Google Cloud: Third Horse In A 3-Horse Race

| About: Alphabet Inc. (GOOG)


Google currently has less than a 5% market share in the public cloud market.

Its principle rivals, AWS and MSFT, are much larger and have greater geographic scope.

Google, to date, has seemingly competed more on price than anything else.

Google has hired well-known industry veteran Diane Greene to run its commercial division that includes Google Cloud Platform (GCP).

Using a highly optimistic assumption, GCP is unlikely to add more than 2 or 3 growth percentage points to Google's outlook between now and 2020.

A paean and a caution about an American Icon:

Google (NASDAQ:GOOG) (NASDAQ:GOOGL) was incorporated less than 20 years ago and it went public as recently as 2004. In less than 20 years, the company has grown to dwarf Oracle (NYSE:ORCL), rival IBM (NYSE:IBM) and it makes something like 5X Amazon's (NASDAQ:AMZN) current reported earnings. It is very nice to develop what is essentially a monopoly position in a key segment of the "New economy."

There are lots of high growth companies I follow and have followed in the Information Technology space. Some vendors, in their early stages particularly, can have share valuations that are hard to justify even for a growth junkie like myself. But to the best of my knowledge, Google/Alphabet is the one of only three companies that I follow in my space, for which the Yahoo Finance charts have to be presented in log scale. (The other's being Apple (NASDAQ:AAPL) and (NYSE:CRM).

Is Google the greatest American success story ever? I suppose the closest rivals would be Standard Oil and Ford (NYSE:F) but their glory days are more a century in the past now - well perhaps 90 years for Ford. But regardless of my penchant for dragging in bits and pieces of economic history, Google has to be considered the quintessential American success story of this century so far. If there is a metaphor for American exceptionalism, then this company is it

But like every other company both in and out of IT there is a sequence of events in the life of an enterprise and at just short of 20 years of age, the question has to be what does Google do for an encore and how can it avoid becoming middle aged? This company has essentially created the online advertising business and in the process, along with Amazon, has changed shopping in this country forever. And the company, unlike many cloud software companies, has been GAAP profitable for a long time now - no need to cut through the weeds of financial statements in order to find the value.

And yet the growth in the online advertising space is showing signs of beginning to flag. Google has been the market in Web advertising for many years now. No one, lest of all myself, has any doubt that the company's paid search businesses will continue its growth. These days the growth is in mobile ads but it really doesn't matter - advertising is still growing in mid-teens for this company. The real question is at what level will that growth continue.

As Google has grown, and as it has come to dominate such a significant area of the "new" economy so completely, it has attracted its share of political problems. I could speculate about the potential outcome of the EU probe regarding the potential competitive abuses of Google's dominance in both advertising and with its Android operating system. SA veteran competitor Dana Blankenhorn has put together an interesting article regarding the inability of Google to monetize its mail, news, maps and finance services. I am not sure I might describe Google as either lazy or poor at this juncture, but the point is well taken.

But I want to go down a different path in looking at this company's prospects going forward. I have no idea if Q1 was spectacular or just good. As the saying goes, "Yer pays your money and yer takes your choices" (in Punch Magazine of 1846). For me, the basic issue is what are the company's prospects in the areas outside of search and its Android operating system. Being the dominant player in a crucial segment of the "new" economy can and will lead to political consequences. Even if the company could monetize something like maps where it has a dominant service, would it really matter in terms of prolonging the company's growth run.

You have posed the problem. Now what is the solution! Or was that all a tease?

Google restructured both its name and its reporting segments a few months ago. I suppose it is supposed to be known as Alphabet these days, but just like the renaming of the Triborough Bridge here in New York, I doubt if the name Alphabet is likely to ever really going to take. The Google brand is simply too valuable a franchise to cast aside. But now we know something a little more about the company's potential growth buckets.

Google made just shy of $16 billion last year on a GAAP basis or $21 billion of non-GAAP net income and it generated $26 billion of operating cash. It isn't going to be particularly easy for the company to find a business segment in which it can generate enough revenues and margin to be significant. Google, while far less expensive than some of the smaller hyper-growth names, is hardly inexpensive. It currently has an EV/S of 5.2X based on consensus revenue expectations for the current year. At the consensus EPS forecast of $34.60 in EPS, the shares sell at a P/E of just less than 26X. Based on the earnings growth forecast, operating cash flow this year should be about $30 billion and free cash flow should be about $19 billion more or less. That suggests that the 2016 free cash flow yield is 4.2%

The mean price target is $930/share, which suggests that the company has to continue to achieve something like mid-teens earnings growth just to support the current valuations. If the growth in advertising and operating systems revenues starts to slow in percentage terms, where can the company possibly find enough revenue with high operating margins to replace the growth slowdown in advertising? And indeed, of course the percentage growth in core Google has already started to slow. Even though online advertising is a very big thing, its growth rate is slowing and Google likely cannot increase its market share that is already high.

If the real CAGR suggested in the PWC report is 12.1% with typical falling percentage trends, where can Google really go for the growth it is going to need to fill the "gap?" Again, the online advertising business is far more complex than some of the detailed numbers in the PwC report. I think the takeaway has to be that percentage growth in core Google is most likely to slow and indeed already has slowed.

Google has a revenue bucket called "other bets." Other bets include such things as Fiber, Nest and Verily. These businesses actually generate revenue, which generate a little bit of revenue. It also includes things such as the autonomous car development and some of the company's hardware initiatives. "Other bets" managed $448 million of sales last year, up 37% from the prior year. But "other bets" managed to lose $3.6 billion on the operating income line, a noticeable amount even for Google. I do not propose in this article to try to handicap the specific pieces of "other bets." My knowledge of autonomous cars is slight. Indeed, the only thing I can say about autonomous cars is that my friends would be more than happy to get me from behind a steering wheel and I concur.

The fact is that many analysts who cover Google make estimates for revenue streams from the constituents of "other bets" and just as regularly, analysts get disappointed. Before the last earnings call, the RBC analyst had estimated that Nest would have brought in $672 million of revenue for 2015. After all, Google paid $3.2 billion for Nest and it has owned the company for a few years now. Google, in its new incarnation as Alphabet, at least breaks out other bets as a whole. It is obvious that estimating $672 million for Nest when all of the other bets contributed $427 million of revenue for the year is an embarrassment. The real number, apparently was more like $340 million.

In the absence of any particular knowledge, it becomes a very fraught exercise to handicap "other bets" and I simply won't try. But it ought to be clear that whatever else is true, "other bets" is not very likely to become a significant Google growth driver any time soon and maybe never. It would take a brave, not to say foolhardy analyst who might think that any of the pieces in the other bets bucket is going to plug the potential growth hole for Google.

So, if "other bets" is unlikely to be the next big thing for Google, what is the opportunity that can keep this company growing at double-digit rates for some time into the future? While management hasn't given up on "other bets," it spends far more time talking about its cloud opportunity. It still mentions Fiber occasionally but CFO Ruth Porat was remarkably reticent in talking about Nest during the last conference call. One Google executive actually gave a presentation last fall in which he forecast that at some unspecified time in the future, Google cloud revenues would surpass advertising revenues at some point

The bucket in which Google's cloud business is included is called "other revenues" and it did $2.1 billion in Q4 2015, which was up by more than 24% year over year. On the last call, CEO Sundar Pichai commented, "Cloud is also a growing area where we see great opportunity and we're building a fantastic service that we will keep investing in." Google does not break out what are the constituents of other income but it did say that the preponderance of the growth in that segment is coming from Google's Play Store. Other sources outside the company have suggested that Google Cloud did all of $400 million last year with 10% of cloud revenue coming from Snapchat.

This will be the third in a series of explorations into the cloud business of major technology vendors. The first two articles, Amazon and Microsoft, were easy. Both of those vendors have a significant cloud services business that has an operating history of several years. Google's cloud business is not the easiest thing about which to make projections. If Google's cloud revenues double this year, then cloud is going to make up less than 1% of the company's projected revenues. Both Amazon and MSFT are far larger in the cloud than is Google. So the question is going to be can Google catch up, and if it can't, what is the outlook for its current cloud business and will it be enough to sustain double-digit top line growth?

Google's Cloud: Are we looking at a rising sun?

Perhaps you remember reading about the speech given by Benjamin Franklin at the end of the constitutional convention. Franklin is said to have told the convention delegates that he had been looking at a painting in Constitution Hall since the start of the convention to determine if the sun in the picture was rising or setting. He is said to have told the delegates, with the work of the convention in hand, Franklin now knew that it was a rising sun. My conclusion, while far more nuanced and hedged, is that for GCP the sun is really not rising that far.

Usually at this point in an article such as this, there is supposed to an attractive graphic with the sun hiding behind or breaking through the clouds. The graphic is nice to look at if you enjoy looking at those kinds of things but the graphic is not going to help to understand the issues Google faces in the cloud. One is left with more rhetoric and less substance than I would like.

While I believe that success in the cloud is an imperative for Google, I simply do not see the steps that Google can readily take to make up so much lost ground in terms of cloud market share. As I shall discuss, even if Google performs flawlessly going forward, it simply has come too late to the party for it to dent the market share leads currently enjoyed by both MSFT and AWS.

One issue for Google is that it currently only has three data centers which will rise to five by year's end. Why is that an issue? Simply because as the saying goes, "speed matters." Yes, this is 2016 and people don't have to work 50 miles from a data center in order for them to usefully subscribe to cloud services. At the moment, Google has but one data center in Europe, one in Asia/Pac and none at all in Latin America. At the moment, AWS offers 33 availability zones, which consist of one or more discrete data centers, each with redundant power, networking and connectivity housed in separate facilities. Of course, the precise comparison between data center numbers is not intended to represent all of the factors in terms of evaluating a cloud service, but it certainly suggests that Google has extreme problems in terms of scale.

When the cloud was mainly for developers or for SMBs which was the case as recently as a couple of years ago, that might not have mattered all that much. Now that the largest corporations are starting to shift standard workloads to the cloud, it does. I suspect that most large multi-nationals that are contemplating moving some of their workloads to the cloud are simply not going to be all that happy with Google's current coverage.

This year, AWS is to open 11 new data center zones including locations such as Montreal, Ohio, India and Ningxa (that is a city in China). Google is going to open two new regions, one in Tokyo and one in Oregon. That is certainly a start but Google is simply going to have to open data centers even where they are not currently justified in order to effectively compete with AWS and Azure.

While size certainly matters, so do price and performance. The latest studies that have been published on the subject in January of this year suggest that Google Cloud Platform has a lower list price than AWS per EC2 instance of between 15%-41%. The same study indicated that Google has an advantage in terms of latency and performance. All of the studies quoted above were either commissioned by or actually put together by Google, so I'm going to take a wild guess and suggest that the reality isn't nearly so one sided as the charts seen on the link.

Another study from a 3rd party source with 3rd party testing suggests that "the playing field between Google and AWS is about level." Pricing for IaaS is not as straightforward as one might wish to see. AWS as I mentioned in my report has different prices for just about anyway one can imagine using a public cloud. You can reserve time, you can get capacity on demand, you can bid for capacity in a Dutch Auction, you can go with a Priceline-like alternative that allows access after all of the higher paying customers have gotten what they need and if you can think of some other way of pricing in the cloud, AWS will probably develop a plan for that as well.

At least on the surface, Google appears to have won the race to the bottom for now, but as I commented in an earlier report, AWS has cut its prices by more than 50 times since it began offering the public cloud. The odds are that a typical user will pay a bit less for a typical Google installation which is what might be anticipated since AWS is the incumbent vendor much of the time and because of the far greater geographic coverage offered by AWS. As I mentioned above, if Latin America is important to a customer, or China or India, they are going to choose AWS ahead of Google because of local data center operations.

Yet another study that I looked at from what appears to be a third party site draws a different conclusion when it comes to performance. This study says that Microsoft Azure and AWS are about equal in latency and GCS runs 3X more slowly. I confess that I was barely able to understand most of the study - it is not written for the most part in terms that are completely comprehensible to this author.

I have, for example, not the slightest clue as to what the following sentence implies - "Google's fautil includes a perfding command and because it is built on boto it can actually be used for both GCS and S3." I think that the sentence says something nice about Google with regard to object storage and concurrency - but what it means exactly is just beyond me.

But one thing that the study does point out is the AWS charge for accessing a bucket from another region is ideal if you have data processing servers in multiple regions... I suspect multi-region buckets (AWS) are more fault tolerant as well. I'm sure with a bit more research, I could find yet more studies to muddy the waters even further.

My conclusion is that GCS is probably cheaper than AWS and Azure. As the third man at the prom, Google has to compete some way as many of the girls have been taken and price is not a bad way to go since the marginal costs of additional customers are close to zero. Google is going to lose lots of competitive bake-offs for no reason other than it does not have enough data centers for large multi-nationals.

Google certainly has the scale to be competitive given that its data center have to run the massive Google search engines, Gmail and YouTube. Google also has - well a certain cachet - although the modern way to say that is cool - simply because it is Google. And Google knows how to innovate as well as anyone in the IT world. Many developers have a strong distrust of Microsoft and feel much more comfortable developing on Google and feel that Google will remain more committed to open standards than Microsoft. I really can't measure that to be sure but the articles all suggest that suspicion of Microsoft persists.

Right now, Google is struggling in the cloud. Its market share is less than 4% with Amazon at 27% and Microsoft at 16%, according to Statista, while Synergy reports that AWS had a 39% market share in IaaS with MSFT at 11% and Google at 6%. It is described as "visionary" by the Gartner MQ report, it's below the centerline in terms of its ability to execute according to the same survey although it made up a little ground in the past 12 months.

Recently, Google hired Diane Greene, former founder and CEO of VMware (NYSE:VMW) from 1998 to 2008, as its head of the newly converged cloud software business. While the principal asset is Google Cloud Platform, Ms. Greene will also be responsible for Google for Work, Chrome for Work, Android for Work and Google Apps. At the same time, Google paid no less than $380 million for Greene's start-up Bebop. Bebop was supposedly developing HR, benefits and training management software, one of the more crowded software spaces in the world. Greene received $148 million for her stake in Bebop but she is donating all of those proceeds "to a donor advised fund," i.e. a charity.

Ms. Green has a formidable challenge in attempting to take Google's commercial assets and its cloud platform and develop them into a unified business that might possibly be able to mimic Microsoft's strategy with Azure. Crucial to her success is going to be developing a sales capacity and a marketing message to sell GCP to large-scale enterprises. Google's marketing message had been "build your app on App Engine - just like we do." The basic problem is that developers, for the most part, don't really want to build their app "just like Google does."

Ms. Greene has said that the reasons to choose GCP are that it is cheaper, better (in terms of security and open source software) and - my favorite - access to innovation. Of the three only price is really palpable and probably isn't enough by itself to propel Google's market share from what it is (around 5% depending on who's doing the counting to what it would actually have to be to be meaningful to this company's potential revenue and earnings growth. While Ms. Greene really does need no introduction, she had a very troubled tenure by all accounts at the end of her time at VMware where she had to deal, mainly unsuccessfully, with that company's core technology becoming commoditized and with Microsoft giving away VMware's product for free.

There are some VCs and industry observers who think that the hiring of Ms. Greene is a step in a longer-term plan for Google to develop a separate commercial products business including the cloud that could be potentially spun off to current shareholders and which would reap a very high valuation even if it was only achieving modest success. I think such speculation might be reasonable given all of the challenges GCP is going to face as it attempts to scale up in the IaaS business.

We're ready for the punch line

The punch line is pretty simple. Based on all the available data, Google is unlikely to build a business in the cloud that would be large enough to fill the growth hole that is likely to show its presence by the end of this decade. And so far as it goes, it seems to this writer that it is highly unlikely that Google can overcome the lead to which it basically staked both AWS and Microsoft by its tardiness in taking the cloud opportunity seriously. The cloud business at scale will be immensely profitable for all participants.

By its nature, Cloud or IaaS is a service business and is accounted for that way. Service businesses simply do not show their profit potential in their early years of existence since it costs so much to capture customers and to build infrastructure. We are starting to see operating margins rise at a prodigious rate for AWS-25% in Q3 of last year to 29% last quarter and they are likely to exceed 50% at some point.

But the issue is going to be one of scale for Google. The only real reason that I can see for Google to gain market share is based on costs. It does have lower and simpler prices than either Azure or AWS. It has huge disadvantages in terms of geographic coverage, its performance is probably the same as AWS and I really can't even begin to comment on whether it has better or more features than Amazon. Just reading about it is the same as reading about two kids playing or fighting in a sandbox.

The problem for Google is that AWS and MSFT have the so-called "virtuous cycle" going for them and that advantage cannot be competed against very readily. AWS is going to have lower costs than anyone for the foreseeable future. Running a price war when a company does not have the lower costs is something that business schools urge their students to avoid at all possible costs.

Absent figuring out what is really better in Google's cloud than in the cloud offered by AWS, and being able to articulate that difference in a fashion that is understandable by reasonably educated people, I am hard pressed to see Google materially expanding its market share in IaaS or developing enough business applications that can sit on Google's IaaS platform for it to gain significant market share.

Sometimes the best way to reach a conclusion regarding something like this is to present the absolutely most favorable scenario and see how that works in terms of what Google would actually need to achieve. Sort of an exercise in reductio ad absurdum

Let's say that Google doubles its market share to 10%. That could certainly happen because there are going to be loads of dropouts and Google has the resources to spend anything necessary to gain market share. After all, if it can afford to blow $3.6 billion on other bets in a single year, then it certainly can invest in the GCP at a prodigious rate. But getting beyond 10% requires some powerful assumptions that Google is going to gain share at the expense of AWS - simply because that is where there is share to be gained. And that just doesn't seem likely. So - 10% share - double what it is now.

The cloud market in which Google competes in 2016 is estimated to be something a bit over $30 billion. (There are many different numbers for just about everything cloud. Sometimes statistics include the cloud application vendors such as Salesforce and Workday (NYSE:WDAY) and Oracle and SAP (NYSE:SAP). I think that the appropriate statistic to use for this exercise is that this year's IaaS public cloud market is a bit over $30 billion.) That's close to 40% growth from the $23 billion of cloud revenue achieved in 2015.

Usually in making a forecast like this, it is typical to degrade growth rates but the public cloud has been confounding predictions on the upside for a long time now. Indeed, many services suggest that growth in public cloud will be more like 30% going forward but the methodology for such forecasts is suspect. So, let's use 35% growth in the market for the years between 2016 and 2020.

That comes out to something over $100 billion in cloud revenue by 2020. There are other forecasts that speak to a larger target but they include things such as apps and private cloud. Big number, but not inconceivable as the migration to cloud is a once in a generation phenomenon. 10% of the $100 billion forecast is of course, $10 billion.

I assume that by that time the economies of scale and the phenomenon of long-term renewals kick in, Google ought to enjoy 50% margins. Another pretty high number but the incremental costs of cloud renewals is almost negligible and renewals will start to become a higher and higher proportion of total revenues going forward. So, that is $5 billion of operating income by 2020.

Google's tax rate is significantly less than the statutory rate. That is mainly a function of the benefit of income being taxed at much lower rates in foreign jurisdictions. In addition, despite its domicile, Google had negative state income taxes last year. I have no idea as to what Google's tax rate is likely to be in 2020 - although I think that higher rather than lower is probably a reasonable assumption. But I will use a 15% tax rate on the income from the cloud anyway. So, that means that if it all goes very well for Google, the cloud could add a bit more than $4 billion of after-tax net income at that time.

Google's total net income next year, non-GAAP, is estimated to be just shy of $24 billion. Obviously, growth estimates are going to vary considerably, but I think using a 14% growth estimate for the period of 2018-2020 is as reasonable an estimate as any. So, that suggests that the earnings for Google excluding the cloud impact are going to be a bit greater than $35 billion. So, the cloud can add $4 billion to a $35 billion estimate using very generous assumptions.

I doubt seriously that adding 11% to the net income in the terminal period would really be adequate to support a positive outlook for the shares. Basically, making very generous assumptions for the success of the GCP adds between 2%-3% to the company's earnings growth rate. Not insubstantial when the growth rate will probably be just into double digits by the end of the period, but surely not enough to change anyone's opinion from hold to buy.

Some Final Thoughts

Google has been one of the quintessential American growth stories of the early 21st Century. It has created a market in terms of search and has become pervasive almost around the world. It has created the online advertising market that continues to grow and which has disrupted the cozy worlds of 20th century marketing.

But like all good things, the market for online advertising is starting to grow more slowly, and in any event, Google's domination of that market is attracting regulatory scrutiny. Google has some other services that it could potentially monetize but none of them have the huge potential of search. And it has a bucket of what it calls "other bets" that are just that, i.e. bets that are really not related to the core of the company's business. Other bets, at least to date, has been a money pit. It lost $3.6 billion last year and the revenue constituents of "other bets" do not provide a great deal of hope for the future.

So, the problem to be addressed is that Google shares are priced as though nothing is going to happen to the growth rate in its advertising franchise and it will continue to enjoy success in the future as it has for all the years since it was initially founded. That is simply an unreasonable basis on which to make a purchase recommendation on these shares.

The only logical place from which growth is going to come is from what Google calls its cloud platform. Google, simply put, is way behind its two rivals Amazon and Microsoft in terms of market share and more critically, perhaps, its geographical scope. It will have five data centers by the end of this year as compared to 44 data centers that AWS will operate by the end of the year.

There's nothing particularly different about the services that Google supplies compared to Amazon or Microsoft. One can find different estimates as to whose service offering has the lowest latency but that is not the best argument for Google to use because of its geographic limitations. It is hard for an independent observer to either know or understand some of the different features that Google offers and compare those effectively to AWS.

The only real difference that Google has going for it is price. But trying to be the price leader when competitors inevitably have lower costs than you do is not an easy undertaking. AWS can charge far less than Google and still make money.

So, I simply do not see how the GCP is really going to be as successful as would need to be to really move the needle sufficiently for this company. It got to the dance a couple of years late and it is now paying a very heavy price for its tardiness. I tried to put together a highly optimistic scenario for the future of GCP, probably far more optimistic than is really warranted. But even using the most optimistic assumptions I could, GCP only adds about 2%-3%/year to the company's growth rate through 2020. I simply do not think that is enough to change my mind on the share price valuation.

Google has been and will remain an American icon that should be saluted as a monument of American technology. That, however doesn't make it a great stock at this point. While I have no idea as to whether or not Google will achieve the so-called whisper number when it reports later this month, I think that long term, its valuation needs to compress.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.