases terminate for cause on 30 days' notice. Amazon is also a large customer and a potential competitor's partner.

    “In some cases, AWS may terminate the agreement for cause upon 30 days’ notice. Termination of the AWS agreement may harm our ability to access data centers we need to host Slack or to do so on terms as favorable as those we have with AWS.” source ↗

    What happened: The AWS relationship held throughout Slack's public life and beyond; no termination, capacity crisis, or pricing squeeze materialized. Single-cloud dependence remained a paper risk.

12. Privacy Shield is under legal attack and GDPR fines scale to 4% of revenue

    partly came true

    regulation · serious

    European operations rest on data-transfer mechanisms — Privacy Shield and model clauses — that were both being litigated at the time of filing, while GDPR exposes the company to fines up to the greater of €20 million or 4% of global turnover. Works councils and data-localization demands can slow enterprise sales in exactly the markets Slack needs.

    “There are concerns about the future of Privacy Shield as a data transfer mechanism as it continues to be subject to legal challenges, which, if successful, would require us to ensure that we had alternative data transfer mechanisms.” source ↗

    What happened: The specific legal challenge the filing flagged succeeded: Europe's top court invalidated the Privacy Shield framework in July 2020 (Schrems II), forcing Slack and the whole industry onto alternative transfer mechanisms such as standard contractual clauses. The disruption was real but industry-wide and absorbed; no material GDPR fine hit Slack.

13. Ten-to-one dual-class stock concentrates control with insiders

    didn't happen

    governance · serious

    Class B holders carry ten votes per share, leaving public Class A buyers with little say over directors, charter changes, or any sale of the company — and likely excluding the stock from major indices under then-current rules.

    “Because of the ten-to-one voting ratio between our Class B and Class A common stock, the holders of our Class B common stock collectively could continue to control a significant percentage of the combined voting power of our common stock” source ↗

    What happened: Concentrated insider control never harmed public holders in practice — if anything it streamlined the one decision that mattered, the sale to Salesforce at a premium, which shareholders approved. The governance structure died with the deal in July 2021.

14. No employment agreements bind Butterfield or any executive

    partly came true

    key person · boilerplate

    The company is closely identified with co-founder and CEO Stewart Butterfield, yet no executive is contractually obligated to stay. Standard language, but sharper here because the product's taste and voice were unusually founder-driven.

    “The loss of one or more of our executive officers, especially our Chief Executive Officer, or key employees could harm our business. Changes in our executive management team may also cause disruptions in, and harm to, our business.” source ↗

    What happened: Butterfield stayed through the listing, the Teams war, and the acquisition — then announced his departure in December 2022 and left Salesforce in January 2023, eighteen months after the deal closed. The key person did eventually walk, but by then Slack was Salesforce's problem, not public shareholders'.

15. The $28 billion TAM is an internal extrapolation

    came true

    other · boilerplate

    The market-size estimate is built by multiplying company counts by Slack's own current customer spend — a methodology the filing itself concedes may not reflect the actual market, in a category it also describes as brand new and rapidly evolving.

    “The estimates and forecasts in this prospectus relating to the size of our target market, market demand and adoption, capacity to address this demand, and pricing may prove to be inaccurate. The addressable market we estimate may not materialize for many years, if ever” source ↗

    What happened: The estimate proved inaccurate — in the opposite direction from the one the risk factor implied. The collaboration market the pandemic revealed was vastly larger than $28 billion: Teams alone reached 145 million daily users by April 2021, and Salesforce paid $27.7 billion — essentially Slack's entire stated TAM — for Slack by itself. The company undersold its market and then lost most of it anyway.

How the S-1 reads

The category-creator S-1 in its purest form: consumer-style love metrics (DAU, hours connected, messages sent) standing in for enterprise durability, a TAM reverse-engineered from the company's own invoices, and a mission statement ('most important software company in the world') doing the work a profitability model would normally do. What makes it singular is the candor — the filing names Microsoft as its primary competitor and describes the bundling mechanism that could kill it, effectively publishing its own bear case. Layered on top are the era's governance fashions (ten-to-one dual-class) and a structural novelty: only the second major tech direct listing, with no underwriters, no lock-ups, and 63 million RSUs vesting into the open. Single product, single cloud (AWS), single revenue line — maximal focus presented as strength, which is the same thing as maximal concentration risk.

• “Our primary competitor is currently Microsoft Corporation. Our other competitors fall into the following categories: productivity tool and email providers, such as Alphabet Inc. (including Google Inc.); unified communications providers, such as Cisco Systems Inc.; and consumer application companies who have entered the business software market, such as Facebook Inc.” source ↗

  The S-1 names its adversary in a single flat sentence — rare specificity that reads as both honesty and warning.

• “We estimate the market opportunity for Slack and other providers of workplace business technology software platforms for communication and collaboration to be $28 billion.” source ↗

  The company's own TAM math put the entire market at $28 billion — a number its first-day valuation would nearly consume.

• “People love using Slack and many become advocates for wider use inside of their organizations. They also tend to recommend Slack when they switch jobs or join organizations that are not yet using Slack.” source ↗

  The growth engine is affection, not procurement — an asset no incumbent had, and one hard to defend at contract-renewal time.

• “We further compete against existing software, programs, and tools, such as email.” source ↗

  Slack's real incumbent was not a product but a habit.

What this one teaches

• When an S-1 names its primary competitor and describes the exact mechanism by which it could lose, treat that as a forecast, not boilerplate. Slack's competition risk factor was the single most accurate paragraph in the document.
• Product love is not a moat against distribution. Nine hours of daily connection and 143% retention lost to a rival that was merely adequate but already installed and priced at zero.
• A single-product company fighting a bundler has three exits: out-innovate, litigate, or get acquired into someone else's bundle. Slack tried all three in eighteen months, in that order.
• Antitrust can validate your complaint and still not save you. The EU ultimately charged Microsoft over Teams bundling in 2024 — four years after Slack's complaint and three years after Slack ceased to exist as an independent company.
• Judge direct listings by the mechanism and the business separately. Slack's listing mechanics worked fine; the reference-price 'pop' just meant early buyers paid $38.50 for a company that would be taken out at $45.86 two years later.
• A TAM derived from your own current pricing can be wrong in both directions at once: Slack's $28B estimate understated the market's size and overstated Slack's claim on it.

The paper trail

1. 2019-04-26 S-1 filing index ↗ document ↗
2. 2019-05-13 S-1/A filing index ↗ document ↗
3. 2019-05-20 S-1/A filing index ↗ document ↗
4. 2019-05-31 S-1/A filing index ↗ document ↗
5. 2019-06-20 424B4 filing index ↗ document ↗

Filed as Slack Technologies, Inc.. All documents are public domain, served by SEC EDGAR.