Optimizing software solutions: Six success factors

Technical report by Dario Waechter

Corona crisis as a driver of digitalization

The current coronavirus pandemic has not only turned the world upside down, but has also revealed shortcomings in many areas: from unfavourable dependencies in supply chains to a lack of digitalization. Many German companies have some catching up to do when it comes to digitalizing their processes. In this respect, the crisis can certainly be seen as a positive: It has (finally) shaken companies awake. Never before has there been a greater awareness that digitalization projects are actually necessary. This fact will lead to companies pursuing digitization initiatives with higher priority and greater seriousness than before. Software manufacturers and developers now have the opportunity to optimize their solutions for this need, taking into account the latest technical developments so that their software can better support user companies in their digitalization efforts. Dario Waechter, Head of Data & Analytics at atlantis media, presents six current approaches and trends that make software solutions even better.

1st trend: platform economy

2020 is the year of platform solutions. It is definitely time for modern business software to integrate seamlessly into existing system landscapes. After all, this is the fundamental prerequisite for setting up cloud-based ecosystems that are tailored to requirements. In order to put the idea of the platform economy into practice, software manufacturers must pursue a platform strategy on the one hand. On the other hand, they need to think in terms of holistic end-to-end processes instead of just mapping sub-processes - as is often the case at the moment. Ideally, the processes are supported by corresponding algorithms and AI-based components.

2nd trend: customer experience

The need to provide customers with a positive shopping experience is, of course, no longer a new trend. However, in the context of customer experience, many companies have recognized that the pure service character of business software has been neglected in recent years. There is still a lot of room for improvement here. Good business software must have the necessary features to enable users to optimally control and analyze the customer experience. It is also conceivable to specifically expand existing software solutions in the direction of service and service processes. For example, the options for integrating various social media channels as required should be significantly improved.

In the CRM environment in particular, the trend is also moving more and more towards customer experience. Against this background, a good CRM system must have two things: Firstly, it must offer the functional options to individually control and reliably evaluate the customer experience. Secondly, it must be able to link seamlessly with other systems in order to create needs-based ecosystems. After all, an optimal customer experience can only be ensured if sales, marketing and service can work together smoothly and exchange data with each other. To make this as easy as possible for users, CRM system providers should also pursue a platform strategy and support holistic end-to-end processes. In addition to the standard CRM basic functions, CTI (Computer Telephony Integration) functionalities and AI-supported processes in the areas of lead and opportunity management, predictive analytics and route planning are increasingly in demand.

3rd trend: augmented analytics

In the area of data analytics, the focus in 2020 will be on simplifying data analysis. Augmented analytics plays an important role here: it automates and accelerates the processes of data collection, cleansing and evaluation. Instead of numerous tedious and error-prone work steps - from database queries to the consolidation of results - machine learning processes and algorithms are used that independently examine data sets, recognize patterns and anomalies and immediately derive causes, hypotheses and recommendations for action. Augmented analytics thus relieves the burden on data analysts and specialist departments in particular. They can make informed decisions and take action more quickly. The challenge here is to ensure a high-quality database on a permanent basis. After all, augmented analytics is only as good as the data that is available for analysis.

4th trend: Natural Language Processing

Natural Language Processing (NLP) is a process that allows data to be made searchable using simple questions. Against the background of the increasing popularity of voice control, NLP is now also increasingly coming into focus. The combination of linguistic knowledge with modern computer science and artificial intelligence allows communication from human to computer to be as natural as possible. As a wide range of questions is now available, NLP-based systems are able to deal with much more complex issues. Although most tools currently only support the English language, the range of languages will expand in the future. Users will then be able to simply "google" important questions in their native language.

5th trend: Opening up the systems

Another important trend is that more and more systems are opening up to open source-based AI and machine learning platforms as well as programming languages such as Python. Python is used by data scientists in particular to analyze data. By opening up their systems, providers can give users access to relevant analysis models and create new models. The open source market for artificial intelligence offers a wide range of opportunities in this respect. By integrating such platforms, manufacturers can improve their solution and thus the analyses, which in turn is essential for users to remain competitive.

6th trend: Explainable AI

In order to really benefit from AI-based models, companies must be able to correctly assess the results delivered. In this context, progress can be expected above all in the field of Explainable AI. Explainable AI makes models transparent and users gain a better understanding of what artificial intelligence does in a technical sense. On this basis, companies can better decide which models or AI solutions are suitable for their data analyses.