Why traditional software is on its last legs

The companies discussed in the survey carried out by AME Research are ADP Nederland, Afas, Exact, Unit4, SAP Nederland, Oracle Nederland and Visma|Raet.

It not difficult to explain why the growth of these traditional software companies is stagnating. Their innovation is considerably hampered by the way in which they develop their software. Nowadays this can be done eight to ten times faster with a low-code platform.

Copying from other industries

With regard to low-code, I see many similarities with other high-tech sectors, such as the automotive and aerospace industries. They frequently use virtual models, for instance to test new designs virtually, to rule out any surprises later on. CAD-CAM is good example of this. You produce a digital design, which can be extensively evaluated and tested before it is manufactured as an actual product. In effect, this is exactly the same methodology that we nowadays apply with low-code, which is almost impossible with traditional software development.

Software as a catalyst

As a result of traditional development techniques, the functionality of the software is embedded in millions of lines of programming code. This code is complex and over time it becomes outdated, which means that this software must be periodically redeveloped. We transitioned, for instance, from DOS to Windows, the web, various mobile platforms…. And for all these platforms the software had to be redeveloped time and again in a very traditional manner. This is not only expensive and risky. It is now happening more frequently, because in this age of digital transformation new technologies and innovations follow each other in increasingly rapid succession. This means that we need more and more of these traditional software craftsmen and these developers are simply not available.

Make software stress a thing of the past

In order to increase their development speed, some companies opt for offshoring, whereby they purchase cheaper development capacity abroad. However, this brings along language problems and cultural differences and the development process remains traditional. In essence, you purchase your problem at a cheaper rate and that is not very ambitious. Another solution is to make use of package software, but the processes are still traditional. Furthermore, there are hardly any package software vendors that have survived more than 20 years of technological development. Another major disadvantage of package software is that you often pay for a lot of features for which you have absolutely no requirements. The only solution to escape from this quagmire of traditional software development is to automate the process. This is exactly the purpose of model-driven low-code software development. The functionality of the software is recorded in visual models and, based on these models, software is created automatically. The development process is much faster because modifications are made in the model and not in complex programming code.

You just have to take into account that automation is domain-specific. After all, you can’t build an airplane in a car factory either. So ask yourself before you start automating; which low-code platform is most suitable for my domain?

Low-code is disruptive for IT

For decades, other industries have been automated through the application of IT. It definitely brought improvements for companies, but it sometimes had major consequences for the work of the employees. Low-code is also disruptive for IT itself. This is certainly not a negative aspect, but experts will have to get used to the fact that their work is changing and, incidentally, will often become more enjoyable.

Actually, we should split the world of developers into two categories. With a significant shortage of technical developers, we should use them to create low-code platforms and not to develop apps. These are, for example, the traditional Java and C# programmers.

The end products should then be produced by business IT specialists or business experts. They are less scarce and will become ten times more productive. This will give a huge boost to the available software development capacity. In addition, this allows the business to be in the driving seat for the project. The typical ivory IT towers no longer exist.

Positioning of the various low-code platforms

Core systems, such as ERP, are very rigid and that is why companies often build additional solutions in Excel or Access to satisfy their requirements. It is perfectly common practice for a company with 10,000 employees and a large ERP package, such as SAP, to have no less than 500 to 1,000 peripheral applications. Of course, this was not initially the intention, but it is how things have grown over the years. A low/no-code platform can be the solution to prevent this proliferation of apps. There are a number of ways in which an organization can deploy low-code/no-code.

1. No-code platforms – These allow companies to build simple apps without any programming. This frequently leads to a love-hate relationship with IT when it comes to responsibility for the management of these apps.
2. Low-code platforms – These are used by software vendors (and sometimes companies themselves) to add apps to an ERP application. These apps are often somewhat larger and more complex and usually have a lifespan of several years. It is also possible to use a System Integrator (SI) that develops business software, based on your specifications, using a low-code development platform.
3. Low-code for core applications – This is used to build an application that actually replaces the traditional core system. What is also important here is that this application can automatically transform to new technologies because it has a life span of decades. That means this is a structural solution that creates enormous flexibility for further development of the software and allows you to adapt it to meet your requirements.

To pursue this last option you must have a clear vision of what you want as an organization. This is also true to a lesser extent for the second option and is rarely or not applicable for the first option. The companies that have such a clear vision are often those companies that want to make a difference in the market and need a unique distinguishing application.

If you want to replace your core system based on your own vision, it is possible of course to do this on your own. You can already do this with a small team consisting of just a few people who don’t even need to be programmers. You can also outsource it to a System Integrator.

If you already have a solution that you want to continue to use, it may be useful to draw the attention of the solution vendor to low-code for ERP. If they can develop ten times faster (which is possible with low-code), it will immediately benefit you as a customer, because functionality will become available earlier and will automatically remain technologically up-to-date.

Ask for advice and avoid selection stress

There is one more option, which is in my opinion the worst one: do nothing… Just keep on using traditional software and ignore the low-code revolution. Everything will remain the same, while the costs for maintenance and further development will continue to increase. This is certainly not what I would recommend, because in the end this can only have one outcome: your company will cease to exist in a few years’ time. However, it does not need to be this way, because there are enough options for you to apply low-code. Making a choice is easier than you think. As long as you make sure you get good advice!

If you want to know more about my view on low-code, watch this conversation with Digital Waves.