This report was created in collaboration with Tegus...
Tegus provides a best-in-class platform for primary research. That includes thousands of investor-led interviews, all searchable through Tegus's intuitive platform. That proved exceptionally useful as The S-1 Club set about researching UiPath. With 47 (!) company-specific interviews available, I can say that Tegus was, without a doubt, the single most useful source of information we accessed. Not only did it give us an inside look at UiPath's operations, it helped illuminate the company's competitive advantages, and the key dynamics within the RPA market.
You'll find references from $PATH VPs, Directors, implementation consultants, and senior clients throughout this document. All of them came from Tegus. If you're an investor looking for an edge, I highly recommend contacting the team.
UiPath in 1 minute
UiPath wants to make you more human. The RPA industry leader has created an impressive business helping companies realize efficiency gains while portraying itself as the solution to mundane, repetitive work. In the last fiscal year, $PATH brought in $607.6 million in revenue, up 81% from the year prior. Gross margins approached 90% and dollar-based net retention hit a remarkable 145%.
Despite its dominance and impressive numbers, UiPath is yet to turn a profit. Perhaps more concerningly, the industry it operates in is changing rapidly, with more technologically advanced solutions disrupting the market. To maintain its foothold, UiPath will need to use its IPO as a catalyst for continued innovation.
Let’s start with a joke.
Which villain from the Marvel Cinematic Universe works as an RPA consultant?
Take a moment.
Thanos. He snaps his fingers, cuts headcount 50%, then moves on.
If there’s a knock on the RPA industry, this is it: though technically impressive and often useful, it comes at a human cost. As software robots take over an increasing number of tasks, jobs vaporize piece by piece, dissolving like one more of the Mad Titan’s victims.
UiPath sees itself differently, arguing that it’s part of Schumpeter’s gale rather than Thanos’s cyclone. Yes, the company argues, some jobs might be lost. But those that remain will be unburdened of the most monotonous, soul-sapping tasks, freed to do meaningful, creative work. Through elimination, we are made more human.
This is UiPath’s vision of itself: a balmy zephyr of creative destruction, ushering in a more genial era. Whether one agrees with this self-assessment or not — the S-1 filing does a convincing job of advancing this narrative, as one would expect — resistance seems not just futile but foolish, akin to a Luddite handweaver dismantling a mill. Of the Fortune 10, UiPath already serves 8 of them; of the Fortune Global 500, the company counts 305 as customers. While Thanos once declared, “I am inevitable,” UiPath might reasonably say, “I’ve already won.” The corporate world needs no convincing that the future lies in automation, and UiPath has established itself as the industry leader. Roll credits.
And yet, questions linger. Anyone who’s watched a Marvel film knows the movie doesn’t end with the credits. Stay long enough, and just before the lights come up, a coda crosses the screen, introducing a new hero or villain, setting up the next battle.
While UiPath has dominated the sector with its offering, the past few years have seen a new competitive set gain strength. API-first companies claim to provide the benefits of UiPath— improving efficiency through automation—with greater reliability. These insurgents contend that while UiPath’s graphical approach made sense in the past, the proliferation of modern APIs makes the methodology outdated and insufficient.
That’s without even mentioning software’s Kronos (a titan of the astrological rather than cinematic variety), the planet-gobbling Microsoft. With a play in the RPA space and a record of devouring market share (see: Slack and Teams), the Seattle-based company represents a formidable adversary.
The next few years could prove critical. While this chapter of UiPath’s journey should rightfully conclude with much fanfare on listing day, the story is just beginning. Perfectly suited to the last decade, UiPath’s triumph has long looked like a foregone conclusion. The company will hope that as the space shifts, it retains that aura of inevitability.
Number of mentions in S-1
- Automation: 425
- AI: 67
- Daniel Dines: 66
- RPA: 46
- “Fully automated enterprise”: 15
- Humble: 4
- Automation Anywhere: 1
Daniel Dines’ journey as CEO of UiPath has hardly been simple. The explicit mention of “humility” in the S-1 is a clue to this non-linear path of determined survival. From its humble origins in Bucharest, UiPath is a company that doesn't take anything for granted.
An entrepreneurial streak
Daniel Dines became a software engineer by happenstance.
Despite dreaming of becoming an author when he was young, Dines discovered he had a proclivity for the numerical and an eye for the entrepreneurial. During college, the Romanian national supported his mathematics degree — he rarely attended lectures — by competing in bridge tournaments and capitalizing on his country’s turbulent leu. A Forbes profiles notes:
He supported himself as a post-Communist arbitrager of Romania’s inflationary currency, buying goods when they were cheaper in Bucharest and sending them home with a markup.
After graduating with a Master’s from the University of Bucharest, Dines turned his attentions to business, founding a job listings company. During this period, he heard about the money software developers could earn working for US companies remotely. Keen to get in on the action, he picked up a book on C++ and taught himself the ropes.
The bet paid off. In 2001, Dines was hired by Microsoft and relocated to Seattle. He spent five years at the company before returning to Romania with a new mission: to start a technology company of his own.
Before there was UiPath, there was DeskOver.
Founded in 2005, this precursor business was part consultancy, part software development shop. DeskOver was a long way from the venture-backed behemoth UiPath would become. Despite his experience in the US, Dines didn’t consider venture capital an option given its scarcity in Romania. Having been inside the Microsoft behemoth, he also wasn’t sure he wanted to build a global scale business, noting as much in an early interview:
My motivation [with DeskOver] was, I think, to build the minimum wealth. To have a decent life, in a way – Romania was never an easy country in which to have that. And especially to be an entrepreneur. And going through the tough years, in the early 90s in Romania, to me was an experience that touched my entire life. So I was always a bit concerned about building some wealth to live a decent life.
As Dines became more connected to the tech ecosystem, though, his ambitions began to change. Without any local entrepreneurial leaders, Dines found inspiration through TechCrunch articles and the essays of Paul Graham. The Y Combinator founder’s pieces, in particular, showed Dines the power of technology and the virtue of building something for the sake of human progress.
What changed my perspective was looking at new startups coming out of Silicon Valley. TechCrunch started talking about new start-ups, and then there was Y Combinator and then Paul Graham, one of my virtual mentors whose thinking really influenced me deeply. And I saw that you could actually build a company out of love for technology. For the sake of doing something for the greater good, of doing something better in this world. It changed my thinking completely. We gave up outsourcing completely and we said, “let’s build the best thing that we can and then we’ll see what happens.” That was really a pivotal thing in our history.
Inspired by the potential to change the world through software, Dines decided he wanted more than just “minimum wealth.” He wanted to turn his budding company into something bigger.
Finding a business model
DeskOver’s newest incarnation began with a classic entrepreneurial mistake: building software without a clear customer in mind.
Operating from an apartment not far from Bucharest’s Old Town, DeskOver started by building automation libraries and software development kits (SDKs) that helped programmers build applications. For years, DeskOver proved a solid if unspectacular business operating with little fanfare, low burn, and spotty revenue.
The turning point arrived in 2013. A business process outsourcing (BPO) company based in India contacted Dines. They’d discovered DeskOver’s software and been impressed, noting its capabilities compared favorably to Blue Prism, a much larger business and coiner of the term “Robotic Process Automation.”
Surprisingly, the Indian firm noted they were employing DeskOver’s technology differently than intended, using the SDKs to conduct data entry automatically. Intrigued, Dines sent his team to spend three months with the firm, sniping the contract from Blue Prism in the process. It proved a lightbulb moment for Dines:
We understood there was a huge market out there of people who just do repetitive processes all day long, for whom our technology, which emulates what people do, is perfect. From that moment on, things changed dramatically for us. We started to grow.
After refocusing DeskOver on the RPA market, Dines launched the “UiPath Desktop RPA Suite,” naming the new product after a reference in the codebase. DeskOver drew inspiration from Dines’s former employer, modeling its positioning on Microsoft’s Workflow Designer.
In that respect, DeskOver’s highly technical team benefitted from a late-mover advantage with the CEO recognizing his company was playing catch-up against global behemoths:
In 2014 and 2015 it seemed to me that we had giant competitors, but without looking at them, we started to grow the team and things later exploded. There comes a moment when you need to get a little crazy.
More than craziness, Dines needed to show perseverance in securing funding for his newly focused software firm. It took the company 14 months to close a $1.6 million seed round from Credo Ventures and Earlybird. The round finally closed in 2014; Seedcamp tacked on the following year.
It was time to grow. A reflection of the RPA Suite’s centrality to the business, Dines renamed the company UiPath and set about signing strong channel partners. These BPOs and consulting firms distributed UiPath’s technology to clients, taking a cut in the process.
UiPath’s big break was in partnering with Ernst & Young’s (EY) Romania team. Partnering with a global firm gave the company the chance to not only win domestic clients but extend its capabilities around the world. Aurelia Costache, advisory services leader at EY Romania and head of the RPA Center of Excellence in Bucharest, noted as much:
EY Romania became aware of UiPath’s technology in 2015 and, since then, we have been developing our partnership to jointly bring our clients the value-added of an end-to-end solution. With the benefit of proximity, we were able to work together in implementing the first RPA projects in Romania… Being a globally integrated firm, we were able to support our offices in other countries with the UiPath expertise, facilitating UiPath’s access to EY’s global clients as well.
EY’s show of faith was critical in UiPath’s future distribution. With money coming in, UiPath entered a period of hypergrowth, opening offices in Bangalore, London, and New York and bringing aboard hundreds of enterprise clients. UiPath even scored a $300,000 contract with General Electric, illustrating the firm’s ability to win and serve the Fortune 500.
That progress caught the attention of Accel, one of the world’s most prominent venture firms. Romanian-born partner Luciana Lixandru was intrigued by what Dines was building:
What I thought was very intriguing early on was that the company did not have any real feet on the ground in the US and yet they were getting a lot of attention from US customers, riding a very timely trend.
That interest became concrete, with Accel leading a $30 million Series A in 2017. Lixandru later recalled Dines’s parting words: “I will make you guys a lot of money!’”
It wouldn’t take long for Dines to demonstrate the seriousness of that promise. Over that year, UiPath opened offices in Australia, France, Japan, and Singapore, bringing headcount to over 400. The company also launched its Academy, a feature that allowed thousands of interested parties to learn how to leverage UiPath’s software.
The following year proved another inflection point. In 2018, UiPath increased its number of clients by 611%, counting Allianz, BMW Group, CenturyLink, Dairy Farm Group, Dentsu Inc., Huawei, Morningstar, and Sumitomo Mitsui Financial Group.
That growth powered another round a year later, this time at a unicorn valuation. UiPath raised $153 million in a Series B that pegged the company at $1.1 billion.
Once again, Accel led the deal, with Rich Wong joining the board. CapitalG, one of Google’s venture arms, and Kleiner Perkins Caufield & Byers (KPCB) contributed. In a far cry from UiPath’s first funding round, the company raised in just two months from a suite of suitors.
The Google connection
CapitalG would co-lead UiPath’s Series C, too, collaborating with Sequoia Capital. That further cemented the relationship between UiPath and Google, a valuable strategic ally in the company’s battle against other tech titans. In particular, Dines highlighted the technical capabilities of CapitalG’s parent, saying, “We chose Google for its interest in AI.”
It doesn’t hurt, of course, that Google competes with Microsoft directly in many markets, establishing an “enemy of my enemy is my friend” dynamic between UiPath and the search giant. Intriguingly, Dines turned down investment from Softbank’s Vision Fund — partially capitalized by Microsoft — in choosing CapitalG to lead that round.
That hasn’t prevented UiPath from securing further mega-rounds. Coatue led a $568 million Series D in 2019, followed by an additional $225 million from the firm, in conjunction with Alkeon Capital.
Earlier this year, with the expectation that UiPath’s public listing was just around the corner, the company instead announced news of a $750 million Series F from the same duo. That final private round valued the company at $35 billion post-money.
Surely Daniel Dines couldn’t have imagined the dizzying heights his business has achieved. But if UiPath’s core values are anything to judge the company by, he won’t be letting that get to his head.
It’s the natural evolution of a human being, to reach humbleness. Ego is the worst enemy in one’s life. And to us humbleness came a bit from the realization that we were not the best software engineers in the world. We don’t have the best ideas in the world. And what can make us successful is really the desire to do something better. To become better. And only people who think within the humbleness framework can improve. Because otherwise you think you are good enough. And humbleness helps you overcome this type of barrier.
Given UiPath’s long road to success, that humility has been hard-earned.
UiPath is after no smaller goal than augmenting humanity. Given the audacity of that vision, it’s worth understanding the intellectual influences that have led to the creation of the RPA market. Below, we’ll discuss that heritage, in addition to analyzing the complexion of the current space.
Intellectual influences: Man and machine
In 1960, computer scientist JCR Licklider published his seminal piece “Man Computer Symbiosis.”
In his work, Licklider theorized that computing would aid human activity, eventually extending our capabilities. Rather than seeing technology as a dystopian agent of obsolescence, Licklider predicted a close, fruitful coupling of man and machine.
Steve Jobs famously echoed this belief. In one of his early interviews, Jobs suggested that computers would advance humanity, serving as a “bicycle for the mind.”
Though Licklider and Jobs influenced many successive companies — Palantir’s founders were particularly enamored with Licklider’s work — this intellectual ancestry speaks directly to the RPA space. With a stated mission to “help humans achieve more,” understanding this heritage informs how observers look at UiPath’s current market and potential future. While defining RPA narrowly might not produce an especially large TAM, such an appraisal might miss the totality of what Dines’ company is chasing.
Market development: The rise of mundanity
As technology has flourished, so has complexity. The last few decades have seen a proliferation in the number of computing devices and software systems. Increasingly, corporations and the knowledge workers they employ must manage extensive, fragmented systems.
That has led to the creation of “meta-work,” or work about work. While one’s stated job might be as a product designer, executing that role could include compiling team goals, following up with individual contributors, porting files from one platform to another, or any number of other mundane, repetitive tasks.
To optimize productivity, knowledge workers often resorted to “hacks” that helped smooth these processes. Running an Excel macro was one solution — an early example of RPA.
Companies began to recognize that many of their most highly trained employees frittered time on low-value tasks. To try and focus available hours on higher acuity tasks, corporations began outsourcing humdrum tasks like document processing or data entry, leading to the BPO boom in countries with a lower cost of labor, like India.
While BPO remains an ample-sized space, the turn of the millennium saw technological innovations undercut the necessity of human assistance. Pioneers like Blue Prism, founded in 2001, leveraged computer vision to create more advanced RPA software. One RPA integration consultant summarized how technology affected the BPO industry:
[W]hat I've kind of seen in terms of patterns for implementing RPA, it's a lot of companies that are offshoring different functions that they have…[W]hen I was working in the Big 4 firm...a lot of their customers came from big banks and financial institutions, where they have these outdated ERP systems…
[W]here they started using RPA was essentially, they hired people in India, in China and all over the world to look at their doc-based ERP system and look at a number on there and type that number into a modern system that they had. And that was the person's job. And they had hundreds of people hired where, essentially, their role would be to key in data by looking at one screen and putting it on the other screen.
So how they started using RPA was to do that process of reading one screen, taking the data values, typing it on to the other screen...they were able to eliminate the need to offshore that function...one robot or a couple of robots...would replace over 100 people in that particular function.
BPOs adapted by adopting the software themselves, recognizing that their manual work could be significantly expedited. Many of these firms served as distributors and implementers of RPA solutions. As shared under “History,” even UiPath’s prominence relied on the savvy and belief of a BPO firm.
RPA: Darling or dud?
More than most spaces, UiPath’s market is a matter of definition. In a previous briefing from The Generalist, we shared that some sources peg the RPA space between $2 billion and $7 billion. Statista projects the market will be worth $5.4 billion in 2021. RPA has been considered a hot space with Gartner dubbing it the "fastest-growing market in enterprise software” after it spiked 63% between 2019 and 2020.
Despite that commendation, RPA is often viewed with skepticism. That’s because integrations often don’t deliver the value they promise. While industry goliaths like UiPath and Automation Anywhere are more than happy to wheel out case studies illustrating the cost savings they’ve captured, delight is not universal. Ernst & Young noted that 30-50% of initial RPA deployments fail with organizational confusion and IT debt contributing to deterioration. Notably, Accenture stopped delivering RPA services to some geographies after struggling to provide ROI for customers. (They were using Blue Prism.)
Perhaps unsurprisingly then, UiPath takes a more expansive view of its market. Rather than seeing itself as competing in the RPA space, UiPath refers to an estimation of the “automation software” space:
According to an estimate by Bain & Company in the report “Beyond Cost Savings: Reinventing Business Through Automation,” the expansion of automation platforms by incorporating broader capabilities and technologies has increased the size of the addressable market for automation software to approximately $65 billion.
Hearkening back to “bicycle of the mind” territory, the S-1 subsequently notes that UiPath sees its market opportunity as “the substantial amount of business processes that could be improved through automation, but are not currently automated,” a capacious notion.
With that framing in mind, what markets will UiPath attempt to attack next?
Frontiers: Where next?
UiPath will wish to aggressively expand into cloud services to stay ahead of the curve. While the company’s technology is solid, it was nevertheless founded more than fifteen years ago. There are some promising initiatives underway (more under “Product”), but successfully managing this migration represents a risk, as well as a significant opportunity.
Beyond that, the company seems to have designs on the low-code space, highlighting the market as an area for potential expansion. UiPath’s “drag and drop” interfaces touch on this movement, but it’s a far cry from the offerings of companies like Appian and Pega that the S-1 highlights.
UiPath’s product suite spans different stages of the automation journey. Along the way, the company showcases different technology, including traditional RPA, AI/ML, Process Mining, and more.
We’ll dig into what the platform does, where it’s going, and how it separates itself from the competition. In the process, we’ll seek to separate myth from reality. As with every provider in the space, UiPath oversells the capabilities of its technology, willfully conflating traditional RPA (rules-based actions) with AI-driven automation.
RPA solutions automate tasks by emulating human behavior on computers (as you’ve probably picked up by now). Bots move the mouse, click on screens, and type on keyboards virtually. In essence, these bots follow a set script, performing the same actions a human might have done previously.
As noted, this basic functionality is not new. Excel supported end-user programming of macros as far back as 1987. The actual change RPA platforms brought about (particularly starting in the 2010s) was scaling these rule-based bots beyond Excel, across applications. This is still where RPA shines, per an RPA implementation expert. (He also noted that RPA is usually not the best solution for more complex integrations.)
It's just like what I tell my customers is that depending on what you're using RPA for, if you're going to use it like you would a macro in Excel, it can be great, and it can really empower the people that are using it. If you're going to use it for integrating 2 systems together, then a better solution is to build an integration between the 2 systems than to code a robot to make clicks on one screen and copy things over. Just from a point of failure standpoint, you introduce more points of failure that way than by just having a program do it...
When executed effectively, RPA provides cost and time savings by reducing the need for human labor. It also puts the power of automation in the hands of business users, with minimal involvement required from IT.
RPA has also proven popular from a compliance perspective, given that it provides an auditable trail of all activities. From a former UiPath VP:
The other things that has made this [RPA] kind of like the darling of the compliance crowd and the risk officers...is the fact that it leaves an unmistakable audit log of every single transaction that you can't mess with, right?
Within the RPA space, automation can be divided into three categories: attended, unattended, and cognitive. While this sounds confusing, they’re straightforward to distinguish:
- Attended automation occurs when a human worker oversees the actions of software robots, prompting their action. These robots exist to take over tedious tasks of the human worker, almost like junior employees. This has historically been UiPath’s focus.
- Unattended automation happens at the infrastructural level, operating through virtual machines. Bots execute end-to-end processes for a community of users without supervision (or even awareness that tasks are being undertaken). In this respect, unattended automation is similar to IT testing.
- Cognitive automation refers to automation conducted with the help of AI, ML, and NLP. For example, this type of automation is used to process an invoice PDF, extracting and structuring relevant data. As the RPA market has heated up, “cognitive automation” has become something of a buzzword, without a definitive definition.
Today, UiPath’s platform centers on traditional RPA functionality. That isn’t a knock: the company has arguably constructed the most complete solution in the space with a range of products across five distinct functions.
UiPath helps companies discover opportunities for automation. Using “AI-enabled” desktop recording and process mining, the platform can help companies identify new flows that should be turned over to bots. UiPath’s process mining capabilities are primarily thanks to the acquisition of Process Gold.
Users can create either unattended or attended automations in UiPath’s low-code builders. There’s also a marketplace through which users can leverage bots built by others, with over 1,200 assets available.
UiPath’s platform allows you to track automation, orchestrate different bots, and ensure deployment happens within a suitable governance framework.
Bots can be deployed across the enterprise, executing automations triggered by an API, event, scheduler, or something else.
UiPath gives users multiple ways to interact with its robots. There seems to be functional overlap with “Manage,” though the company calls out its “Assistant,” which allows the user to set reminders for automations, and “Chatbots” features, among others.
UiPath handles all three types of automation (unattended, attended, cognitive) across these five functions, speaking once again to the completeness of the offering. Beyond this core functionality, UiPath has made moves to protect its future.
To fend off bottom-up disruption, UiPath has rolled out features made for the AI and cloud era. These overlap with cognitive automation and extend beyond it. Notable, these new initiatives fit the five core functions listed above, but are worth highlighting given their importance to the company’s future.
A multi-tenant SaaS product that allows for automation without expensive hardware or infrastructure. As with other cloud software, updates are delivered automatically. This will be a key product as more companies migrate to the cloud.
This is the focal point for UiPath’s AI efforts, giving customers the ability to upload their own ML models or use models created by other organizations. As with UiPath’s other products, this power is delivered via an intuitive interface, allowing models to be dropped into existing flows.
A single source of data for the organization, Data Source is also cloud-based. Relevant data is easy to manipulate with low-code tools.
UiPath purports to leverage “proprietary computer vision technology” and optical character recognition (OCR) to extract and structure data from documents. This is a nice example of cognitive automation.
Investors will want to keep a close eye on the evolution of these products in the coming years. As UiPath admits in the S-1:
We have offered cloud-based products for only a short period of time and cannot predict how increased adoption of cloud-based products will change the buying patterns of customers or impact future ARR or revenue.
Like few other businesses, UiPath has succeeded in positioning itself as the “best of both worlds.” Management has shown an uncanny and unique ability to fight on two fronts: situating the company as the best bet for IT departments and citizen developers. That’s no small feat — it’s rare to find a product robust enough for technical users and intuitive enough for laypeople.
Key features that bolstered this dual-appeal:
- Easy build interface. UiPath made it easy for non-developers to create bots, offering a slick designer with accessible themes, layouts, configurable fields, and other out-of-the-box features. Even if developers ended up using the platform more, this positioning allowed clients to believe that integration would be relatively simple and broadly applicable.
- Solid IDE. Developers can use a slick “Integrated Development Environment,” boasting reasonable version control and testing, and endemic integrations with Git/SVN/TFS. Technical users have the tools to get the most out of the platform and use it on their terms.
- Deep catalog. UiPath has created an accumulating advantage by consistently adding to a library of pre-built routines and activities with good configurability. Users can jumpstart their deployment by leveraging these existing assets.
- Computer vision for surface automation. Though there might be others with more native AI capabilities, UiPath has nevertheless done an impressive job of integrating its offering. For example, the company’s desktop recorder is an easy way to bootstrap the creation of new bots.
- Academy and free-tier to reduce the cost of learning. An RPA integration expert noted that though trained on Blue Prism, Automation Anywhere, and UiPath, he continually gravitated towards UiPath. Why? “I'm probably most proficient with UiPath since I am able to learn -- with me having my own business, I can't afford to pay licenses to all of the different vendors. So UiPath is just very attractive because they offer it for free for learning purposes.”
Though UiPath’s product offering is complex, its differentiation is simple: succeeding in being the tool of choice, regardless of a user’s technical ability.
Before turning to the individual revenue streams, it’s worth noting how UiPath sells, doing so in three ways: direct to customer, with a partner, or to a partner.
According to a former sales executive from the company, selling directly to the end customer typically takes six months, regardless of customer size.
So from a timeline perspective, an average sales cycle is about 6 months, right? This has nothing to do with the size of the implementation. This just has to do with the education, the POCs, I want to touch it and feel it before I buy it kind of situation.
This timeline is accelerated when UiPath engages a partner, typically a BPO or consultancy. UiPath sometimes sells with one of these partners, being brought in to propose their offering to an end client. An example from the same executive:
So Deloitte says, "Hey, JPMorgan Chase, I know this use case is going to come up in your digital transformation. Let me introduce you to Automation Anywhere or UiPath," right? And why don't all 3 of us talk and we can make something happen here. So that's a sell with, right. Those tend to be quicker because the partner is already in there. They already have the relationships. They already have the paperwork, so on and so forth.
A sell to also tends to take less time (3 to 5 months), given that there’s less convincing and education required. In this scenario, the partner purchases UiPath’s technology to subsequently white-label and resell to customers. In addition to the shorter sales cycles, partnerships also reduce the number of salespeople UiPath needs. This, at least in part, explains why the company has grown so effectively leveraging a B2B2B playbook.
UiPath uses all three approaches to bring in licensing, implementation, and maintenance revenue.
UiPath predominantly monetizes through software licenses. For the fiscal year of 2021, licensing contributed 57% of the company’s total revenue.
Within licenses, pricing is typically defined by the number of “bots” used. Though different providers define this unit differently, the notion of a bot succeeded in communicating what enterprises were buying, becoming commonplace nomenclature.
Effectively, a bot is a discrete “virtual worker” with defined capacity. Once a bot is fully deployed and at 100% capacity, buyers purchase or activate another license, accessing new bots. In that respect, bots are a proxy for usage and grant enterprises the ability to scale flexibly. This is a vital part of UiPath’s motion: upselling. As we’ll discuss in “Financial highlights,” the company has done a remarkable job of “landing and expanding.”
While bot licensing represents UiPath’s core model, there is greater complexity under the hood. UiPath also sells licenses for its Studio (drag and drop RPA creation) and Orchestrator (RPA management and measurement) products.
Though robust, UiPath may worry about the durability of this business model. As competition in the RPA space has heated up, more players have approached product parity with UiPath, resulting in negative pricing pressure. One executive-level client of UiPath’s noted as much in an interview:
[Why are the vendors reducing their prices?] I think for a couple of reasons. One, there are more players emerging, trying to compete with each other.
On the subject of pricing pressure: what’s the only thing worse than a competitor offering a similar product for a lower cost? A competitor offering a similar product for free.
In March, Microsoft made its “Power Automate Desktop” free to all users with Windows 10, essentially offering a good portion of UiPath’s value at no cost. Students of Slack will recognize the significance of such a move. By improving its Teams solution and bundling it into Office 365, Microsoft sniped Slack’s market, capping growth. UiPath will fear Microsoft using its formidable distribution and default power to similar effect.
UiPath’s product has been designed to be easy to implement. Much of its success is down to a remarkable knack for simplifying complex development, particularly relative to the market’s first mover, Blue Prism.
As mentioned under “Product,” this has proved to be powerful positioning, no doubt contributing to UiPath’s growth. UiPath makes it easy for companies to get up and running, recognizing that it will be able to multiply LTV over time if successful.
It’s perhaps no surprise then that the company earns little here with just 5% of revenue from the last fiscal year coming from implementation.
Maintenance and support
Given that UiPath makes so little from implementation, it seems almost paradoxical that it should earn nearly 40% of its revenue from maintenance and support.
That begins to make more sense once one considers the relative newness of the RPA space and the velocity with which it has advanced. With the industry relatively inchoate, the past 5-6 years saw significant advancements, with RPA vendors regularly shipping updates to customers. That has involved considerable maintenance and support.
The last couple of years have seen platforms stabilize and innovation around pure “1.0 RPA” falter. (Secondary evidence of that is UiPath’s declining investment in R&D; down 30% in 2021). That may cause UiPath problems, with customers questioning the value of high maintenance fees when support is no longer delivering the next game-changing update.
Looked at holistically, UiPath’s business model appears to have been ideally suited for its time. Fortune 500 companies would likely have been loath to undertake an automation revolution without the assistance and service of BPOs and consultancies, a channel that UiPath capitalized upon. Perfecting the art of getting the foot in the door, UiPath subsequently nailed the second part of the equation: upselling customers and growing LTV over time. The pace of development justified maintenance costs.
But the market seems to be shifting, posing a threat to this way of doing business. Investors will want to be confident UiPath can move into this new era on the front foot. Otherwise, it may see its top two revenue engines splutter.
While co-founder Daniel Dines sits at the top of the UiPath org chart as CEO, recent additions to the management team reflect the organization’s mission to deeply infiltrate top global enterprises. Leaders come from sales-driven enterprise software and IT behemoths like SAP, HPE, and VMWare.
Key personnel includes:
- Marius Tirca, co-founder and Chief Technology Officer: Tirca has been with Dines since the beginning. Interestingly, his LinkedIn notes he started as a software engineer in 2004, leveling up to attain CTO and “co-founder” status after the transition to pure RPA in 2011. Strangely, Tirca is entirely missing from the S-1.
- Ashim Gupta, Chief Financial Officer: Gupta transitioned internally from Chief Customer Success Officer to CFO in November 2019. Before his time at UiPath, Gupta was a veteran at General Electric for almost two decades. He originally championed UiPath and RPA at GE, with Dines bringing him over in 2018.
- Thomas Hansen, Chief Revenue Officer: Hansen is a relatively new addition to the leadership team, joining in April 2020. At UiPath, he is responsible for worldwide sales, supported by extensive experience at VMWare Carbon Black, Dropbox, and Microsoft.
- Ted Kummert, EVP of Product and Engineering: Kummert is another recent hire, joining UiPath in 2020 after a long career, including 24 years at Microsoft and four years as EVP of Product and Engineering at Apptio.
- Bobby Patrick, Chief Marketing Officer: Patrick brings a wealth of enterprise marketing experience. He previously served as CMO of Hewlett Packard.
Reflecting their seniority, Hansen and Kummert are the two highest compensated members of the Executive Team, with Dines earning a token compensation package for 2021:
Outside of the executive team, the board consists of select venture partners. With its 28% stake, Accel has two directors: Philippe Botteri and Rich Wong. Botteri replaced longtime Accel partner Luciana Lixandru (who, you’ll remember, spearheaded the firm’s initial investment) after she departed for Sequoia. Speaking of Sequoia, the firm is represented by Carl Eschenbach, one of the world’s most experienced and respected enterprise investors. Filling out the roster is Laela Sturdy, General Partner at CapitalG, riding a remarkable record with ten unicorns under her belt.
Despite the loaded roster, Dines has done something quite remarkable for an IPO’ing unicorn: concentrate voting control. While we’re used to founders using a dual-class system to ensure they retain power, Dines has taken the mechanism to its inevitable, absurdist conclusion, holding the entirety of UiPath’s Class B shares, which have 35x the weight of Class A shares. The result is that Dines holds 88% voting power. Including individual board members, Accel has 6.2%, the next largest.
Given UiPath’s success, many investors may have no issue with this set-up, trusting Dines to steer the ship. Others may worry the structure stifles discussion, depriving dissenting voices of oxygen.
That authoritative approach may speak to Dines’s management style. A 2019 Forbes profile portrayed the UiPath chief as a gregarious, mercurial figure operating by his own rules. Dines described himself as a “lazy person” who starts work at 11 am when in UiPath’s Romanian office, only after a morning nap and an hour of reading. Though UiPath’s origins embody the company’s commitment to “humility,” Dines does not always seem to walk the talk. From the profile:
“Humility” is one of UiPath’s four main tenets, akin to Google’s early maxim “Don’t be evil,” and it’s practiced just as inconsistently. At lunch with Dines in Bucharest, the founder says he doesn’t consider himself a great coder, just a very good one. By dinnertime, he says he’s still better than anyone at UiPath.
Of course, others paint a different picture. One early employee and former Director gave Dines arguably the greatest compliment a CEO can receive, comparing UiPath’s founder to two visionaries:
Daniel reminded me so much of what people talked about when they would describe Bill Gates. I mean the guy, he is so smart.
And he came up in Romania, he’s self-taught as a developer, and he got recruited by Microsoft to come to Seattle to be a software engineer... And he really knows his stuff…[H]e could also call BS on developers because he can go right down to the code lines if anybody is trying to tell him that something can't be done. So he's kind of a little like Elon Musk in that sense.
As long as Dines delivers growth, contradictions and quirks are unlikely to concern investors; they remain worth keeping an eye on.
Although UiPath went nearly ten years between its founding and first major institutional round, it has attracted many of the most storied names in venture to its cap table. In the last three years alone, it welcomed checks from Sequoia, Meritech, IVP, Dragoneer, Altimeter, and Tiger Global.
As alluded to above, Accel is UiPath’s largest external shareholder with 26%. The blue-chip firm led UiPath’s $30M Series A in 2017, doing the same with the company’s $153M Series B. At February’s reported $35B valuation, Accel’s stake is worth more than $9B — a 4.5x return on its 2016 vintage $2B growth fund.
They’re followed by Earlybird, the European firm first to back Dines. Earlybird protected its stake by participating through the Series B, reaching IPO with 10.3%. That commitment is set to secure the firm its biggest ever win. Earlybird’s stake is worth roughly $3.6B — an 18x return on its €150M 2013 vintage (~$200M) based on the last round.
UiPath’s largest strategic investors are Tencent and Google (via CapitalG), with the latter owning 7.5%. That makes CapitalG UiPath’s third largest outside shareholder.
Below you can see a summary table of UiPath’s financials. Their fiscal year ends in January. As such, FY 2020 can be used as a proxy for calendar year 2019: the 12 months ending January 2020 are nearly the same as the 12 months ending December 1.
We’ll dig into the most relevant details below, highlighting UiPath’s growth, margin profile, operating expenses, net expansion, and more.
UiPath has an incredible growth profile, expanding revenue 81% over the last 12 months. This trajectory has been impressively consistent, showing no signs of deceleration.
When looking at the broader universe of public cloud companies (a useful if imperfect analogy), UiPath compares favorably. Over the past twelve months, UiPath is growing sixth-fastest.
One knock on the company’s growth story comes from reviewing the last couple of quarters of ARR YoY expansion. Generally, ARR growth is a better proxy than revenue growth for evaluating business growth. (Are you sick of the word “growth” yet? Hang tight.)
Unlike revenue YoY growth, ARR YoY growth did decelerate, dropping from the mid-80%s to the mid-60%s. Because of the way UiPath recognizes revenue, this deceleration doesn’t show up in the revenue growth line, but it’s certainly something to monitor closely.
GAAP Gross Margins
In late 2019, UiPath was forced to make significant layoffs. In response to that turnover, Bobby Patrick fended off questions about whether UiPath was a WeWork-in-waiting; another high-burn business struggling to make the math work. Patrick responded, “We are a software company with high gross margins; we are not real estate. We are not remotely similar.”
Looking at UiPath’s gross margins today, the comparison seems as ludicrous as it must have felt to Patrick at the time. Hovering close to 90%, UiPath’s gross margins are extremely impressive and industry-leading. The only cloud businesses with better numbers are Alteryx and Autodesk.
GAAP Operating Margin
UiPath has demonstrated impressive operating leverage. Over the last eight quarters, the company has taken GAAP operating margins from -260% to 7%. Not many public cloud businesses can claim positive GAAP operating margins, favorably marking UiPath out.
UiPath spends quite a bit of money on sales and marketing expenses. That isn’t surprising given the company’s focus on serving large enterprise customers, including 61% of the Fortune Global 500. Over the last 12 months, S&M expenses represent 63% of total revenue, the seventh-highest of all cloud businesses.
There is some good news. First, S&M spend as a percentage of revenue is on the decline, dropping to 48% in the most recent quarter. From 2020 to 2021, this figure fell from 144% to 62%. Second, UiPath’s efficiency isn’t too bad, as shown by the company’s payback period (more below).
Research and Development (R&D) and General and Administrative (G&A) expenses as a percentage of revenue are both on the low end (18% and 27%, respectively), below the median for public cloud businesses. The former is arguably a concern given the technological advancements in AI and other related fields. Investors might have hoped to see a more significant commitment to remaining ahead of the curve.
Dollar-Based Net Retention
A critical SaaS metric, UiPath benchmarks quite well in terms of dollar-based net retention (DBNR). This metric communicates the growth in spend the company captures from existing customers, illuminating the success of a “land and expand” strategy.
At 145%, UiPath has the fifth-highest DBNR out of all cloud businesses, behind Agora, Snowflake, nCino, and On24.
(One might argue Zoom’s DBNR is also higher, but instead of reporting an exact figure, the video-conferencer reports the metric as “>130%.” It’s unclear how far above 130% that means.)
Gross Margin Adjusted CAC Payback
This metric is used to calculate how long (in months) it takes to pay back the cost to acquire a customer with gross profit dollars.
It’s a great way to look at the operating leverage of a business. The lower the number, the better — that indicates it takes less time to pay back that CAC. Anything under 12 months is exceptional; UiPath comes in at ~22 months.
This isn’t particularly impressive, nor is it terrible. The median for all cloud businesses is 21 months.
Rule of 40
A classic of the software industry (particularly SaaS), the “Rule of Forty” seeks to distill growth and profitability into a single metric. It does so by combining revenue growth and free cash flow over the last twelve months. The idea is that companies should exceed 40% when adding those two metrics, with higher numbers considered better.
UiPath’s performance is very strong here, reaching 85%. That places it seventh among public cloud businesses.
Overall, the financial profile of UiPath is quite impressive. The growth is staggering, and margins are scaling nicely. Investors will want to keep an eye on S&M spend, ARR growth, and gross margin adjusted CAC payback over the coming quarters.
UiPath faces competition from multiple angles. We’ll discuss the different players below, but first, it’s worth outlining its most severe threat: the changing automation landscape.
As we noted earlier, UiPath was the perfect company for its era, offering a strong rules-based automation platform. But the market is changing. The last few years have seen a proliferation of entrants, each bringing a unique spin to the space. That’s led to niche offerings designed to appeal to specific industries. Whereas customers might have previously chosen a platform, increasingly they are choosing a platform with pre-built use cases. A bank does not pick Automation Anywhere; it chooses Automation Anywhere’s financially-tailored functionality.
With an intentionally broad product, how will UiPath adapt to a more selective customer base? Or to put it in other words: can the dominant horizontal player go vertical?
This critical question influences how one perceives UiPath relative to four competitive sets:
- The Big Three
- Rising players
- Horizontal giants
- Bottom-up insurgents
Let’s dig into each.
The Big Three
UiPath competes with different categories of rivals within the enterprise space. Most directly, UiPath vies with Blue Prism and Automation Anywhere. As summarized by a former UiPath VP:
These are kind of what we call the big boys of the industry... [T]hat is because of their global reach and their clientele list...and the fact they're just massive from an RPA perspective and spread into every single vertical that you can think of.
With roughly equivalent offerings (all are horizontal) and similar reach, the trio often fights for the same customer base. Though an early mover, Blue Prism seems to have fallen off relative to UiPath and Automation Anywhere, often lagging in product development. Critically, some experts consider Automation Anywhere to be the most “cloud-ready” of the group.
While currently the dominant players in the space, all three risk the comparative degradation of their core technology, and the risk of verticalized offerings. To stave that off, all have made forward-looking acquisitions in recent years. These acquisitions fall primarily into two buckets: API ecosystem expansion (Cloud Elements acquired by UiPath in 2021; Thoughtonomy acquired by Blue Prism in 2019) and deeper process automation (StepShot and ProcessGold acquired by UiPath in 2019; Klevops acquired by Automation Anywhere in 2019).
Notably, customers do sometimes use more than one solution. From a former UiPath executive:
It's also very easy for very large companies to be using both [UiPath and Automation Anywhere] or [all] 3...So Finance knows some guy at Automation Anywhere...he rusn a pilot in his silo, was successful and continues to grow. HR goes with UiPath... [W]e've seen it happen...global companies are using Automation Anywhere in India and UiPath in the U.S.
Outside of this “Big Three,” UiPath also competes with companies like WorkFusion, Kofax, Antworks, and others. Though these players don’t yet have the same range as the Tier 1 behemoths, each has succeeded in carving out a place in the market, often around a particular technology. Again, referring to a former VP:
[I]f you take the example of WorkFusion and Kofax, they were very kind of OCR based automation and then they kind of beefed up their RPA game.
But they're tier 2 for a reason...they are either very strong in a couple of geos, are very strong in a couple of industries or more than a couple of industries, but not as big and don't have the sales force. They don't have the partner relationships, or the cloud that a UiPath or Automation Anywhere has. And even from a product portfolio perspective, they're far behind in terms of capability.
This class of competition threatens UiPath by selling specific use-cases and aggressively targeting specific customers.
Lastly, within the enterprise segment, even more verticalized offerings fight for UiPath’s client base. These businesses explicitly focus on specific industries, providing re-packaged, templated use cases. For example, Olive AI has made a name for itself in the healthcare space.
Perhaps the most annoying question a founder can ever face: “Why can’t Google do this?”
UiPath won’t fear Google’s entrance with the search giant on the cap table. Daniel Dines might, however, spend nights worrying about Microsoft. Though a late-mover, Microsoft has built a credible RPA platform, particularly since it acquired Softomotive for $60 million. The Seattle megalodon does not yet have as complete an offering as UiPath, but what it lacks in comprehensiveness, it makes up for in price, with access offered for free with an Office 365 license. When asked why clients might change RPA solutions, an integration consultant answered as follows:
I mean, I think really the only element would probably be price...I could see a company saying, "Hey, we want to switch from UiPath or Automation Anywhere to Power Automate just because we moved to Office 365, and now it's included and it's free. So we could entirely eliminate the licensing cost of a Blue Prism or a UiPath if we just port this over to Power Automate." The question then would be, is Power Automate robust enough to be able to do the same things that you were doing in that regard?
With more time and investment, could Microsoft reach feature parity with UiPath and undercut it on price? Other industry spectators believe the horse has bolted. When asked if Microsoft had missed the bus, a senior UiPath client said:
Yes, too late to the game. Too late for the game is the short answer, I would say...Microsoft Flow was always just focused on making things easier for people that are already using Microsoft products, but it wasn't really thinking horizontally like how would you integrate with other kinds of solutions.
And RPA, obviously, was a big missing component so they acquired Softomotive, but that wasn't a strong performer or a leader. I mean, it was a pretty small player maybe, but I think maybe Microsoft could inject it with some excellence and development. But I think it's a little bit too late... It's very difficult to catch up.
With UiPath and Automation Anywhere out of reach from a valuation perspective, will Microsoft look for further acquisitions down market?
Amazon also cannot be ruled out, though it might make a more natural partner for an existing player.
UiPath focuses on companies with 200+ employees. Yet, process automation is needed everywhere, from early startups to the Fortune 50. As UiPath dedicates itself to the top-end of the market, insurgent product-led companies are starting at the bottom. The risk for UiPath is that, in time, those small fish may become big enterprises with legacy integrations.
The company will hope to develop a similar positioning as Salesforce. Though $CRM is seldom the solution with which companies start, they tend to matriculate after reaching a certain scale. Can UiPath manage something similar? Or will the depth and complexity of RPA integrations make that conversion considerably tricker?
Just as companies like Airtable, Clay, Orbit have found vectors of attack against Salesforce, so too, will UiPath likely have to fend off advances from startups like Alloy Automation, KissFlow, Bryter and others, all operating with a different GTM playbook.
There is no doubt UiPath has created an incredible machine.
There is also no doubt the IPO market is showing signs of cooling. Multiples of high-growth tech companies have fallen from 2020’s lofty heights, seemingly altering UiPath’s valuation trajectory along the way.
As noted previously, UiPath raised $750 million in February, valuing the business at $35 billion. Yet, in its amended S-1 filing, UiPath disclosed that it expects the IPO to price between $43 and $50 per share, a $22-26 billion range. Of course, such pricing is preliminary, and there is a (perhaps likely) chance of $PATH debuting north of that range. But even at the high end of that range, $26 billion would represent a 25% discount from Series F.
So, how should investors think of UiPath’s more modest valuation? Are public investors getting a steal?
Like so many other venture-backed businesses, UiPath is not profitable. The company lost $520 million in FY2020 before narrowing the deficit to $90 million in FY2021. Of course, UiPath will argue that this spending above its means is simply a result of its “foot in the door” strategy: costs are incurred upfront, but effective upselling brings in more revenue per customer over time. UiPath would suggest opex will be meaningfully lower as a percentage of revenue moving forward.
While the company’s revenue growth, margins, retention, and expansion are all awe-inspiring, it’s also worth noting that UiPath has likely been a beneficiary of the coronavirus. The pandemic provided management teams an opportunity to cut headcount and execute RPA strategies with minimal scrutiny. Is automation becoming an enduring priority for the corporate world, or should we expect this ardor to dwindle once life returns to normal?
On a longer time horizon, investors might worry about the company’s vulnerability to disruption from low-code, cloud-native competitors that don’t have to rely on such heavy S&M spend. UiPath’s March acquisition of API/cloud specialist Cloud Elements is tacit recognition of its susceptibility. If the company wants to compound over time, it will need to head off this threat.
Where does that leave us?
Even at $26 billion, UiPath looks pricey with a 43x LTM sales multiple. That puts $PATH in the company of Zoom (48x), but miles ahead of mature enterprise SaaS giants like Salesforce (7x), Microsoft (6x), Oracle (5x).
UiPath has done a stellar job penetrating its market from a franchise perspective. It offers a valuable product with high-network effects, but ultimately, the wisdom of valuation will depend on top-line growth and the company’s ability to grow DBNR.
If UiPath maintains its pace, $26 billion might look like an excellent deal in hindsight.
Digital transformation is a hot topic for every enterprise in the world. UiPath’s growth and DBNR indicate that it grasped the generational opportunity presented by covid, further insinuating itself into the world’s most prominent organizations. Even as more technologically advanced solutions come to market, replicating this level of integration takes time, money, and effort.
Now is an opportune time for UiPath to deepen its moat with capital. The listing will allow the company to be more active from an M&A perspective, granting the opportunity for the company to extend its capabilities through acquisitions. As the former UiPath VP stated with regards to M&A:
[Consolidation?] It’s going to happen. The way it's carrying on now is not sustainable because...for the number of [sales] deals that we step into every month, right, 80%, 90% of the deals are UiPath, Automation Anywhere. UiPath -- I don't even hear of the others, right? So it's going to happen.
New capital will also provide firepower for UiPath to chase the secondary markets it seems to covet, including the no-code space.
For a success story sixteen years in the making, going public seems like the end of one blockbuster, but just the start of UiPath’s saga. After all, augmenting humanity is a mission that cannot be automated.
The Generalist’s work is provided for informational purposes only and should not be construed as legal, business, investment, or tax advice. You should always do your own research and consult advisors on these subjects. Our work may feature entities in which Generalist Capital, LLC or the author has invested.