The best way to understand how Red Hat built a multi-billion dollar open source business is to look at IBM. Red Hat founder Bob Young spoke at the All Things Open conference in 2014 about this :

There is no secret. It takes a lot of hard work to retain customers, think through and understand development opportunities. What competitors are not doing, what can you do better? One of the great examples that inspired us at the beginning, besides the fact that I didn’t have enough money to pay for a rented apartment, is the example of Lou Gerstner and IBM ...

Gerstner came to IBM and transformed the company magically in three years ... His insight was that he spoke to a whole bunch of IBM customers and found out that they really didn’t like any of the company's products. Everything is normal, but with each personal conversation it turned out that there was always some more suitable product from another company ... He asked: “So why do you buy from IBM?” Customers answered: “IBM is the only technology company with offices everywhere we do business, ”and as a result, Gerstner realized that IBM is not selling products. She sells service.

He talked about it in public, and we also say in Red Hat: “OK, we do not have a product for sale, because it is open source and anyone can use our innovations as well as ourselves. Therefore, we do not sell the product. But Gerstner and IBM's example show that customers do not buy products, but services: the things that make them more successful. ” And this is one of the earliest ideas about our business: the idea that we are actually in the service industry. Even when we sold shrink-wrapped boxes with Linux, we saw this as an intermediate step to real growth, so that we could sign service contracts with real customers.

Yang's story went through a full cycle when IBM bought Red Hat for $ 34 billion, 60% higher than the closing price of Red Hat shares on the previous day of trading. IBM hopes that it will also go through a full cycle: it will return Gerstner's magic, which depended not only on his understanding of the services business, but also on the epochal shift in corporate computing.
')

How Gerstner Transformed IBM

Earlier, I wrote about the transformation of IBM Gerstner in the context of Satia Nadella’s attempt to do the same at Microsoft, and that the company's culture is extremely difficult to change and its nature cannot be changed. From Microsoft's Monopoly Hangover :

The great thing about monopoly: the company does anything, because there is no competition. But when the monopoly disappeared, the company is still able to do anything, but only at a mediocre level, nothing high, because it has become fat and lazy. In other words, for the former monopoly, “big size” is the only truly distinctive asset. This was Gerstner's key understanding when it came to making a plan for the future of IBM ... in the Gerstner vision, only IBM had the scope to deliver solutions instead of products.

However, a solution-based strategy requires the existence of a problem. Another factor that helped Gerstner's transformations was the Internet. By the mid-1990s, companies were faced with a completely new set of technologies that were nominally similar to their IT projects of the last fifteen years, but in fact are completely different. Gerstner described the problem / opportunity in the book “Who Said Elephants Can't Dance?” :

If the strategists were not mistaken and the cloud would indeed become a place for all such interactions, two revolutions would occur - one in data processing and the other in business. The world of data processing has changed, because the load was transferred from personal computers and other so-called client devices to large systems within companies and to the cloud itself (the network). This changed the look at personal computers as a center for innovation and investment, with all the obvious implications for IT companies that have made a fortune on PC technology.

More importantly, the global connectivity that the cloud symbolized would revolutionize the relationships of millions of companies, schools, governments and ordinary consumers. She changed commerce, education, health care, government services, etc. It triggered the biggest wave of business transformations since the advent of digital data processing in the 1960s ... Terms like “information superhighway” and “e-commerce” were not enough to describe what we were talking about. We needed other words to help the industry, our customers, and even IBM employees to understand what we saw behind access to digital information and online commerce. It was supposed to change the relationship and interaction between companies and people. In the end, our marketing and internet services offered the term “e-business”.

Older people will surely remember IBM's ubiquitous prefix "e-":


IBM spent more than $ 5 billion on e-business marketing. Gerstner called this investment "one of the best brand positioning" that he saw in his career. It worked because it was true: large enterprises, most of which always interacted with customers indirectly through a long chain of wholesale distributors and retailers, suddenly had the opportunity — even the need — to directly interact with end users. It can be as simple as a website, online store or customer support, not to mention the possibility of using all other parts of the value chain in real time. The technological challenges and business opportunities — the problems, if you will — were enormous, and Gerstner positioned IBM as a company that could solve these new problems.

This was an attractive proposition for almost all non-technology companies: the problem with the Internet in the 1990s was that the underlying technologies were very diverse and immature. Different companies offered solutions to each problem, many of them were startups with no experience working with large enterprises, and even with the best offer, they didn’t have IT departments capable of managing and integrating many suppliers. On the other hand, IBM offered universal services of the full spectrum (one throat to choke). They promised to solve all the problems associated with this new-fashioned "Internet". In addition, all IT departments were already familiar with IBM.

This strategy also made it possible to squeeze profits out of the value chain:



The actual technologies underlying the Internet have been discovered and transferred to the public domain. This meant that IBM could form an integration point and make a profit. This is exactly what happened: IBM's revenues and growth have steadily increased - often quickly! - over the next decade. The company managed everything from data centers to internal networks and operations of external sites and e-commerce. Naturally, the company controlled middleware for connecting the system together, and here it earned most of its profits. IBM took care of everything, gradually locking customers in its ecosystem. She again became fat and lazy.

When IBM lost the cloud

In the last paragraph of the book Who Said Elephants Can't Dance? Gerstner wrote about his successor, Sam Palmisano:

Yes, I have always been an outsider. But that was my job. I know that Sam Palmisano has the opportunity to create links with the past that I could not create. But he will have to do this without going back, realizing that the centrifugal forces that have made IBM a closed and self-centered company are still strong.

Palmisano failed miserably, and there is no better example than his announcement of 2010 with a development plan until 2015, which focused on promising to provide $ 20 per share in profit by 2015. Palmisano said then:

[The consensus is that] product cycles will stimulate industry growth. The industry is consolidating, and in the end consumer technology will destroy all computer science in the last 20 years. I'm a guy from the East Coast. We have a slightly different point of view. Product cycles will not be able to stimulate steady growth. Customers in the future will require a quantitative return on their investments. They are not going to buy fashion and trends. Enterprises will have their own unique model. In the cloud, it is impossible to provide services such as ours.

By this time, Amazon Web Services was running for four full years and two months. It was the height of folly not just to mock the idea of ​​the cloud, but to fix future profits in the face of an existential threat that implied huge infrastructure costs.

Gerstner pinpointed exactly what Palmisano was wrong: he was so “blinkered and self-absorbed” that he could not imagine a corporate solution better than that of IBM. He did not see the facts. Here is what I wrote in 2014 , when the company officially abandoned its 2015 profit plan:

The reality ... is that the companies that IBM served - and the sole reason for having an IBM business - bought customized technology solutions not for their own pleasure, but to achieve their business goals. The main idea of ​​Gerstner was that many companies had a problem that only IBM could solve, and not the intrinsic value of individual solutions. And since the universally provided cloud services were slowly but surely improving, IBM’s monopoly on problem solving disappeared.

Since then, the company has claimed for several years that it is seeking to catch up with competitors in cloud services, but the truth is that Palmisano closed this path when he was unable to invest in it ten years ago. In fact, one of the most important conclusions from the Red Hat acquisition is the recognition that IBM’s efforts in cloud services have come to naught.

IBM's torment

So what is the point of acquiring Red Hat, and how is it related to Lou Gerstner?

Well, firstly, IBM has not worked very well for a long time: last year’s annual income was the lowest since 1997, partly due to the Gerstner transformation. Of course, on this graph ZDNet need to consider that $ 79 billion in 1997 today is equal to $ 120 billion.


Source: ZDNet

At the beginning of this year, the company finally returned to growth after a recession for 22 consecutive quarters, but the numbers declined again last quarter: IBM's ancient mainframe business grew by 2%, and the traditional services business — by 3%, but technology services and cloud platforms remained at the same level, and cognitive solutions (ie, Watson) decreased by 5%.

Meanwhile, the mentioned commitment to the cloud is mainly an accounting fiction derived from the re-classification of existing businesses. It is more appropriate to look at the volume of capital expenditures of the company, which in 2017 amounted to $ 3.2 billion. This is less than $ 3.6 billion in 2016. Charles Fitzgerald writes in Platformomics :

We see that IBM’s capital costs are slowly falling, just like the company itself. IBM has always spent a lot on capital investments (up to $ 7 billion a year in a more glorious past), long before the cloud era, so one cannot assume that the absolute cost is associated with the cloud. The entire big three outstripped IBM in capital expenditures in 2012-2013. Resisting their rise, like all other cloud service providers, IBM simply doesn’t play cloud over-scaling.

Red Hat Acquisition

This is where Red Hat purchase makes sense. Of course, IBM is pleased to take over a profitable business with a subscription to RHEL, which generates real money, but the real prize is Openshift, a software package for creating and managing Kubernetes containers. I wrote about Kubernetes in 2016, in the article “How Google Challenges AWS” :

In 2014, Google introduced Kubernetes, the open source container cluster manager based on the internal Borg service, which abstracts Google’s massive infrastructure, so that any Google service instantly receives the necessary computing power without worrying about the details. The key element is the containers I wrote about in 2014 : engineers work in a standard interface that retains (almost) full flexibility. They do not need to know anything about the underlying hardware or the operating system (this is an evolutionary step compared to virtual machines).

Kubernetes differs from Borg in absolute versatility: the manager works on AWS, on Azure, on the Google cloud platform, on any infrastructure, you can even run it in your home. More importantly, this is the perfect antidote to AWS’s ten-day odds in infrastructure as a service: while Google has made great strides in its own infrastructure offerings, Kubernetes’s potential impact in particular on container development in general is that it doesn’t matter Which infrastructure provider are you using? Not surprisingly, this is one of the fastest growing open source projects of all time: it is not tied to any platform.

That's exactly what IBM is counting on, according to a press release announcing the deal :

This acquisition brings together best-in-class hybrid cloud solution providers and will allow companies to safely move all business applications to the cloud. Today, companies are already using several cloud services. However, research shows that 80% of business applications have not yet moved to the cloud. The proprietary nature of today's cloud market interferes with the portability of data and applications between clouds, data security in a multi-cloud environment, and consistent cloud management.

IBM and Red Hat offer a decisive solution to this problem, accelerating the deployment of hybrid multi-cloud solutions. Together, they help customers create cloud-based business applications faster, increase the portability and security of data and applications in multiple public and private clouds with consistent management. In doing so, they will rely on their leadership in key technologies such as Linux, containers, Kubernetes, cloud management and automation.

This is a serious move: in the 90s, the complexity of the Internet became a problem for enterprises and gave IBM the opportunity to sell solutions, and now IBM claims that having three centralized cloud providers makes it difficult for enterprises to choose. IBM is betting that, along with Red Hat, will provide a solution that easily integrates private data centers and all public clouds.

IBM's untrained mind

The best thing about this strategy is its pragmatism: ten years ago, IBM abandoned competition in cloud services, has simulated work in this direction for the last five years, and now finally recognizes that the best option is to build a business on top of all other clouds. However, this is also a weakness of the strategy: it seems more focused on the needs of IBM itself than of its potential customers. After all, if the enterprise is concerned about dependency on a single vendor, is IBM really the best option? And if the answer is “Red Hat is an open software”, then at what point will advanced enterprises build their own solutions?

The problem with IBM is that they do not create solutions for ignorant IT departments that are confused by a dizzying array of open technologies: instead, they build on top three cloud service providers, one of which (Microsoft) specializes in exactly the kind of hybrid solutions they target. Ibm. The difference is that Microsoft really spent money on infrastructure, so its ability to extract money from the value chain is correspondingly higher, and IBM must pay rent:



Perhaps the biggest problem goes back to Gerstner: the first thing IBM needs is rethinking its culture. What benefits the company will derive from the acquisition of Red Hat depends on a similar transformation. Unfortunately, this seems unlikely; The current CEO, Ginny Rometti, who took over the management of the company in early 2012, not only supported Palmisano’s catastrophic plan until 2015, but also actually implemented a significant part of the cuts and financial decisions before finally abandoning it in 2014. Meanwhile, the most famous marketing company is built around Watson, whose capabilities are greatly exaggerated. Not surprisingly, sales are declining after frustrating applications of the system .

Gerstner knew that change was difficult: he called the appearance of the Internet “successful” in terms of his time at IBM. But, as Louis Pasteur wrote: "Good luck favors only a prepared mind." Gerstner determined the strategy and began to change the culture of IBM, so when the problem arose, the company was ready. Today, IBM claims to have found a problem. It is not known whether the problem really exists. But even so, unfortunately, there is even less evidence that IBM is really ready to take advantage of the situation.