What is SAP? And why is it worth $163 billion?
Every year, companies spend $41 billion on software for enterprise resource planning, known by the acronym ERP. Today, almost every large business has implemented some form of ERP system. However, most small companies typically do not purchase ERP systems, and most developers likely have not seen them in action. So for those of us who have not used ERP, the question arises... what's the deal? How does a company like SAP manage to sell ERP worth $25 billion a year?
And how did it happen that , including 78% of food supply, is processed through SAP programs?
ERP is where companies store their core operational data. We are talking about sales forecasts, purchase orders, inventory levels, and the processes triggered by these data (for example, payments to suppliers upon order placement). In a sense, ERP is the "brain" of the company—it holds all important data and all actions initiated by that data in workflows.
But before fully taking over the modern business world, how did this software actually come to be? The history of ERP begins with serious efforts to automate office activities in the 1960s. Previously, in the 1940s and 1950s, the focus was mainly on automating the mechanical work of "blue-collar" workers—remember General Motors, which created its automation department in 1947. However, the automation of "white-collar" work (often using computers!) began in the 1960s.
Automation of the 60s: the emergence of computers
The first business processes automated using computers were payroll calculations and invoicing. Previously, entire armies of office workers manually recorded employees' working hours in ledgers, multiplied them by the hourly rate, and then manually deducted taxes, benefit withholdings, and so on... all just to calculate one month's payroll! This labor-intensive, repetitive process was prone to human error and was perfectly suited for computer automation.
By the 1960s, many companies were using IBM computers to automate payroll calculations and invoicing. Data processing is an outdated term, remaining only as a company name. . Instead, today we use the term “IT.” At that time, the software development industry had not yet formed, so IT departments often hired analysts and trained them to code on the job. Purdue University opened the first Computer Science department in the U.S. in 1962, and the first graduating class in the field followed a few years later.
Writing programs for data automation/processing in the 1960s was a challenging task due to memory constraints. There were no high-level languages, standardized operating systems, or personal computers—only large, expensive mainframes with limited memory, where programs ran on magnetic tape reels! Programmers often worked with the computer at night when it was free. For companies like General Motors, it was common practice to write their own operating systems to get the most out of their mainframes.
Today, we run application software on several standard operating systems, but this was not the case before the 1990s. 90% of all software was custom-built, with only 10% sold off the shelf.
This situation profoundly influenced how companies developed their technologies. Some believed that the future lay in standardized hardware with a consistent OS and programming language, such as for the aviation industry (which is still in use today!). Most companies continued creating their own fully isolated software, often reinventing the wheel.
The birth of standard software: the extensible SAP program.
In 1972, five engineers left IBM to secure a software supply contract with a major chemical company named ICI. They founded a new company called SAP (Systemanalyse und Programmentwicklung, or 'System Analysis and Program Development'). Like most software developers at the time, they primarily engaged in consulting. SAP employees would work in clients' offices and develop software on their computers, mainly for logistics management.
Business was going well: SAP finished its first year with a revenue of 620 thousand marks, which is just over $1 million in today's dollars. Soon, they began selling their software to other clients, porting it to various operating systems when necessary. Over the next four years, they gained over 40 clients, revenue increased sixfold, and the number of employees grew from 9 to 25. Perhaps this is a bit far from , but SAP's future looked promising.
SAP's software was special for several reasons. At that time, most programs operated overnight and printed results on paper tapes, which you would check the next morning. In contrast, SAP programs operated in real time, with results displayed not on paper but on monitors (which at that time cost around $30,000).
But most importantly, SAP's software was designed to be extendable from the very beginning. In the initial contract with ICI, SAP did not create software from scratch, as was customary at the time, but wrote code on top of a previous project. When SAP released its financial accounting software in 1974, it initially planned to develop additional software modules on top of it in the future and sell them. This extensibility became a defining feature of SAP. At that time, interaction between client contexts was considered a radical innovation. Programs were written from scratch for each client.
The importance of integration
When SAP introduced its second production software module in addition to the first financial module, the two modules were able to easily interact with each other, as they shared a common database. This integration made the combination of modules significantly more valuable than merely having two separate programs.
Since the software automated certain business processes, its impact largely depended on access to data. Purchase order data is stored in the sales module, product inventory data is stored in the warehouse module, and so on. And because these systems do not interact, they need to be synchronized regularly, meaning that a worker manually copied data from one database to another.
Integrated software solves this problem by facilitating communication between the company's systems and enabling new types of automation. Such integration—between various business processes as well as data sources—is a key feature of ERP systems. This became especially important as hardware evolved, opening up new automation opportunities—leading ERP systems to thrive.
The speed of information access in integrated software allows companies . Compaq implemented a new 'just-in-time production' model using ERP (meaning assembling a computer only after receiving a specific order). This model saves money by reducing inventory levels, relying on rapid turnover—precisely what effective ERP facilitates. When IBM followed suit, it reduced component delivery time from 22 days to three.
What ERP Really Looks Like
The term 'enterprise software' is not typically associated with a stylish and user-friendly interface, and SAP is no exception. The basic SAP installation contains 20,000 database tables, 3,000 of which are configuration tables. Within these tables, there are about 8,000 configuration decisions that must be made before the program can even be started. This is why is a real profession!
Despite the complexity of setup, SAP ERP software delivers key value through extensive integration of multiple business processes. This integration leads to thousands of use cases within the organization. SAP organizes these use cases into 'transactions', which represent business actions. Examples of transactions include 'create order' and 'display customer'. These transactions are structured in a nested catalog format. Therefore, to find the 'Create Sales Order' transaction, you would navigate to the 'Logistics' catalog, then 'Sales', followed by 'Order', where you will find the actual transaction.
If we refer to ERP as a 'transaction browser', it would be an incredibly accurate description. It functions much like a web browser, featuring a 'Back' button, zoom buttons, and a text field for the 'TCodes', which is the equivalent of the address bar in a browser. SAP supports , making navigation through the transaction tree complex without these codes.
Despite the dizzying array of available configurations and transactions, companies still encounter unique use cases that require fine-tuning their actions. For handling such unique workflows, SAP has a built-in programming environment. Here's how each part operates:
Data
In the SAP interface, developers can create their own database tables. These are relational tables much like standard SQL databases: different types of columns, foreign keys, value constraints, and read/write permissions.
Logic
SAP developed a language called ABAP (Advanced Business Application Programming; originally Allgemeiner Berichts-Aufbereitungs-Prozessor, which translates from German as 'General Report Creation Processor'). It allows developers to execute custom business logic in response to specific events or on a schedule. ABAP is a richly syntaxed language, containing about three times as many keywords as JavaScript (see ). When you write your program (SAP has a built-in editor for programming), you publish it as your own transaction, along with an individual TCode. You can customize existing behavior using an extensive system of hooks called 'business add-ins,' where the program is set to run when a specific transaction is executed—similar to SQL triggers.
UI
SAP also comes with a UI builder. It supports drag-and-drop and includes convenient features such as generated forms based on the database table. Despite this, it is quite difficult to use. My favorite part of the builder is drawing table columns:
Challenges of Implementing ERP
ERP is not cheap. A large multinational corporation can spend between $100 million and $500 million on implementation, including $30 million in licensing fees, $200 million for consulting services, and the rest on hardware, training managers, and employees. Full implementation takes between four to six years. stated: 'The competitive advantage in the industry will belong to the company that can carry out the SAP implementation better and cheaper.'
And it's not just about the money. Implementing ERP is a risky venture, and results can vary significantly. One of the success stories is the ERP implementation at Cisco, which took 9 months and $15 million. In contrast, the implementation at Dow Chemical cost $1 billion and took 8 years. . As many as believe that implementing ERP systems carries a 'moderate chance of harming the business.' You rarely hear that when evaluating software!
The integrated nature of ERP means that its implementation requires the efforts of the entire company. And since companies only benefit after widespread implementation, this is especially risky! Implementing ERP is not just a purchasing decision: it is a commitment to change one's operational management methods. Installing software is easy; reconfiguring the workflow of the entire company is where the real work lies.
To implement an ERP system, clients often hire consulting firms like Accenture and pay them millions of dollars to work with individual business units. Analysts determine how to integrate the ERP into the company's processes. Once integration begins, the company must start training all employees on how to use the system. Gartner reserve 17% of the budget just for training!
Despite all the challenges, most companies on the Fortune 500 list had implemented ERP systems by 1998: the process accelerated due to Y2K fears. The ERP market continues to grow today . It is one of the largest segments in the global software industry.
The modern ERP industry
is dominated by major players like Oracle and SAP. Although both are market leaders, their ERP products differ remarkably. The SAP product was mainly built in-house, whereas Oracle aggressively acquired competitors such as PeopleSoft and NetSuite.
Oracle and SAP dominate so much that even instead of its own ERP product, Microsoft Dynamics.
Since most industries have quite specific ERP needs, Oracle and SAP have ready-made configurations for many sectors, such as food, automotive, and chemical, as well as vertical configurations like sales organization processes. However, there is always room for niche players who typically focus on a specific vertical:
- for universities
- and McKesson offer ERP for healthcare organizations
- for manufacturing and logistics
Vertical ERPs specialize in integrations and workflows specific to their target market: for instance, ERP for healthcare .
However, specialization is not the only way to find a niche in the market. Some startups are trying to bring more modern software platforms to market. An example is : it offers integration capabilities (with various ERPs!) on a subscription basis. Startups like Anaplan and Zoho offer the same.
Is ERP on the rise?
In 2019, SAP is doing great: last year, revenue totaled €24.7 billion, and its market capitalization is now But the world of software is no longer what it used to be. When SAP first emerged, data was isolated and difficult to integrate, making storage in SAP seem like the obvious answer.
But now the situation is changing rapidly. Most modern enterprise applications (such as Salesforce, Jira, etc.) have a backend with good APIs for data export. Data lakes are being formed; for example, facilitates connections between databases that were impossible just a few years ago.
Source: habr.com
