Microsoft tones its HPC cloud with Cycle Computing

Contact: Scott Denne, Csilla Zsigri

In an effort to extend Azure into a potentially lucrative corner of the cloud market, Microsoft picks up Cycle Computing, a company that enables HPC applications in multi-cloud environments. Microsoft’s move fits with a larger trend of cloud providers building and buying software assets to attract those applications with the most appetite for compute and storage.

Cycle Computing has more than a decade of experience orchestrating, provisioning and managing HPC and other intensive computing applications across multiple environments. First developed to take advantage of grid computing, it has more recently launched CycleCloud, and joined Microsoft’s Accelerator program in 2016.

HPC is about three to five years behind enterprise computing when it comes to new technology adoption – the applications are generally more sophisticated, and engineers are conservative. Yet the HPC cloud market is accelerating, and compute- and data-intensive applications in areas such as big data, machine learning, deep learning and IoT are also moving to the cloud. We believe that Microsoft is taking advantage of these trends and is looking to use Cycle Computing’s technology to enhance Azure’s current data-processing capabilities and build virtual supercomputers in the public cloud.

By investing in HPC and other data and analytics applications, Microsoft makes Azure fertile soil for new workloads. According to 451 Research’s Voice of the Enterprise Cloud Transformation survey, 21% of data and analytics workloads will move to public clouds in the next two years, a larger share than any category excepting web and media deployments, which, not coincidently, is where Amazon has focused its recent M&A with acquisitions of Thinkbox Software and Elemental Technologies.

Moreover, that same survey showed that IT departments have a greater threshold for price increases for mission-critical data analytics workloads. Almost half (44%) said they would be willing to pay an additional 26-50% to ensure quality of service, compared with just 30% who would pay such an increase for web workloads.

For more real-time information on tech M&A, follow us on Twitter @451TechMnA.

Don’t bet against Bezos

Contact: Brenon Daly

A day after Amazon’s Jeff Bezos put out an open-ended tweet to the world asking where he should donate his money, we now know at least one early recipient of his philanthropy: Whole Foods Market. OK, the $13.7bn acquisition of the grocer isn’t exactly charity, but nor is it an example of a hardened dollars-and-cents M&A strategy.

Instead, it might be most accurate to think of the Amazon-Whole Foods pairing as a blend of giving and buying, a deal that’s being attempted by one of the few CEOs who could possibly get away with spending billions of dollars of shareholder money to effectively take his company backward in time. For lack of a better term, think of Bezos’ move as a ‘patronage purchase.’

Whole Foods, which was being stung by a gadfly hedge fund, needed a buyer for the 430-store chain. (From our side, we were half expecting the grocery chain’s CEO, John Mackey, to try a Kickstarter-funded management buyout.) Amazon — or more accurately, Bezos — is convinced that the world’s largest online retailer needs a brick-and-mortar presence.

Undoubtedly, there’s a certain logic to building up the distribution network for physical goods, which account for the bulk of Amazon’s revenue. However, those sales aren’t particularly attractive, at least economically. To put some numbers on that, consider the operations for Whole Foods, a real-world business that Amazon is buying, compared with AWS, a cloud business that Amazon has built. Conveniently, both businesses generated roughly the same amount of revenue in the most recent quarter, $3.7bn. Leave aside the fact that AWS grew 43% while Whole Foods flatlined and just look at the operating margin: Whole Foods posted just $171m of operating income, only one-fifth the $890m that AWS generated.

Conventional corporate strategy would typically encourage a company to allocate resources to the business with the highest return (AWS), rather than spending billions of dollars to buy its way into a low-growth, low-margin adjacent market. But then, Bezos has never been conventional.

Historians will remember that Bezos pushed ahead with a $1.25bn convertible note offering for Amazon in 1999. At the time, the deal — the largest-ever convertible by a tech vendor — flew in the face of conventional corporate finance, giving those investors bearish on the money-burning company even more reason to mock ‘Amazon dot bomb.’ However, given that those notes converted at a price of $156 each, compared with the current market price for Amazon shares of nearly $1,000 each, it’s fair to say that Bezos has created a certain amount of goodwill on Wall Street. (Investors gave him the benefit of the doubt on the Whole Foods pickup, nudging Amazon shares slightly higher after the announcement.)

Similarly, by all accounts, Bezos’ purchase of the existentially threatened The Washington Post in 2013 has brought renewed growth to that stalwart newspaper. And while that $250m acquisition was done from Bezos’ own pocket (rather than Amazon’s treasury), it actually lines up fairly closely with the proposed reach for Whole Foods. Both groceries and newspapers represent once-thriving industries that have been decimated by a combination of technology and shifting consumer consumption patterns. In contrast, Amazon has built a half-trillion-dollar market cap on both of those trends, making it hard to bet against Bezos.

For more real-time information on tech M&A, follow us on Twitter @451TechMnA.

Onapsis on the block?

Contact: Brenon Daly

Enterprise application security startup Onapsis quietly kicked off a sale process about a month ago, according to our understanding. Several sources have indicated that Onapsis, which focuses on hardening security for SAP implementations, has hired UBS to gauge interest among buyers. And while there undoubtedly will be acquisition interest in the startup, Onapsis may ultimately prove to be a bit of a tough sell. The reason? The most obvious buyers for the company don’t typically pay the type of valuations that Onapsis is thought to be asking.

In many cases, the heavy-duty SAP systems that Onapsis helps secure were implemented by one of the big consulting shops. So at least theoretically, it’s not a big leap to imagine one of these consultancies buying Onapsis and offering its platform, exclusively, to help safeguard these mission-critical systems and the data they generate. (Indeed, Onapsis already has partnerships with many of the big consulting firms, including KPMG, PWC, Accenture and others.) While that strategy may be sound, M&A always comes down to pricing. And that’s why we would think it’s probably more likely than not that eight-year-old Onapsis remains independent.

According to our understanding, Onapsis is looking to sell for roughly $200m, which would be twice the valuation of its September 2015 funding. The rumored ask works out to about 8x bookings in 2016 and 4.5x forecast bookings for this year. For a fast-growing SaaS startup, those aren’t particularly exorbitant multiples. Yet they may well price out any consulting shops, which have typically either picked up small pieces of specific infosec technology or just gobbled up security consultants. Any reach for Onapsis would require a consulting firm to pay a significantly richer price than the ‘tool’ or ‘body’ deals they have historically done.

Xactly exits

Contact: Brenon Daly

Two years after coming public, Xactly is headed private in a $564m buyout by Vista Equity Partners. The deal values shares of the sales compensation management vendor at nearly their highest-ever level, roughly twice the price at which Xactly sold them during its IPO. According to terms, Vista will pay $15.65 for each share of Xactly.

Xactly’s exit from Wall Street comes after a decidedly mixed run as a small-cap company. For the first year after its IPO, the stock struggled to gain much attention from investors. Shares lingered around their offer price, underperforming the market and, more notably, lagging the performance of direct rival Callidus Software. However, in the past year, as Xactly has posted solid mid-20% revenue growth, it gained some favor back on Wall Street. In the end, Vista is paying slightly more than 5x trailing sales for Xactly.

The valuation Vista is paying for Xactly offers an illuminating contrast to Callidus, which has pursued a much different strategy than Xactly. Although both companies got their start offering software to help businesses manage sales incentives, the much-older and much-larger Callidus has used a series of small acquisitions to expand into other areas of enterprise software, notably applications for various aspects of human resources and marketing automation. According to 451 Research’s M&A KnowledgeBase, Callidus has done seven small purchases since the start of 2014. For its part, Xactly has only bought one company in its history, the 2009 consolidation of rival Centive that essentially kept it in its existing market.

Although Xactly is getting a solid valuation in the proposed take-private, it’s worth noting that Callidus – at least partly due to its steady use of M&A – enjoys a premium to its younger rival with a narrower product portfolio. Even without any acquisition premium, Callidus trades at about 7x trailing sales. Callidus is roughly twice as big as Xactly, but has a market value that’s three times larger.

For more real-time information on tech M&A, follow us on Twitter @451TechMnA.

No ray of sunshine from Cloudera IPO

Contact: Brenon Daly

As far as Wall Street is concerned, the outlook for the tech IPO market is still cloudy after Cloudera’s offering. Sure, the data analytics platform vendor priced shares higher than its underwriters expected and investors pushed the freshly minted stock about 20% higher in aftermarket trading on Friday. But that solid start isn’t likely to necessarily draw other startups to the public market because Cloudera’s capital structure got so uniquely inflated.

Few startups could even imagine – much less collect – an investment of three-quarters of a billion dollars from a single investor in a single round, as Cloudera did from Intel three years ago. The chipmaker paid up for the privilege, putting a ‘quadra unicorn’ valuation of $4.1bn on Cloudera. Altogether, Cloudera raised more than $1bn from private market investors, making the $225m raised from public market investors seem almost like lunch money.

And then there’s the small matter of valuation. In its debut, Cloudera is only worth about half of what Intel thought it was worth when it made its bet. (As we noted in our full preview of Cloudera’s IPO, Intel’s investment appears even more bubbly when we consider that, at the time, Cloudera was generating less than half the quarterly revenue it currently puts up and its operating loss actually topped its revenue.)

As a longtime corporate investor, Intel can chalk up the overpayment for the stake of Cloudera to ‘strategic’ considerations. (Much like the chipmaker effectively wrote off its massive bet on security, unwinding half of its underperforming acquisition of McAfee at roughly half the valuation it initially paid in the largest infosec transaction in history, according to 451 Research’s M&A KnowledgeBase.) Besides, Intel can afford it: the day that Cloudera priced its IPO – thus confirming Intel’s overpayment – the chipmaker reported that it earned $3bn in the first quarter of this year alone.

http://www.the451group.com/images/upload/4-28-17_kbi_pic.png

Okta’s growth-story IPO finds an audience on Wall Street

Contact: Brenon Daly 

The unicorn parade on Wall Street continued Friday as security vendor Okta nearly doubled its private market valuation in its debut on the Nasdaq. The subscription-based identity and access management provider initially sold shares at $17 each, but investors bid them to about $24 in midday trading. With the surge, Okta is valued at some $2.4bn. (See our full preview of the offering.)

Okta becomes the third enterprise IT startup to come public so far this year, and it extends the strong performance of these new issues. It also joins the two previous IPOs – MuleSoft and Alteryx – in sporting a rather stretched valuation. Based on a market cap of $2.4bn, Okta is trading at about 15x trailing sales.

Granted, Okta’s sales are growing quickly, having nearly quadrupled in just the past two fiscal years to $160m. Still, the company is commanding quite a premium compared with fellow secure identity specialist CyberArk, which also just happens to be the last information security startup to create more than $1bn of value in its IPO. (To be clear, CyberArk, which went public in 2014, also sells identity-related products in the form of privileged identity management, but doesn’t really compete with Okta.)

Wall Street currently values CyberArk at about 8.2x trailing sales, or just slightly more than half the level that investors are handing to the freshly public Okta. Bulls would argue that Okta merits the premium given that it is growing twice as fast as CyberArk. But others might counter with a question about what that growth is costing each of the companies. Okta lost a mountainous $83m on its way to generating $160m in sales last year. In contrast, CyberArk, which has run in the black for the past four years, netted $28m from its 2016 revenue of $217m.

If nothing else, the valuation discrepancy underscores that growth is still the key metric for investors. Okta’s IPO is simply supply meeting demand, same as it ever was on Wall Street. Indeed, CyberArk has also experienced that. Shares of the company reached an all-time high – nearly 50% higher than current levels, roughly Okta’s current valuation – in 2015, when revenue was increasing north of 50%, compared with the mid-30% level now.

For more real-time information on tech M&A, follow us on Twitter @451TechMnA.

An earthbound IPO for Cloudera

Contact: Brenon Daly 

Looking to extend the current bull run for enterprise software IPOs, Cloudera has taken the wraps off its prospectus and put itself on track to hit Wall Street in about a month. Assuming the debut follows that schedule, the heavily funded Hadoop vendor would be the third infrastructure software provider to come public in six weeks, following MuleSoft and Alteryx. Unlike the debuts of those two other software firms, however, Cloudera’s IPO will almost certainly be a down round.

Three years ago, when Cloudera’s quarterly revenue was less than half its current level, Intel acquired 22% of the company at a valuation of $4.1bn. Since then, both the company and other equity holders agreed that ‘quadra-unicorn’ valuation got a little ahead of itself and have priced Cloudera shares below Intel’s level of just less than $31 each. (In contrast, MuleSoft has more than doubled its final private market valuation on Wall Street.) Cloudera – along with its nine underwriters, led by Morgan Stanley, J.P. Morgan Securities and Allen & Co – should set the inaugural public market price for shares in about a month.

Because Wall Street likes to use a ‘known’ to help assign value to an ‘unknown,’ investors will look at Cloudera’s future trading valuation relative to the current trading valuation of fellow Hadoop provider Hortonworks. However, that comparison won’t particularly help Cloudera get any closer to its previous platinum valuation. Hortonworks currently has a market capitalization of just $650m, or 3.5x its 2016 revenue and 2.7x its forecast revenue for 2017.

The two Hadoop-focused companies actually line up fairly closely with one another, financially. Cloudera and Hortonworks hemorrhage money, largely because of huge outlays on sales and marketing. (Both firms spend roughly twice as much on sales and marketing as they do on R&D.) Cloudera is nearly one-third bigger than Hortonworks, recording $261m in sales in its most recent fiscal year compared with $184m for Hortonworks. Both are growing at about 50%.

Within that revenue, both Cloudera and Hortonworks wrap a not-insignificant amount of professional services around their product, which weighs on their margins and, consequently, their valuations. Both are consciously shifting their revenue mix. Cloudera is further along in moving toward a ‘product’ company, with professional services accounting for 23% of revenue in its latest fiscal year compared with 32% for Hortonworks. That progress is also reflected in the fact that Cloudera’s gross margins are several percentage points higher than those at Hortonworks, although both are still low compared with pure software providers. (For instance, MuleSoft, which also has a professional services component, has gross margins in the mid-70% range, about seven percentage points higher than Cloudera.)

With its larger size and more-efficient model, Cloudera will undoubtedly command a premium to Hortonworks. (That will come as a relief to Cloudera because if Wall Street simply valued the company at the same multiple of trailing sales it gives Hortonworks, Cloudera wouldn’t even be a unicorn.) We’re pretty sure Cloudera will come to market with a ‘three-comma’ valuation, but it won’t be near the $4bn valuation Intel slapped on it. Perhaps Cloudera can grow into that one day, but it certainly won’t start out there.

MuleSoft gets a thoroughbred valuation in its IPO

Contact: Brenon Daly 

After a four-month shutout, the enterprise tech IPO market is back open for business. Infrastructure software vendor MuleSoft surged onto the NYSE, more than doubling its private market valuation. It sold 13 million shares at an above-range $17 each, and the stock promptly soared to $24.50 in late-Friday-afternoon trading. That puts the fast-growing company’s market valuation at slightly more than $3bn, twice the $1.5bn value that venture investors put on it.

MuleSoft’s debut valuation puts it in rarified air. Based on an initial market cap of $3.1bn, investors are valuing the company at a stunning 16.5x its trailing sales of $190m. That multiple is twice the level of fellow data-integration specialist Talend, which went public last July. Talend currently trades at a market cap of about $875m, or 8.3x its trailing sales of $106m. The valuation discrepancy indicates that investors are once again putting a premium on growth: MuleSoft is larger than Talend and – more importantly to Wall Street – it is growing nearly twice as fast. (See our full report on the offering as well as MuleSoft’s ‘hybrid integration’ strategy – what it is and where it might take the company in the future.)

The IPO netted MuleSoft $221m, or $206m after fees. That’s undoubtedly a handy amount, but we would note that it is still less than the $260m it raised, collectively, from private market investors. (Somewhat unusually, there are three corporate investors on the company’s cap table.) All of those MuleSoft backers are substantially above water on their investment following the IPO. That bullish debut is likely to draw more high-flying startups to Wall Street after a discouraging 2016, when only two enterprise tech unicorns went public. This year will likely match that number next month, when Okta debuts. The identity management startup revealed its IPO paperwork earlier this week, putting it on track for a mid-April debut.

 

With Okta, infosec no longer conspicuously absent from the IPO market

Contact: Brenon Daly 

Even as several other fast-growing enterprise IT sectors have all seen unicorns gallop onto Wall Street, richly valued information security (infosec) startups have stayed off the IPO track. The sector hasn’t seen a $1bn company created on a US exchange in more than two-and-a-half years. Infosec has been conspicuous by its absence from the tech IPO market, especially considering that no other single segment of the IT market has as many viable public company candidates. Fully one-quarter of the startups in the ‘shadow IPO’ pipeline maintained by 451 Research’s M&A KnowledgeBase Premium come from the infosec space. (See related report.)

At long last, one of the infosec unicorns is (finally) ready to step onto the public market: cloud-based identity management startup Okta has publicly revealed its paperwork for a $100m offering that should price next month. The company, which raised nearly $230m in venture backing, had already achieved a $1bn+ valuation in the private market – and will head north from there in the public market.

Wall Street will undoubtedly find a lot to like in Okta’s prospectus. The company is doubling revenue each year, with virtually all of its sales coming from subscriptions. (Professional services accounts for roughly 10% of total revenue, a lower percentage than most of the big-name SaaS vendors.) Subscription revenue gives a certain predictability to a company’s top line, especially when coupled with the ability to consistently expand those subscriptions. Okta notes in its prospectus that its customer retention rate, on a dollar basis, is slightly more than 120%, an enviable rate for any subscription-based startup. Put it altogether and revenue at Okta for the fiscal year that ended in January is likely to be in the neighborhood of $160m, up from $86 in the previous fiscal year and just $41m in the fiscal year before that.

Having quadrupled revenue in just two years, Okta’s red ink isn’t likely to worry many investors. Through its first three fiscal quarters (ended October 31, 2016), Okta lost $65m, up from $55m in the same period the previous fiscal year. As is often the case with SaaS providers, Okta’s losses stem primarily from heavy spending on sales and marketing. Early on, Okta was spending slightly more than $1 on sales and marketing to bring in $1 of subscription revenue. It has since slowed the spending, with the result that in its latest quarter it spent $32m on sales and marketing to bring in $38m in subscriptions. (For comparison, Box – one of the more egregious spenders – shelled out $47m on sales and marketing to generate exactly the same subscription revenue as Okta ($39m) in its most recent quarter when it originally filed to go public in 2014.)

Okta’s IPO would represent the first new $1bn valuation for an infosec vendor on the NYSE or Nasdaq since CyberArk’s offering in September 2014. Sophos went public (rather quietly) in 2015 on the London Stock Exchange, and the two domestic infosec IPOs since then (Rapid7 and SecureWorks) both currently trade underwater from their offering. In contrast to the recent infosec shutout, startups from several other IT sectors have all been able to enhance their $1bn private-market valuation on Wall Street, including Nutanix, Atlassian, Twilio and Pure Storage. That list will get a little longer as MuleSoft is set to debut at more than a $2bn market cap, up from $1.5bn in its final round as a private company.

A mule that’s actually a unicorn

Contact: Brenon Daly 

Unlike a fair number of late-stage startups, MuleSoft is no donkey trying to pass itself off as a unicorn. The fast-growing data-integration specialist has tripled revenue over the past three years, and appears to be on track to put up about $250m in sales this year. More importantly, MuleSoft is not hemorrhaging money. That should play well on Wall Street, which has telegraphed that it will no longer reward the growth-at-any-cost strategy at startups that want to come public (ahem, Snap Inc).

Assuming MuleSoft does indeed make it to the NYSE, where it will trade under the ticker MULE, it would mark the first enterprise technology IPO since last October. Of course, Snap is currently on the road, telling potential investors that its business model, which consists of hardware and disappearing messages, is the next Facebook rather than the next Twitter. But we’ll leave aside the offering from that consumer technology vendor, which just might be able to convince investors that losing a half-billion dollars last year, which is about $100m more than it booked as revenue, is a sustainable or even desirable business model.

Instead of Snap’s planned IPO, MuleSoft’s offering lines up more closely with fellow infrastructure software provider AppDynamics. (At least up to the point where Cisco comes in with a too-good-to-be-ignored $3.7bn offer.) A glance at the prospectus from each vendor shows both growing at a rapid clip (AppDynamics posted a slightly higher rate, even off a bigger base) and posting GAAP numbers that were at least headed out of the red (MuleSoft lost less than AppDynamics, on both an absolute and relative basis). Also, both firms had annual customer retention rates, measured by dollars spent, of roughly 120%. Wall Street eats up that sort of metric.

MuleSoft raised roughly $250m in total funding, most recently announcing a $128m round in mid-2015. With investors clamoring for growth tech companies right now, MuleSoft could certainly start life as a public entity with a double-digit multiple. Maybe not the nearly 18x trailing sales that AppDynamics commanded in its sale to Cisco. (After all, that was terminal value, not trading value.) But MuleSoft could almost undoubtedly convince Wall Street that it’s worth a premium to Talend, a rival data-integration vendor that came public last summer and currently trades at about 7x trailing sales. MuleSoft is larger than Talend and – more importantly to Wall Street – it is growing twice as fast. That profile will likely boost MuleSoft’s initial valuation on Wall Street to north of $2bn, or 10x its trailing sales of $190m.

2016 enterprise tech IPOs*
Company Date of offering
SecureWorks April 22, 2016
Twilio June 23, 2016
Talend July 29, 2016
Apptio September 23, 2016
Nutanix September 30, 2016
Coupa October 7, 2016
Everbridge October 11, 2016
BlackLine Systems October 27, 2016
Quantenna Communications October 27, 2016
Source: 451 Research *Includes Nasdaq and NYSE listings only