SaaS Applications: [Subscription] Software is Eating the World!
SEE LAST PAGE OF THIS REPORT Paul Sagawa / Artur Pylak
FOR IMPORTANT DISCLOSURES 203.901.1633 /.1634
psagawa@ / email@example.com
May 15, 2015
SaaS Applications: [Subscription] Software is Eating the World!
SaaS enterprise applications make up a more than $31B market that is growing at more than 30%, and is projected to sustain 20% growth through 2018. SaaS products are a bundle, displacing not just the application, but also the enterprise infrastructure and personnel needed to deliver the service to users. With prices for IaaS hosting rapidly falling, infrastructure investment is no longer an entry barrier, enticing start-ups and threatening SaaS pioneers. In this context, having an excellent application with strong customer support in an attractive, cloud ready sector; exploiting low cost cloud infrastructure; and achieving relative expense efficiency are the critical factors separating winners and losers. We evaluated 43 SaaS focused public companies, considering their category market share and sales growth, their gross margin and capex-to-operating cash flow, and their sales cost to new revenue and R&D to new revenue ratios as indicators of positioning. Our framework reveals, DATA, PCTY, TNET, and VEEV as particularly well positioned, but shades CTCT, CVT, EPAY, ININ, QLYS, RP, and ULTI as relatively vulnerable longer term. We also subjectively valued the SaaS initiatives at traditional software leaders, with MSFT, ADBE, ADSK and INTU standing out against disjoint offerings from rivals like IBM, ORCL and SAP.
- SaaS is big and growing fast. According to Gartner, SaaS applications will be a $31B+ market in ’15, already 21% of the global enterprise application software market, but we note that by bundling data center costs into subscription fees, SaaS actually addresses the $1.1T spent annually on enterprise data center software, hardware AND services, with further savings in customer IT personnel as well. Gartner is expecting 20% forward growth in SaaS apps, a forecast that we believe may be conservative, given the growing rivalry fueled by falling IaaS prices.
- Not all SaaS is created equal. Valuations for public SaaS companies are all over the map, with WDAY’s 20x EV/Sales nearly 3x the 7.2x median in our universe of 43 public SaaS companies with market cap >$1B, and more than 9x the cheapest, CTCT at 2.1x. We used a framework to assess the relative positioning of the companies in this universe and also subjectively assessed the SaaS initiatives of several traditional SW players.
- SaaS is a bundled product. SaaS applications displace more than just packaged software applications, they also displace the hardware, infrastructure/platform software, communications services, and IT personnel necessary to deliver and support the application from the customer’s internal data centers. These elements are provided in a bundle with the application, and thus, winning long term SaaS solutions must succeed on both application functionality and data center cost/performance. This greatly increases the spending pool addressed by SaaS, and, until recently, has been a key competitive barrier for SaaS pioneers who have invested heavily in their own data center infrastructure.
- IaaS levels the playing field. The steep downward price trajectory for cloud-hosted computing and storage has opened the door wide for SaaS startups, who no longer must contend with playing infrastructure catch-up against the likes of CRM, N, IBM, ORCL and SAP. Indeed, we believe that those early leaders will find themselves disadvantaged in the long term by their commitment to sub-scale internal data center operations that are much more costly, lower performance, and less flexible than the IaaS offered by AMZN, MSFT and GOOG and purchased by their rivals. We assessed our universe of SaaS companies for PP&E and lease obligations relative to sales preferring companies without firm commitments to costly internal data center operations. Smaller, newer companies, such as GWRE and EGOV rated well, while older or larger SaaS names, like CRM, BOX, and CVT, trailed.
- The right product in the right market. As infrastructure investment fades as a barrier and scale becomes somewhat less critical, SaaS success will be more driven by the quality of the application itself and the attractiveness of the market niche that it targets. We created a compound metric to balance application and market attractiveness, multiplying SaaS category market share by the relative growth rate of company revenues vs. the category growth rate, then adding a market category attractiveness factor determined by the projected 5th year category market size. Here, CRM, ZEN, HUBS were the leaders, while QLYS, RP and SPS rated lowest.
- Operations matter. Over time, we expect falling barriers to entry to drive SaaS prices lower, potentially squeezing profitability for many players, even those already taking advantage of low cost IaaS infrastructure. As this happens, operating efficiency will become an increasingly important differentiator for SaaS companies. We looked at R&D and sales costs relative to next year revenue growth as an indicator of this efficiency for the companies in our universe. VEEV, SNCR, WDAY were the most efficient, while RP, ININ, and CTCT were the least.
- Considering all 3 factors reveals likely long term winners. 5 companies from our universe of 43 SaaS vendors appear in the top half of all of our metrics – infrastructure flexibility, application quality and operating efficiency. These are DATA, TNET, VEEV, HUBS, and EIGI. Another 18 companies missed the perfect rating by one category. For the companies that are carrying above average infrastructure costs, the answer will be an eventual transition to an IaaS host, something that most will find challenging. For companies that are not gaining traction or that are addressing a modest target segment, the answer likely lies in building a defensible and efficient niche business. For companies that trail in operating efficiency, most are small and fast growing and the answer is obvious – get more efficient. The framework also reveals 6 companies challenged on all of these fronts – CVT, EPAY, ININ, QLYS, RP, and ULTI.
- Most traditional SW players will struggle w SaaS. The large enterprise software competitors do not break out costs for their SaaS operations, so it is difficult to evaluate them in our framework. However, most of them rely on their own data centers, with MSFT the only one that we believe can boast a modern, web-scale, cost competitive infrastructure. Most also face Clay Christensen’s “Innovator’s Dilemma”, propping up lucrative packaged software franchises while simultaneously trying to build and buy SaaS product lines to compete with them. We looked at the issue, measuring each company’s cloud share vs. its traditional share, preferring companies with a strong relative showing in SaaS. With this perspective MSFT, ADSK, ADBE and INTU rated as the most attractive.
The SSR TMT Heatmap
Life in the SaaS Lane
According to Gartner, SaaS applications constitute a $31B market anticipated to grow 20% over the next 5 years. While the overall application software market is pegged at just over $140B, The market addressed by applications sold as a service, because they include not just the application but also the computing, storage, communications, platform software, facilities costs, and IT support needed to serve users, is much, MUCH bigger. Total annual global data center spending tops $2 trillion, and while we expect the shift to be highly deflationary, the long-term, all in opportunity set for public cloud computing is easily north of $1T. It is an exciting time to be a SaaS player, disrupting the market.
However, not all SaaS operators are created equal and it will take more than participation to win. Valuations are all over the map, from 20x EV/S for WDAY down to 2.1 for CTCT. Returns for a universe of 43 SaaS companies with market caps over $1B have averaged 57.9% over two years, but the spread is very wide, ranging from QLYS’s 228.4% returns at the top down to CSOD’s –23.4% at the bottom. We expect the future to be every bit as volatile, so how is an investor to choose? We believe that winners must succeed on three fronts. The first, obviously, is the application itself, which, ideally, should target a particularly attractive market segment and outgrow its competition. With market data and forecasts from Gartner, we built a compound metric – (company growth rate/segment growth rate)*company market share + (5th year segment share of the total SaaS Market) – to assess each company’s position vs. that ideal.
The second factor is infrastructure, the hidden part of an implicit bundle. Traditionally, this has been a substantial entry barrier, as SaaS pioneers struggled to raise the considerable capital needed to build their data centers. Today the barrier is falling, as vigorous rivalry amongst IaaS behemoths AMZN, MSFT and GOOG puts world class cloud hosting within a swipe of a credit card for startups at prices already below the costs of the subscale and less-than-cutting-edge data centers run by their older rivals. We measured each company’s sunken investment in infrastructure by their PP&E and lease commitments relative to sales. Finally, SaaS winners must have efficient operations – with barriers to entry low, investments in improving the product and acquiring customers must pay off quickly. To capture that cause and effect, we calculated operating expenses as a percentage of the following year’s sales on a rolling 12 month basis.
Of the 43 SaaS companies in our analysis, DATA, TNET, VEEV, HUBS and EIGI delivered above median marks in all three metrics. 10 more companies carried high infrastructure commitments, including many of the biggest and best known – i.e. CRM, WDAY, and NOW. All of these companies are growing quickly, but we believe that must begin planning a move to IaaS hosting, or risk long term competitive threats. Companies with relatively weak applications and/or unattractive target segments may face limited futures as niche providers. Players with inefficient operations will wear out investor patience if they can’t become more efficient in what we will believe will be increasingly fierce price rivalry.
We also considered the SaaS initiatives at the big packaged software vendors. Here, the big and early commitment to the cloud by ADSK, ADBE, INTU and, in particular, MSFT stand out. In contrast, the M&A happy approach by ORCL, IBM and SAP is partly hamstrung by the impetus to milk their older franchises and further burdened by subscale and less than cutting edge infrastructure.
Send in the Clouds
SaaS applications are already a $31B annual market (Exhibit 1). It is tempting to simply compare that to the $140B that Gartner estimates will be spent on enterprise applications globally in 2015, but there are some complications. First, SaaS applications are sold as a subscription – typically per user, per year – while packaged applications are sold as a license – largely paid for upfront with a small annual software maintenance fee ongoing. This results in a very different flow of revenues and expenses. Packaged software sales follow boom-bust upgrade cycles, while SaaS subscriptions miss the upfront boom, but deliver steady predictable revenues. Costs are also cyclical for traditional software, with R&D rising ahead of upgrade releases and sales commissions following sales.
Exh 1: Worldwide SaaS Enterprise Applications Market, 2012-2018
Second, SaaS applications are a bundle, containing not just an application, but also the data center infrastructure and personnel needed to support the application. This second part of the bundle displaces significant costs that enterprises would otherwise have to bear in operating their own data centers. As such, SaaS applications address not just $147B but rather the $1.1T that global enterprises spend on their internal data center hardware, software and operations (Exhibit 2). Because of these factors, simple market shares based on contemporary revenues between traditional packaged software and SaaS applications are basically apples and oranges, despite both products addressing a similar enterprise need.
Exh 2: Worldwide Data Center Hardware, Software, and Operation Spending, 2013-2019
As such, we believe most analyses of the SaaS market overstate the current penetration of SaaS applications, but also drastically understate the long term potential for SaaS growth. Gartner estimates that the SaaS application market will grow at a 20%+ CAGR over the next 5 years, a projection that we believe is quite conservative. Importantly, we believe that the SaaS market will have legs and that 20%+ growth rates could persist for over a decade, as wave after wave of new applications come to market and as wave after wave of enterprise customers bite the bullet and adopt subscription software for their key applications.
Who’s on First?
While SaaS applications operate with a very different delivery and revenue model than packaged software applications, they compete for the same business computing use cases. For customers, SaaS offers some distinct advantages over the traditional model. First, subscription software frees the enterprise from responsibility for the data center capacity needed to run it and the personnel necessary to support it (Exhibit 3). Upgrades happen automatically and frequently – users are always on the latest version and transitions are gradual without the typical downtime and investment necessary for swapping out one version for the next. User support is usually included and is often far better than can be provided internally. Security is also included and is also often far better than can be provided internally, even though SaaS is inherently friendly to mobile access by employees. SaaS subscriptions are also flexible – it’s generally easy to adjust the number of users or the range of functions under contract and there are typically no hidden fixed-cost burdens or scalability limits. Considering the all-in costs of internal data centers and staffing, SaaS also tends to be cheaper.
Exh 3: Basic On-Premise versus Cloud Cost Comparison
However, there still are some advantages to the traditional software model. The most intractable is latency – SaaS applications may be hosted hundreds of miles from the customer premises and accessed via the internet, imparting some 20-40ms in delay between the time data is sent on one end and received on the other (Exhibit 4). For the large majority of applications, this is unnoticeable, but for a few, such as shop floor controls for precision manufacturing or automated trading, latency is a critical requirement. Another is customization – businesses may prefer idiosyncratic software tailored specifically to their unique requirements. These enterprises may use packaged software as a starting point, writing their own modifications. This is not generally supported within the SaaS model, although there is often flexibility within the applications to support customized interfaces on top of the standard functionality.
Exh 4: Effect of Distance on Content Delivery Performance
Exh 5: Universe of Enterprise Application Software – Suitability for the Cloud by Application and Current Spending on Cloud
We believe that the balance of advantages and disadvantages broadly tips toward the SaaS model in general, but that there are differences between the many categories of application software that make some better near term candidates for SaaS than others (Exhibit 5). Gartner divides the application software market into 10 categories: Business Intelligence (BI), Customer Relationship Management (CRM), Digital Content Creation, Enterprise Content Management, Enterprise Resource Planning (ERP), Office Suites, Project and Portfolio Management, Supply Chain Management, Web Conferencing/Teaming/Social, and Other. Across this taxonomy, ERP is currently the largest segment at over $27B in projected 2015 spending, while just $2B is expected for Project and Portfolio Management. The forecast growth rates for SaaS applications within each segment also vary widely – Digital Content Creation is expected to rip at a nearly 40% pace through 2018, while the Conferencing/Teaming/Social bucket, already more than 70% SaaS, is projected to just make double digits.
The contrast between the various segments within the SaaS market is a substantial contributor to the disparate performance of the many SaaS-focused stocks. Two-year returns amongst the 27 names in our universe of $1B+ capped SaaS vendors that have been trading that long range from Cornerstone On Demand’s (-21.6%) decline up to Proofpoint’s 200.8% appreciation (Exhibit 6). Amongst 16 stocks that have IPO’d in the past two years, PayCom and Tableau have more than doubled since their debuts, while RingCentral, New Relic, Cvent, Box, Veeva Systems, BenefitFocus, and Opower are all down, in some cases down hard (Exhibit 7). Valuations are also all over the map – OPWER trades at just 2.9x EV/S, while NewRelic gets a 13.6x multiple.
Exh 6: 2-Year Performance of SaaS Companies
Exh 7: Performance of Recent SaaS IPOs
How, then, to make sense of SaaS? We built a three factor framework, assessing each company for its infrastructure flexibility, for the attractiveness of its application, and for its operating efficiency.
Exh 8: SSR’s SaaS Framework
In the beginning, the pioneers of SaaS – companies like Salesforce.com and Netsuite – had no choice but to build the data center infrastructure upon which they would run their applications. In the late ‘90’s, when these companies were formed, that meant clustered Linux servers running structured data base (i.e. Oracle) infrastructure software operated in house. This was an enormous capital commitment for a start-up and it represented a substantial competitive moat for first movers.
Meanwhile, Google was hatching a revolution on the infrastructure side, inventing a new data center approach that did away with server clusters and structured data bases. This technology, invented to scale to the size of the internet and contributed into the open source community, became the basis for a new architecture embodied by Hadoop, HBase, NoSQL and other innovations (Exhibit 9). Data centers built on these principles no longer were bound by the scale limits of the past, enabled millions of processor cores and disk drives to work in concert, delivered extraordinary performance, and critically, drove the cost of computing and storage to unprecedented lows (Exhibit 10).
Exh 9: MapReduce/Hadoop parallel computing flow
Exh 10: Costs of 100TB of Storage, Internal versus Cloud
Intrinsically, SaaS applications are a bundle containing not just the application, but also the infrastructure and operations necessary to deliver it. The capital commitment that SaaS pioneers made in ‘90’s era data centers have evolved from a strategic moat to a potential anchor. With the rise of the modern data center architecture, cloud hosts, offering data center “infrastructure as a service” or IaaS gained enormous momentum and scale. With scale came dramatically lower costs and with lower costs came the ability to drive prices ever lower. We believe that the top three IaaS players, Amazon Web Services, Microsoft Azure and Google Compute Engine have cost structures that may be a magnitude or more lower than the average enterprise data center, and less than half that of even the most efficient clustered server operations. This advantage is widening, emboldening the big three to keep slashing prices and, in the process, enabling a new generation of SaaS startups who no longer have to raise capital to build out private data centers. VC-backed SaaS startups nearly doubled between 2011 and 2014, with more than 800 deals in BI and 400 deals in CRM alone over that time (Exhibit 11-12).
Exh 11: VC-backed SaaS Investments by Application Type, 2011-14
Exh 13: Quarterly Financing Trends of VC Backed SaaS Companies, 2011-14
The impact of plummeting IaaS costs and prices has not yet really had effect on the SaaS market, but we believe that it is a matter of time. The most recent market entrants largely exploit the low costs in the IaaS market, and a few of the older players, most notably Workday, have explored using hosting for some of their products. Our thesis is that the cost advantages of the web scale players will drive dramatic consolidation in the IaaS/PaaS markets, and that SaaS companies will have little choice but to transition to a hosted model. Those companies with substantial capital investment sunken into data center PP&E or with funds committed to leases for the same will be more vulnerable to the transition to the hosted model.
Exh 14: SaaS Companies Ranked by Infrastructure Flexibility
We assessed the PP&E and lease commitments of the 43 companies in our universe as percentage of their annual sales (Exhibit 14). On this metric, many of the larger companies fared poorly, with Box, Salesforce, Splunk, Cvent, Workday, Athena Health and New Relic all generating less than $3 in annual revenue for each dollar in capital committed to their business. On the other extreme, GuideWire, Nic Inc. and LifeLock all generate more than $20 of sales per dollar of capital commitments. This is potentially a harbinger of things to come, and we note that Workday has already begun thinking about the transition, having launched its BI products on Amazon Web Services.
Of course, a good SaaS company should have a strong application that is gaining share in a large and growing market segment. We built a compound metric to capture both elements with the average percentage share gain of the application against other SaaS companies in its category weighted as 2/3rds and the projected size of the category in 2018 relative to the total SaaS market, expressed as a percentage, as 1/3. We recognize that this metric is a bit kludgy and arbitrary, but believe that, given the available data, it is a reasonable proxy for the potential of the application (Exhibit 15).
On this metric, some obvious candidates rise to the top. First ranked Salesforce.com is the dominant player in CRM, the largest SaaS application category and one of the fastest growing. Zendesk is moving quickly to build share with its customer service solution. Third ranked HubSpot, which makes inbound marketing and sales software, also gets a lift from the size of the CRM market. At the bottom are small, focused players like Guidewire, Qualys, and SPS Commerce. For companies ranking poorly in this metric, the goal should be to establish a high margin, defensible niche position, a strategy less likely to appeal to investors looking for the next double unicorn.
Exh 15: SaaS Companies Ranked by Application Attractiveness
Growth is good, but it comes at a cost. Applications have to be developed and updated. Customers have to be contacted and won over. Ideally, investments in R&D and Sales expenses should pay off in new subscription revenue in subsequent periods. We compared these expenses from a year ago and compared them to the growth in in current year sales – expressed as new sales per dollar spent on R&D and S&M (Exhibit 16).
The most efficient company in the 43 stock universe is Veeva Systems, a life sciences cloud specialist, which clocked in at an impressive $1.49 of new revenue in 2014 for every dollar of R&D and S&M spent in 2013. This is more than 50 cents higher than number 2, Synchronoss Technologies, which operates a cloud platform used by carriers to manage the activation of new devices. Following just behind in the top ten are several familiar names – Workday, Tableau Software, Service Now, Zendesk, and Salesforce.com.
At the bottom of the list are companies that are either newish companies obviously investing in growth or established players groping unsuccessfully for sales growth. RealPages brings up the rear, just behind Interactive Intelligence Group, Constant Contact and Broadsoft. These companies will need to gain efficiency, either by growing into it or cutting down to it, if they are to be successful.
Exh 16: SaaS Companies Ranked by Operating Efficiency
Exh 17: SaaS Overall Rankings
Exh 18: SaaS Company Summary Performance
Exh 19: SaaS Company Summary Performance – Horizontal Focus
Exh 20: SaaS Company Summary Performance – Vertical Focus
Exh 21: Above Average on All Three Metrics
Tic Tac Toe – I Win
Five SaaS companies ranked in the top half of our universe for all three metrics (Exhibit 21). Of these, TriNet Group would stand out as the leader and an outlier across all of the metrics without adjusting for the company’s unusual revenue streams from its insurance services. Even after making the adjustment and netting the insurance business, the company appears to be above average in every metric. The stock has appreciated 12.6% since its IPO in March 2014, making it a middling performer in the group, and it trades at a serious discount at just .88 P/S and 5 year PEG ratio of less than 1, while growing revenues at a healthy 23% clip.
Tableau Software is the largest cap player amongst this group, well positioned in the attractive Business Intelligence category, with notably strong operating efficiency and decent infrastructure flexibility. It has been a very strong performer, up over 120% since its IPO nearly two years ago, and trades at a 13.2x EV/S against sales growth of nearly 75%. We have included Tableau (DATA) in our large cap model portfolio since 4/15/2014.
Veeva Systems is a life sciences vertical application with very efficient operations, combined with above average application attractiveness and infrastructure flexibility. Veeva’s SaaS CRM solution, tailored to the health care industry, faces a crowded field, but has been growing sales at better than 35% with a relatively reasonable 2.2 PEG ratio and a 10.3x EV/S. HubSpot and Endurance International Group Holdings round out the group. Both ranked well on infrastructure flexibility and both just topped the mid-line on operational efficiency. HubSpot, which offers a SaaS solution for managing in-bound marketing activities, was the 4th ranked company on application attractiveness. It is growing at better than 50% YoY and trades at a group normal EV/S ratio of 10.7x. EIGI, which offers website management for small businesses, rates just above median on application attractiveness. It is growing sales at a robust 22% and is consistently profitable, trading at just a 0.74 PEG ration and 5.6x EV/S, perhaps due to the more mundane nature of its application.
Exh 22: Above Average on Addressable Markets and OPEX
Eleven companies fell out of the top group because of capital investments and lease commitments that put them in the lower half of our infrastructure flexibility metric (Exhibit 22). In this group are some of the biggest names in SaaS – Salesforce.com, Workday, Service Now, Box, Zendesk and Splunk, along with lesser known Dealertrack, New Relic, Paycom Software, Synchronoss Technologies, and Ellie Mae. With the exception of Box and New Relic, these stocks have been strong performers over the past two years, and the risks associated with a future transition to an alternative data center strategy may be a few years into the future.
However, we believe waiting until new competition and falling IaaS prices begin to pressure prices is waiting too long. Workday already runs its fast growing analytics products on AWS, suggesting an openness to increasing its use of IaaS in the future. We suspect that all of these companies will quietly begin planning for their own eventual transitions, with the greatest onus on Box and Salesforce.com, which stand out in this group for their unusually large infrastructure commitment. The recent rumors of talks between Salesforce.com and Microsoft – an idea that would have been unthinkable even a year ago – suggests a recognition of the synergies that could be gained by combining Salesforce’s application with Microsoft’s Azure infrastructure.
Exh 23: Above Average on Addressable Markets and Infrastructure Flexibility
Four companies, Cornerstone OnDemand, Marketo, 2U, and NetSuite, missed the top rung for operating expenses that generate relatively meager new revenues, and we note that Cornerstone and Netsuite just made it over the cutoff for application attractiveness, or they would have had two strikes against them. All of them are growing at a decent clip, ranging from Cornerstone’s 28% to Marketo’s 42%, but not surprisingly, none of them makes a profit. None of these companies is amongst the very worst on our efficiency metric, but Cornerstone and 2U each spend $3 for every new dollar of annual revenue, and Marketo and NetSuite spend more than $2. If the coming years bring the IaaS fueled competition that we expect, these companies will have to do something about their inefficient spending (Exhibit 23).
Exh 24: Above Average on OPEX and Infrastructure Flexibility
The Niche Crowd
Four more companies are admirably efficient and have not overcommitted on infrastructure, but simply rate poorly on application attractiveness – Paylocity, Opower, NIC Inc., and LogMeIn (Exhibit 24). These companies are all small – annual sales below $280M, market caps below $1.7B – and while three of the four are still delivering strong growth, none serves a market that would seem capable of sustaining that growth for very long. Paylocity is a cloud-based payroll processor for small and medium businesses, Opower helps electric utilities work with their customers to reduce usage, NIC Inc. helps Government agencies interact with citizens over the internet, and LogMeIn provides online meetings and screen sharing for collaboration and customer service. The logical long-game for these companies is either as a small, profitable, and defensible niche player or to be acquired.
Exh 25: Below Average on 2 or 3 Metrics
Two and Three Time Losers
19 companies fell below median in at least two of our metrics, including seven that failed across the board. Of these, just two made it across with a better than average application – Athena Health and Constant Contact (Exhibit 25). These companies would seem to have a bit more going for them, although it should be noted that Constant Contact is near the very bottom in terms of its operating efficiency. Still, the one-star ratings are an indictment of management and its responsibility to spend its money wisely.
The remaining 17 all suffer from less than attractive applications relative to the 43 stock universe, with only Proofpoint standing out as subjectively underappreciated to our eye. We have held Proofpoint in our small cap model portfolio for a few quarters, and after a bit of a slowdown in 2013, it has picked up the sales pace to better than 35% growth in the past year. Unfortunately for Proofpoint, our attractiveness metric looks at two years of growth, thus the low rating, but it would seem that a future look might rate the company more kindly.
The Big Guys
Traditional enterprise software companies understand the threat that SaaS applications represent, and most have moved to launch or acquire their own SaaS entries. Unfortunately, none of these players break out their financials in a way that would allow us to evaluate their cloud efforts within our three metric framework, but it is worth a more subjective consideration of their cloud efforts.
Microsoft is the valedictorian of this group. Its quixotic competition with Google in search forced it to be an early adopter of the modern data center architecture and practices that its rival pioneered. Its considerable consumer facing franchises – not just Bing, but Xbox Live, Skype, Hotmail/Outlook, and others – gave it the scale necessary to be competitive and an impetus toward moving its core franchises, and in particular, its Office cash cow to a SaaS model. Today, Microsoft’s Azure is one of just three commercial IaaS platforms that we believe have the architecture and cost structure necessary to succeed in the hosting game. This gives the company a serious leg up in its SaaS applications. The shift to Office 365 is well underway, with positive long-term implications for the profitability of what is already a stunningly profitable business. Dynamics ERP is growing nicely, with obvious opportunities to add to the scope of its functionality.
Adobe, Autodesk and Intuit have also been appropriately held up as examples of traditional software companies who have made successful moves to the cloud. All three companies are now SaaS-first in their focus, and while the trio still host from their own internal data centers, they have begun to exploit the advantages of commercial hosting. Intuit plans to shift entirely to AWS, while Adobe and Autodesk have made their applications available on AWS and Azure as well for customers that prefer to keep their documents and working files on those platforms.
LinkedIn is obviously not a traditional enterprise software company, but the financials for its successful recruiting/HR application are knotted with its professional social network. Considered on its own, LinkedIn obviously has an attractive application, and appears likely to be relatively efficient, but, like many SaaS pioneers, lacks scale and cost/performance in its internal data centers. In our framework, we suspect that it would fall into the same group as companies like Salesforce.com and Workday.
Most of the remaining traditional enterprise software leaders have moved to build out internal data centers for hosting while snapping up SaaS applications through M&A. Oracle has acquired over a dozen SaaS companies, led by CRM platform RightNow. IBM has likewise been busy buying small, private SaaS applications, also dropping $2B for IaaS platform SoftLayer in 2013. SAP has been the most acquisition happy, dropping almost $20B on Concur, Ariba, SuccessFactors, Hybris and Fieldglass in the last three years (Exhibit 26).
Exh 26: SaaS Acquisitions by Major Enterprise Software Companies
Despite the substantial investment, it will be difficult for these companies to make a successful transition to the cloud model. All suffer from the perverse incentives described in Clayton Christensen’s popular “The Innovators Dilemma” impelling them to favor their highly profitable packaged software franchises to the detriment of their less profitable but more promising cloud initiatives. All will suffer from their commitment to internal data center hosting, without the cost efficiency and expertise born of massive web-scale consumer franchises. All will struggle with integrating their newly bought cloud properties with their big company, software licensing culture. Meanwhile, the clock is ticking ever faster. We are not optimistic.