The decline of traditional IT vendors has become a trend. Not only in the context of China's move to IOE, but on a global scale, these once-prosperous IT predators are also declining. As an old employee who has served IBM for nearly 14 years, I would like to briefly talk about the decline of IBM and the entire traditional IT industry.
Many people blame Sam Palmisano for IBM's fall. After the famous Luis Gerstner retreated bravely, it was SAM that, in order to please Wall Street, adopted a method of operating the company without increasing sales but increasing gross profit margins, and then frantically compressing costs and reducing R&D expenses, resulting in IBM's lack of innovative products in recent years. Although SAM was very successful during his tenure, IBM took the lead among traditional large-scale computing equipment manufacturers through HP's stupor, but the consequence was that the entire IBM was immersed in bubble-like blind optimism. Before leaving office, SAM frantically proposed the so-called 2015 plan, which led IBM to continue the strategy that had led to its lack of innovation capabilities.
All this is of course SAM's problem. What we can't avoid is that the big traditional IT equipment manufacturers are not having a good time right now. CISCO, HP, HDS, EMC, and even VMWare, which has been shining brightly in recent years, are gradually feeling the crisis. This is not a personal problem for Sam, but an industry problem. This industry is traditional enterprise-level IT system product manufacturing and service providers.
If it is an industry problem, the first thing we need to analyze is the business model and profit model of this industry. Take IBM as an example, because IBM may be regarded as the most comprehensive and comprehensive one in this regard. IBM's products are divided into software, hardware, and services.
IBM's business
IBM's software is mainly middleware software. In layman's terms, it is an indispensable support software for building a business system, including database and information management software (Information Management), application server middleware and software integration software (Application & Integration Middleware), software development software (Rational), collaboration software (Loutus), and system management software (Tivoli). (The process of IBM's fall is mainly discussed here, so it is divided according to the organizational structure before IBM's fall)
IBM's hardware mainly includes two categories of high-end enterprise-class servers and enterprise-class storage.
IBM's services are divided into two divisions: Technology Services and Consulting Services.
The main business models of the Technology Service Department are: 1. Warranty: it is the warranty of all IBM equipment, as well as value-added services based on warranty; 2. System integration services; 3. IT infrastructure outsourcing.
Generally speaking, the consulting service department is engaged in software development. At the same time, because it often involves the consulting design and development of some large-scale enterprise software such as corporate finance, ERP, supply chain, and customer relationship management, it is essential to involve some corporate strategy consulting. , management consulting, after all, a considerable part of the logic of these software is based on enterprise strategy and management strategy. The software development as its foundation is generally divided into two parts: one part is software development in our usual sense. That is, people ask for IBM GBS to help code. This is relatively basic software development. Another large part is the customized development of business suites, usually the customization of ERP, SRM, CRM, SCM and other software of enterprise business suite software manufacturers such as SAP and Oracle.
IBM also has some other relatively small businesses, such as financial services of IGF, implementation services based on IBM software of laboratory services, marketization of some innovative assets of research institutes and so on. None of these constitute mainstream.
IBM's business model
Whether it is IBM's hardware, software or service business, it is a relatively complex business. Some people may not understand, for example: "What's so complicated about a warranty? Isn't it just to quote a price for renewal, and after the customer buys it, if the equipment is broken, will it be replaced?" Familiar with the entire enterprise market. Let's put it this way: All IBM sales, without detailed calculations by engineers, simply cannot be priced, and literally not even the warranty price of a fixed known device. This is caused by the complexity of the enterprise-level equipment itself, which is very different from the online automatic quotation of direct websites in the consumer-level market.
Based on the characteristics of such a product and service, the sales model of IBM and all enterprise-level IT equipment manufacturers is complex. Such a complex sales model requires the support of a huge sales team, including sales and pre-sales engineers, and the help of channels, that is, agents, to complete the entire sales process. As a result, such a complex and huge team is bound to require huge profits to support. So that's a lurid discount rate: a device priced at $10 million might ship from IBM for $500,000.
What's the difference between enterprise-grade and consumer-grade devices
The difference between enterprise-grade equipment and software and consumer-grade products is that enterprises serve millions of users, while consumer-grade products serve only one person or a small group of people. This determines that the enterprise-level business system cannot fail. A securities trading system cannot go down so that people cannot buy and sell stocks; a banking system cannot fail so that people cannot deposit and withdraw money; the system of the State Taxation Bureau cannot stop and people cannot pay taxes. For a consumer-grade system, if it goes down, it will go down, and it will be fine after a while. For the stability of this system, enterprise-level systems must have numerous redundant hardware and software to provide reliability, maintainability, manageability, and sustainability for the entire business system. The combination of these technologies is ever-changing and must be designed and maintained by an excellent team of engineers. For example, if you build a kennel for your dog, you can do it yourself; if you build big pants for CCTV, if you don’t have a high-level team of architects, there will be no drama at all.
changes in the market
If the market doesn't change, the good times for companies like IBM will continue. But the market is never static. With the development of technology, changes in the market are also accumulating, and those companies that are not ready will be eliminated when the final outbreak occurs.
We all know that IT technology is developing rapidly, but we don't know whether the rapid development is good or bad for us.
I think that after Oracle acquired Sun, it released an explosive information that Oracle will no longer support servers with IA64 architecture. Why do you do this? At that time, there were only three types of enterprise-level database servers on the market, IBM Power series, HP IA64 architecture Superdome series, and Sun. Sun has already been acquired by Oracle, so you can't help but support yourself; IBM is the market leader, if you don't support it, the sales will drop directly. Let's get rid of HP first. Although this policy was not implemented in the end, it still had a great impact on the sales of the market. IBM's sales were crazy, and HP's servers languished. But when IBM was jubilant, some discerning people had foreseen the decline of IBM's minicomputers. When the market share of IBM minicomputers finally increased to 70%, which has almost become a one-man show in the market, IBM suddenly found that people no longer pursued the technology of minicomputers and databases. This is common sense. When you are the only player in the market, it is difficult to do it alone. Sure enough, in a few years, everyone's promotion of enterprise-level computing became less and less, the enterprise-level market grew weakly, and the cloud computing market grew rapidly, so IBM was a little frustrated.
Why does the cloud computing market bring so much change to the enterprise market? With cloud computing, is there no need for enterprise-level computing? To answer these questions, the first thing to discuss is the changes in the main business of enterprises in the process of economic development. Everyone has learned the definitions of the primary industry as agriculture, the secondary industry as industry, and the tertiary industry as service industry. There is no doubt that agriculture is what meets the most basic survival needs of human beings. With the development of the economy, industry replaced agriculture as the largest industry, and gradually, the service industry replaced industry as the largest industry.
Computing requirements go through a similar evolutionary process. Initially, businesses used computing to complete accounting system calculations, or record changes in account information. The change of account information undoubtedly requires a very strong computing system. The strong here does not mean large, but the stable, manageable, continuous, etc. described above. This is the need for so-called enterprise-class computing.
And enterprise-level computing needs aren't that much per se. The proportion of core business in an enterprise's data center is only 1/10 to 1/5. There are many businesses that do not need any enterprise-level computing power at all. For example, there are many off-site illegal processing systems in the computer room of the traffic police team. In fact, the problem of downtime is not a big problem; there is also a traffic guidance sign system, which is the one A system that displays traffic information in a city; I once saw a lottery system in this office from a client; many bank branches have some small SMS publishing systems; now there are big data analysis systems, and so on. Today's computing requirements are far greater than the scope of the original core business systems, and a large number of non-core business systems do not require enterprise-level computing capabilities. However, in the past, the construction of these non-core business systems basically followed the architecture of the core business system, so it constituted a market with puffy enterprise-level computing requirements. The development of cloud computing enables these non-core business systems to have a set of solutions built on non-enterprise systems, and even some small core business systems can be built on cloud computing platforms. The market demand for enterprise computing architecture collapsed on a massive scale. Although the computing architecture requirements for the core business of enterprises are still growing, such as the substantial upgrade of the bank's core accounting system caused by Taobao's Singles' Day, the growth rate has been unable to support the increase in sales of enterprise-level computing equipment.
To summarize: Just as the development of agriculture, industry, and services rotates, the development of computing needs has gone through a similar process. Non-core businesses now dominate the computing equipment and services market.
Let's go back and analyze how IBM's products and services can't keep up with the form of the market.
First look at the hardware: IBM's mainframes, Power servers, and high-end storage are undoubtedly serving enterprise-level computing architecture; IBM has sold X86 servers to Lenovo; so IBM hardware no longer has products suitable for current mainstream computing development needs . Some people may ask: In this case, is it a foolish move for IBM to sell X86 servers to Lenovo? No, although the cloud computing market is mainstream, it does not mean that X86 servers will generate high profits. IBM's entire marketing system is built around the enterprise-level sales method. This marketing system is very effective for the sales of enterprise-level computing needs, but it is too complicated for low-end equipment, which will bring heavy burdens to X86 servers. Manage overhead. So selling the X86 was the right decision, and it was in line with IBM's strategy of pursuing high-margin products and services.
The software division has been IBM's main profit growth point over the years. But if we take a closer look at IBM's software, we can see that all but mainframe software can find alternatives to free products in the open source market. This means that those non-mainstream businesses may gradually abandon IBM's software.
IBM's software and hardware features are the software and hardware of the supporting architecture for applications. That is to say, no specific application is used, which means that the user's final business needs cannot be controlled and can be replaced at any time. This is a good strategy, avoiding competition with SAP and Oracle ERP, and refining one or several aspects of the business system. But the problem is that when the overall market demand in this segment declines, there must be enough R&D reserves to cope with the new demand, rather than taking the bone marrow from the already declining demand. In terms of R&D reserves, IBM has done too badly over the years.
Several major businesses of the Science and Technology Service Department: the warranty business will naturally fail to thrive with the decline in software and hardware sales; the outsourcing business mainly deals with the relatively stable core business of large enterprises. Non-core businesses with more changes in enterprise IT cannot be negotiated because they cannot be priced. The integration business of the Science and Technology Service Department is a solution for more complex enterprise-level systems. This part is also decreasing.
Similarly, the consulting service department, the large-scale ERP construction period has passed, and of course it will continue and develop, but the development speed has slowed down. The supporting enterprise E-Business project is still going on, the scale is not as big as before, and some enterprises have already started to use cloud computing to do it. Well-known virtualization software maker Citrix uses Salesforce entirely for sales management.
IBM has no shortage of elites, senior leaders have long recognized these issues, and IBM has already started the layout of cloud computing as early as 2007. So why is it still not keeping up with the rhythm of cloud computing? As with the decline of every empire, the presence and survival of vested interests is at the root of the decline. IBM's senior management has been occupied by these vested interests. They can only survive and make a profit with the original business model. They are embracing the new business in a high-sounding manner, but in practice they simply cannot break away from the old operating model. Wall Street also played a very bad role in the process. Financial analysts are ruthless, they only look at statements. As a result, IBM executives who have no guts can only smash their bones in traditional businesses, because the development of new businesses cannot be achieved overnight and cannot immediately satisfy the appetite of Wall Street. At the same time, these people also hope that they can survive the days when their interests are realized, and then pat their butts and leave. So, like the decline of every dynasty, everyone died in a wise decision.
How did IBM come out of that crisis in the late 80s and early 90s? Are there any ideas to learn from? We go back to the late 80's. During that crisis, Hitachi, Siemens, NCR, Wang An... all the mainframe companies went bankrupt, and only IBM remained. The essence of that crisis was the mainframe crisis, and it was not fundamentally different from this one. In that era, the mainframe was used to run the core business of the enterprise, while the minicomputer was used to run some non-core business. The mainframe market experienced an avalanche of decline with the development of minicomputers and the shift to minicomputers at the core of their mainframe-adapted business, and none of these mainframe system makers were ready to accept this shift in market demand. Why in the end IBM can only survive? Because a Luis Gerstner came, he didn't understand IT at all, so he could only do things according to the market demand. He took the initiative to embrace the new market demand. The vigorous development of IBMAS400/RS6000 and its supporting software saved IBM.
Luis Gerstner did several important things: 1. Selling houses, land and famous paintings, improving IBM's cash flow a little (only a little); 2. Large-scale reorganization of IBM, independent software department, independent service department; 3. Increase support for AS400/RS6000 minicomputers.
The first and third clauses are easy to understand, but the second clause is a bit incomprehensible. Why would the independence of the software department and the independence of the service department improve IBM's business performance? Those who know IBM's history know that in that era, IBM software was software attached to IBM hardware, mainly DB2, CICS, MQ, computer language compilers (C, COBOL, RPG) and so on. The hardware is mainly based on mainframes, so all software is oriented to IBM mainframes, and there is no idea of its own development. After independence, the software department has its own development strategy. It acquired encina and developed CICS on minicomputers, as well as the development of DB2 and MQ on minicomputers, and the development of integrated middleware. It acquired software companies such as Lotus and Tivoli to develop enterprise collaboration. Software and systems management software, and later acquired Rational. These strategies have made IBM Software Packages IBM's most profitable division today. The service department was originally attached to the mainframe's warranty department and had no right to speak. After independence, the service department has made a good development in value-added services, and gradually developed outsourcing services and integrated services. After the consulting service department became independent, it started consulting and implementation services for ERP, the core business of the enterprise, and is currently the top three in the implementation of ERP software in the world. In fact, the reason is very simple. When a person or department is dependent on other people or departments, his creativity cannot be exerted. After independence, neither the software department nor the service department had the original burden, adapting to the development of the market and becoming the winner in the market.
Does IBM's current CEO, Ginni Rometty, have the courage and ability of Luis Gerstner? This question can only be answered by time. Just a brief description of what I've seen.
When Luis Gerstner took office, there was a more favorable situation than the current Ginni, that is, IBM in 1993 was on the verge of dying. The whole Wall Street hoped that what Luis did was to break up IBM. IBM has reached the point where it can no longer be bad. At this time, taking over IBM, Luis Gerstner can do whatever he wants, and there is no group of vested interests to stop his innovation, and there is no interest. The situation facing Ginni is different. She takes over the brilliant record and 2015 plan left by Sam, and there are a bunch of vested interests around her, including some IBM executives and Wall Street chaebols. Ginni couldn't use his punches effectively, even with innovative ideas. In general, some of the moves Ginni has made since he took office are good: selling X86 servers and low-end hardware business; selling semiconductor business; aggressively entering cloud computing, including the acquisition of SoftLayer, the establishment of an independent cloud computing department; according to new market needs Restructure IBM's organization. Success depends on her ability to execute, and a clear example of resistance is CMS.
CMS - Cloud Management Service - formerly known as SCE+ (Smart Cloud Enterprise +), is an extension of IBM's earliest public cloud brand SCE (Smart Cloud Enterprise). SCE turned out to be a total rubbish fail product. SoftLayer was abolished after IBM acquired it, but SCE+ remained and was renamed CMS. CMS is also a downright rubbish product. Have you heard of any public cloud product that takes more than 20 days to deploy a virtual machine system? CMS is such a thing, it is actually too embarrassing to call it a cloud! ! ! CMS is a set of public cloud platforms developed in accordance with the concept of IT strategic outsourcing. It is a bad thing developed completely behind closed doors. I guess the main reason why IBM continues to promote it is that the investment in it is too large, and I am embarrassed to stop and beat myself in the face. It is said that CMS sells well in the European and American markets. I don't know if it is true or false. But in any case CMS can't make waves because it's too small. Only 10,000 virtual machines have been deployed throughout China, and they cannot be sold. What is the concept? A virtual machine is a vCPU, an Intel Core can be divided into 4 vCPUs, a two-way 8-core server can be configured with 64 virtual machines, and about 160 servers for 10,000 virtual machines are enough. Just 3 to 4 cabinets. Now Amazon, Microsoft, Ali, Kingsoft, Tencent, any cloud is not in the scale of tens of thousands to hundreds of thousands of servers! What's ridiculous is that IBM is still selling this CMS in the traditional enterprise-level system sales model, that is, seeing a group of sales from a customer. That point of turnover is not enough for these sales to take a taxi to eat.
The way to provide services on the Internet is: Laozi's products are good enough, if you love them or not, I don't have the energy to serve them one by one. Simply put, there are two: 1. The product is absolutely standardized, and it is the same for anyone; 2. The product is sufficiently competitive. The sales method of the enterprise-level system is: the customer is God, the fart is fragrant, and the customer satisfaction is the top, I will give you what you want. The puffy non-core business needs of the past no longer exist, and it is obviously inappropriate to continue to sell non-core business support environments in the same way as selling core business support equipment. It is not ruled out that IBM can come up with different sales methods, but it is by no means the same traditional sales method as IBM used to.
Although some sales directors are also aware of these problems and take some fine-tuning methods to reduce the proportion of sales costs, for example, orders less than a certain amount are not considered sales performance, but the general direction is wrong, and only fine-tuning will not change Nature.
Another IBM cloud brand, Bluemix, is a PaaS cloud platform. It's a great idea, and IBM has the technical skills to do it well. But on the other hand, some related executives are not very reliable. They have become a kind of bureaucracy. Bureaucrats generally don't want to do a thing well, but just want to meet KPIs, and doing business well is just a by-product of meeting KPIs. So we saw some ugly faces and decisions.
The example is no longer cited whether IBM still needs to be rescued when it is dying. Let time decide. In fact, the entire traditional IT industry is facing similar problems. HP has been playing with the public cloud Helion for a few years, and recently announced that it is abandoning Helion. The fundamental reason is that the traditional IT sales method and the new cloud computing sales method are completely different, requiring different management methods and corporate culture. When an old company needs to rejuvenate, it needs the ability to change. This ability to change is not just about developing a new product, but it needs to overcome many difficulties. The corporate culture and the internal obstacles of the management organization may be the biggest. When IBM acquired SoftLayer, our colleagues chatted, and many people held a view: if IBM let SoftLayer develop itself like EMC managed VMWare, SoftLayer still has a future. Sure enough, SoftLayer CEO Lance Crosby left IBM less than two years after the acquisition.
Whether Ginni can lead IBM out depends on her ability to shake off these traditional forces that don't adapt to new markets. Some people have said that we have to wait until 2016, because most of the main vested interests will cash in in 2016, and maybe there will be less resistance after that. 2016 is almost here, let's wait and see. As for the entire traditional IT industry, I can only say that there are talents and talents in the country, each leading the way for decades.
Author: Duke Yu
Article source: Sina Weibo