Photo: Associated Press
Salesforce.com has a real problem and it’s not one of the usual suspects (Oracle, SAP or Microsoft).Google’s Marketplace announcement a few weeks back just confirmed if there is one company that is likely to be giving CEO Marc Benioff a really bad toothache, it’s Google, and unfortunately dentures might be the only longer term solution.
While the Apps Marketplace poses the biggest threat yet to Salesforce.com’s Force.com development platform and the AppExchange store, it’s only the latest indication of a collision course that Google and Salesforce.com have been on for a number of years. It’s not likely to be a head-on collision — more like a grinding side-by-side collision — but the outcome is what you would expect when a Hummer and a Mini are vying for the right of way. The Google Marketplace is likely to be a huge blow to both Force.com and AppExchange, but Google Apps is also going to undermine Salesforce.com’s CRM app by changing pricing expectations for online apps.
Email and CRM are Different, but Pricing Disparities Shouldn’t Span Galaxies
In a nutshell, the big problem that Salesforce.com faces is pricing. Some may think that comparing CRM and email pricing is like comparing rice to risotto. They are different applications. But they are delivered through the same model (SaaS) and they are both core enterprise apps, so there should be at least some semblance of pricing parity. And for those wondering why email is the point of comparison with CRM, it’s because it’s the anchoring app (along with Calendar) of Google Apps.
Since CRM is relatively more specialised than email, it would seem reasonable that it should be somewhat more expensive. On the other hand email, is significantly more computing-, storage-, and and bandwidth-intensive than CRM, and it’s the workhorse app for virtually all enterprises. It’s more than 40 years old and the original killer app of the Internet. In contrast, CRM software just started to get going about 10 years ago and it addresses functionally specific needs in marketing, sales and customer support. Everyone working at a company has an email address; only some use CRM apps.
But even if CRM should be somewhat more expensive, it shouldn’t be dramatically more expensive. Yet it is.
Google Apps Premier Edition (GAPE for short) costs $50 per user per year. If you add on the related Google Postini Antivirus, Anti-spam, message archiving, and other features, you increase the cost by another $50, to $100 in total. By comparison, Salesforce.com’s most popular version, the Enterprise Edition, has a list price of $1,500 per user per year. The cheaper professional edition runs for $780 per user per year. Even if a negotiated price is lower, we are speaking about significant pricing differences; between 8 and 15 times the Google Apps price.
It just seems inevitable that at some point an enterprise CIO is going to wake up and wonder: why am I paying between 8 and 15 times more for my Salesforce.com CRM application than I am paying for my GAPE email (Gmail), calendar (Google Calendar), messaging (Google Talk), video (Google Video for enterprises), intranet content management app (Google Sites), and online spreadsheet, word processing and presentations (Google Docs) solution? Some of the stuff like Google Docs is half-baked, but the core parts, such as Email and Calendar, are fully baked. The half-baked stuff will finish baking, but at worst, it’s an ignorable freebie.
Exactly when this happens is unclear — and Salesforce.com is doing just fine at the moment, aside from a nose-bleed valuation multiple of 40x trailing free cash flow. But given the growth rate of Google Apps, it might become an issue sooner than most think. And while Salesforce.com and Google Apps have been flying the SaaS flag together, until recently, Google’s Marketplace is now competing directly with Salesforce.com’s AppExchange. It’s a battle of CRM vs. email as the anchor app of a new platform, and it’s no secret that the broader reaching and more generic email application is the better candidate here.
Why Pick on Salesforce.com?
A few quick thoughts on why Salesforce.com is a good point of reference for Google Apps. Without a doubt, Salesforce.com has been the most successful independent SaaS company, by far. It has succeeded, largely because it’s cheaper than the shrink wrapped solutions it has replaced. Being the most successful SaaS company, it offers the best relative comparison to Google Apps, when viewed at similar points in their lifecycle. And, of course, it also shares the same pricing model as Google Apps, which offers an easy pricing comparison.
But Salesforce is just likely to be the first among the many enterprise software/IT casualties that lie ahead. And it’s not that Google is going to steal the revenue of others, it’s more like it is just going to suck out all the waste in the industry, which means that the trillion dollar enterprise IT sector is facing a shrinking future. When companies adopt Google Apps, there is dis-intermediation at the software, hardware, services, networking and data centre level. Some aspects, such as services, are not entirely eliminated, but significant portions maybe.
One last thing, before we get going. The Marketplace announcement is very significant. It could be the equivalent of the iPhone App Store for the enterprise cloud market — despite Salesforce.com’s much older AppExchange. But this discussion shouldn’t overshadow the bigger problem caused by the new economics — think Wal-Mart style pricing — that Google Apps Premier Edition introduced in 2007.
Comparing Growth Rates
Before Google Apps, Salesforce.com was the king of the SaaS heap, but after more than 10 years, it’s still a very pitiful heap. And it’s not that the concept of SaaS (or now cloud computing) didn’t make sense, it was just that all the contenders never had heft to begin with and in the enterprise space that is what you need to move the elephants out of the way. Salesforce.com is the only independent SaaS company with more than $1 billion in revenue and most others are well below half a billion in revenue, despite many having been around for 10 years (or almost 10 years). Aside from CRM software, all the other SaaS companies have only found success in niche offerings. In many cases they were not displacing incumbents so much as creating new applications. This is just a testament to how difficult it is to displace anyone within the enterprise IT space.
Here is a quick list for reference. Note: Oracle and SAP have sizeable SaaS businesses, but the bulk of their business is still shrink-wrapped software. I added them to provide some humbling context of the state of the enterprise cloud industry. The primary reason is simple: you can’t disrupt the shrink-wrapped industry by occupying niche application markets.
Company Ticker Sector Trailing 12 Month Revenue $ millions
Oracle ORCL ERP $23,266 SAP SAP ERP $15,320
Salesforce.com CRM CRM $1,306
Ariba ARBA Spend Management $339 Taleo TLEO Human Resources $198 Ultimate Software ULTI Human Resources $197 NetSuite N ERP for SMB’s $167 SuccessFactors
SFSF Human Resources $153 Kenexa KNXA Human Resources $134 Constant Contact CTCT Email Marketing $129 Vocus VOCS PR Management $85 DemandTec DMAN Consumer Demand $79Google launched the ad-supported Google Apps for your Domain (ads are only shown only in the email app) in August 2006, and the paid version, GAPE, roughly half a year later, in February 2007.
To compare the relative growth of Google Apps and Salesforce.com in their first few years of growth, just imagine the latter growth on steroids. Within roughly 3 1/2 years, Google Apps has amassed 25 million users and more than 2 million businesses. In contrast, it took Salesforce.com 10 years to reach 2 million users and 72,000 businesses. At the current run-rate of new businesses added, Google Apps grows its customer base by 3,000 per day, or more than half of the amount that Salesforce.com added in their most recent quarter. In case that 3,000 per day number doesn’t really resonate, that is a run-rate of 1 million businesses per year. Salesforce.com’s Investor Deck claims that the US business opportunity is: 6 million small businesses (with payroll), 100,000 medium sized businesses and 9,000 enterprises; it looks like the reference data comes from here). Not all of Google Apps customers are based in the US but the majority most likely are.
This is not a totally apples-to-apples comparison because the majority of Google Apps users are using the ad supported versions and are not paying a subscription fee. Even so, I estimate that GAPE accounted for roughly $204 million revenue in 2009, 264,000 businesses and 2.7 million paying subscribers (see my estimate assumptions at the end of the article). The estimate is not scientific but it doesn’t matter too much because there is an order of magnitude difference in growth rates even with conservative estimates.
The main point is that it took Salesforce.com roughly 6 years to reach $250 million in annual revenue, while it took Google about 3 years, or half the time, to reach about $300 million in revenue (roughly $200 from GAPE and $100 from the ad supported version).
Here is the story in a few charts. The first one simply shows total customers as reported by both companies. You can’t see much of a bar for Salesforce in 2002, but it’s useful to have it in there since that was Salesforce.com’s 3rd year.
This is the chart showing estimated GAPE (paying) customers. Even if only 11% of Google Apps customers are using GAPE it still amounts to more than 3x the number of Salesforce.com’s customers. I assume the ratio of GAPE to total Google Apps users has been consistent historically.
Here is the chart for users.
And here is the one for paying users (estimated) only. The difference here is less dramatic, with GAPE having less than 2x the number of Salesforce.com paying users (subscribers). This is not so suprising since Google has much greater penetration in smaller businesses, which have fewer subscribers.
Again, this is just comparing Salesforce.com to Google Apps; the other SaaS-lings are proverbial drops in the bucket by comparison. More Data Points for the “Google Apps is for Kids” Crowd
Undoubtedly there are some who still think most of Google Apps sales are still going to people doing business at home in their underwear. That’s what Steve Ballmer was saying about 2 years after the launch of Google Apps. Specifically, he was claiming (regarding Google Apps) that “nobody uses this stuff” and that “growth has flat-lined”. But we also know how he famously pooh-poohed the iPhone and Android at launch, so maybe we should just consider him a contrarian indicator. When Ballmer starts mouthing off boisterously, just remember to follow-up on his claims a year or two later; it’s at that time that he is likely to be removing his feet from his mouth.
Here are some other data points comparing enterprise adoption at similar stages in the CRM/GAPE product lifecycle. Salesforce.com registered its first 25,000 subscriber accounts in the quarter ending 1/31/2007, or 7 years after the company was founded. Google Apps reached this point within 2 years of the GAPE launch with the signing of Valleo for 30k seats in May 2009. Yes, it’s not entirely fair to compare email/CRM seats, since email seats are almost always higher in deals as everyone uses email (not so with CRM), but this also underscores how email is more of a cornerstone, and consequently a better candidate as a platform app.
Key Customer Wins Customer Size Date Konica Minolta 7,000 Mar 2010 KLM 11,200 Feb 2010 Fairchild 6,000 Jan 2010 Sanmina 15,000 Jan 2010 City of Orlando 3,000 Nov 2009 Delta Resorts & Hotels 4,000 Nov 2009 Rentokil 35,000 Oct 2009 Jaguar Land Rover 15,000 Oct 2009 City of Los Angeles 30,000 Oct 2009 Motorola 15,000 Jul 2009 Johnson Diversy 10,000 Jul 2009 Valleo 30,000 May 2009 Avago 4,100 Feb 2009 Genentech 10,000 Nov 2008 Washington, DC 38,800 Jun 2008
Notice how the uptake at large enterprises has mostly happened in the last 18 months, indicating it’s just picking up steam. It’s also noteworthy that Microsoft appears to have changed its mind about the Google Apps threat late last year. A month after Google beat out Microsoft for the 30,000 seat deal for Los Angeles, Microsoft dropped the price of its Exchange Online service by 50% (they simultaneously increased the mail storage per user from 5GB to 25GB, which happens to be the Google Apps size offering). They also dropped the price of the online BPOS (Business Productivity Online Services) by 33%. One thing you can’t avoid with such huge panic-struck discounting is validating your competitor.
Its the Platform, Stupid
The really big disadvantage that Salesforce.com faces is the scale at which its cloud platform operates. Salesforce.com developed its CRM solution to be competitive with shrink-wrapped alternatives, but not with the massive scale of Google’s infrastructure. A good reminder on what efficiency at scale can do is to remember the initial launch of Gmail in April 2004, when Google launched with 1 GB of email for free. Yahoo’s paid email offering of $60 for 100MB of mail storage was instantly vaporized.
What’s tough for Salesforce.com is that Google’s Enterprise business is able to leverage the giant cloud infrastructure that was always designed to have rock bottom operating costs. Unlike Salesforce.com’s CRM app, Google Apps is able to leverage all the R&D and fixed cost investment that has gone into optimising its search infrastructure. So Google has the luxury of only considering the marginal cost of entering new businesses.
We don’t know what the cost differential is but there are good reasons to believe its in that orders of magnitude of difference realm again. Rich Skrenta wrote presciently about the platform economies of scale that Google was developing 6 years ago; its really worth reading to understand this issue. To put the article in historical perspective, at the time it was written Google had spent less than $400 million on capital expenditures (cumulatively). Since that time the company has spent more than $8 billion on infrastructure and servers. Ironically the impetus for his article was the release of Gmail and how it leveraged the Google infrastructure. As it turns out, the core part of the Google Apps package is just an enterprise version of Gmail, built from the same codebase as the consumer version and run on the same types of servers in the same data centres. The company pointed out in a research paper in 2003 that it was building a generic multi-purpose infrastructure that scale to run different apps. Most companies buy special class hardware, such as NetApp servers for email, but this dramatically increases the cost of operations. The rise of the massive scale cloud with Wal-Mart economics has been in progress for a long time. And for all the hype around cloud computing, there is only 1 other company that has a comparably sized infrastructure: Microsoft; which happens to operate at a very large loss.
Coming back to Salesforce.com, Benioff proudly announced at a Credit Suisse conference in December 2008 that Salesforce.com operated with only 1,000 servers in two third-party data centres. (They’ve subsequently opened a 3rd in Singapore and they have around 2,000 servers now). Google doesn’t say how many they have, but they have stated that they are making tools to automate 1-10 million servers. Once again, an order of magnitude of difference.
This is not to say Salesforce.com is a slouch when compared with the rest of the industry. See these interesting comments from Salesforce.com’s earnings call this past January.
I asked Craig Mundie at the World Economic Forum, I was sitting next to him. I said, “You know, if Salesforce.com did not exist, how many servers would you have had in SQL servers and application servers and software and hardware and data centres would have to been bought and installed, how much energy would have had to be consumed as a direct comparison to what we are doing?” And he said, “Well, yes, probably more than a 100,000 servers would have had to be bought and consumed and all the energy and all the waste that goes with that.”
I mean we have only a couple thousand servers in our whole company, they are PCs and those are shared by those 75,000 customers, I mean that’s an incredible economy of scale.
So the claim here is that Salesforce.com is 50x more efficient at the hardware level than a shrink wrapped self hosted version of their application. Of course it’s kind of fascinating that Mundie (of Microsoft) would supply Benioff such ammunition to tote around and to be used against his own company, amongst others. But the statements are on public record and Mundie didn’t try to rebuke them.
The other reason why Salesforce.com lags in cloud operating efficiencies is simply because they are a sales driven company. The CEO of Zoho, Sridhar Vembu, first made this good point in April 2008 by noting the disparities in R&D and S&M spending by the company. Using the financial data from the last 12 months, Salesforce.com’s sales and marketing spending of $605 million was almost 5x its R&D spending of $131 million. At Google, R&D spending of $2.8 billion was almost 1.5x that of its S&M expense of $2 billion. Salesforce.com even spent more on G&A than they did on R&D, at Google G&A was less than what they spent on sales. Salesforce.com is not alone here; Oracle also spent almost 1.5x on S&M than what they did on R&D for the quarter ending in November 2009.
Its the Platform on the Platform Stupid.
Salesforce.com had absolutely the right idea when it decided to try and leverage its CRM business to create the Force.com platform and the AppExchange store; the problem was that CRM is not a critical enough enterprise app to get the developer traction it needs to succeed. It turns out that if you want an app in the enterprise around which to develop something like an Apple App Store, then email is way better than CRM; all employees use it and it’s a much more heavily used app.
When the Google Marketplace was announced, it had — or will soon have — third-party apps from prominent SaaS vendors such as Zoho, Netsuite, Intuit, Concur and SuccessFactors — none of which offer their wares in the AppExchange. Salesforce.com was notably absent from the announcement, even though competitors of theirs, such as Zoho and Netsuite, were present. Zoho has written before about how Salesforce.com tried to blackmail them into dropping their CRM app in order to offer their other products in AppExchange. It would not be surprising if this is the same reason why other key SaaS vendors have not hopped onto the Salesforce.com dream-wagon.
The really big difference for all the other SaaS vendors is that if they want to ride a platform in the enterprise, you want one with the largest reach and the fastest growth. Integrating apps into Google Apps and making them available through the Google Marketplace provides access to 25 million users, vs. roughly 2 million for Salesforce.com. So a no-brainer here. What’s also likely to happen is that the Google Marketplace is finally going to galvanize the SaaS vendors in the way that Salesforce.com has been trying for a long time. Up until now, the various offerings have been siloed. By integrating with Google Apps, there is the analogue of a coalition that can oppose the shrink-wrapped consortium. It’s just like getting the iPhone and getting all the apps through the App Store — seamless and quick.
Android + Chrome and Google Apps
It’s worth pointing out some of the other key stuff that is likely to further the Google Apps case. Many companies are locked into the BlackBerry platform for mobile email and Google Apps provides integration for this now. But for companies that do use Google Apps, Android devices start to become another no-brainer. Google has indicated that enterprise will get more attention for Android in the future. For companies that are considering moving to Google Apps, they can now contemplate ditching BlackBerry solution costs (its expensive) along with all the other server hardware, networking, data centre, and shrink-wrapped related goodies (licensing + maintenance, IT admins etc.).
The other obvious point is that the browser as an application delivery platform is going to advance at a much quicker rate for the next 5 years than it did for the last 5 years. IE6 was the dominant browser in 2005 and Microsoft’s browser share has dropped at an increasing rate from 89% to 61% since then. The reinvigoration of browser technology ensures that all the obstacles to web-based apps will just melt away at an increasing rate. Some people will simply never be convinced of this trend, but it’s a one-way trend, regardless.
Where does this leave Salesforce.com?
It’s kind of sad that in a way Benioff paved the way for Google Apps, only to have it come back repave the road in way way that favours them. Salesforce.com is not showing any ill-effects yet from the Google Apps tsunami — and likely won’t for some time — but it’s hard to imagine how declining prices in the IT world will allow them to maintain such a big price premium for CRM when compared with Google Apps.
Already we are seeing that Microsoft is engaging in some panic-stricken discounting to protect its email franchise — and more importantly keep Google out of the enterprise — and this is likely to continue. Microsoft has the same channel conflict issues that Barnes & Noble had when it tried to fight Amazon.com. It realised it needed to compete by offering an online store, but at the same time it didn’t want to undermine its physical store channel. We know how this played out; it lost because ultimately the online channel was the most efficient one.
Microsoft has the same issue. It is offering online alternatives but it has huge investments in its existing shrink-wrapped model. These compete with each other and you can bet that internal factions are rife at Microsoft. Microsoft also has to consider its VARs which get little upside from selling online versions of Exchange vs. the shrink-wrapped solution. It’s simple to understand they make more money from the latter. But Microsoft will grudgingly keep fighting, and more discounts (or bundling things such as their Dynamics CRM!) is certain.
The bottom line is that pricing is going to keep going down for corporate email and it’s not going to stop there. There is no reason not to anticipate that the enterprise IT sector is in a mature phase that might be on its way to ultimate decline (at least in revenue terms). The only solace IT fans should take is that mainframes are still alive and well; enterprise dinosaurs are not really susceptible to rapid extinction.
How much revenue does Google Apps Make?
Google doesn’t breakout the numbers GAPE revenue. It’s lumped into the “licensing and other” revenue category which accounted for $762 million in 2009. Here is my rough stab at Google Apps revenue (explanation following the table).
2004 2005 2006 2007 2008 2009 DoubleClick
$300,000 $300,000 Google Apps
51,343 139,836 203,759 Enterprise S
73,558 112,289 130,000 227,200 258,000 Total $45,935 $73,558 $112,289 $181,343 $667,036 $761,759
DoubleClick — Google acquired DoubleClick in 2008. The acquisition accounted for $148 million in the first six months of 2008 (see the 10-Q for reference), or annualized for 2008 that would be roughly $300 million. The licensing and other category already accounted for $112 million in 2006, which was prior to the launch of Google Apps Premier Edition (presumably most of this was Google’s Enterprise Search Appliance).
Other significant components of this category are Google Enterprise Search Appliance and DoubleClick’s ad serving business. Assuming relatively consistent growth in the search appliance business and flattish DoubleClick revenue growth (presumably because the DoubleClick model was not one that Google bought the company in order to push) then Google Apps could account for roughly $204 million in 2009 revenue. Note, Postini revenue would be included here and they had a run rate of approximately $60 million in 2007, and they were acquired in September 2007.
What we do know is that Google stated in June 2009 that Google Apps was generating hundreds of millions in revenue. Its not clear whether this includes ad revenue, but the estimates indicate most likely not.
As a side reference, the current 25 million Google Apps users could possibly be generating $100 million per year. This assumes an RPM of $0.50 (50 cents for every 1,000 for email page views), 200 workdays per year and an average of 40 emails read per day. The RPM rate is really conservative because email ads are notoriously low (anyone remember the last time they clicked on a Gmail email ad?).
Assuming a GAPE revenue number of $204 million yields a user estimate of 2.7 million for GAPE. This also assumes that half of the GAPE subscribers are using the Postini services then the average revenue per GAPE subscriber is $75. This would be roughly 10% of the stated Google Apps user base, and this number is likely understated as well.
This is a guest post from Gregor Schauer, who has worked in tech in Silicon Valley since 2000. Including working at Google. Gregor has also recently spent 2 years in equity research at JMP Securities and Jefferies, covering the Internet sector and enterprise software. You can follow him on Twitter here. Disclosure: Gregor is long Google and has no position in Salesforce.com.
Business Insider Emails & Alerts
Site highlights each day to your inbox.