- Expectations are often too high – there is no „magic out of the box“
- Having data and having usable data are two completely different things
- The group of participants, the necessary process changes and the redefinition of responsibilities are massively underestimated
- Only holistic project planning and support secures project success
There is no magic – mathematics is just a natural science too
Many medium-sized companies from traditional industries are currently engaged in projects to digitalise their products and services. The core competence of the people involved typically lies in other areas, and so modern „magicians“ are often invited in to help painlessly and without complications.
Potential customers are sold buzzwords such as artificial intelligence (AI), neural networks, the Internet of Things (IoT) and so on – and the provider with the most convincing „buzzword bingo“ wins the contract! In this context, a not entirely serious English aphorism comes to mind that aptly describes the situation:
Difference between machine learning and AI:
If it is written in Python, it’s probably machine learning
If it is written in PowerPoint, it’s probably AI.
Unfortunately, in practice it still happens again and again that the sober description of the statistical-mathematical methods being used does not appear sophisticated enough for those who believe in technology. People sometimes expect real miracles that humans alone could never have come up with. In fairness, it must be said that there are indeed areas in which algorithms „see“ more than humans. However, the sometimes very poor data situation in the German SME sector reliably ensures that this is rarely encountered in practice; instead, one has to consider which questions can even be asked and how the volume of data can be enriched as quickly as possible so that anything usable can emerge at all.
Nevertheless, it seems that people are more receptive when simple answers are given to complex problems. So here, too, marketing is everything!
Mathematics solves problems, but it is not a sales argument.
Data – the oil of the future or „shit in = shit out“
When you come to a new client as part of a digitalisation project, you often hear „We have all the data!“. On the surface, this statement is also correct at first. But if you then follow the entire process chain, you find data in the most varied forms:
- Machine data (raw data)
- Pre-processed machine data (averaged values)
- Data from ERP, MES and SCADA systems
- Excel spreadsheets
- Proprietary formats
- Notes in running text (electronic)
- Handwritten notes (e.g. shift logs)
- Information passed on verbally
This data is then usually also stored at the most varied intervals. This can lead to evaluation problems when, for example, data from source X, which records a value every minute, has to be reconciled with data from source Y, which does so every millisecond. It is very likely here that the reference value is missing at most points. Solutions then have to be found to determine these mathematically – with all the inaccuracies that can arise.
A further challenge lies in harmonising the data formats. As long as you are dealing with electronically available data, solutions can still be found relatively easily. It always becomes difficult when records that are understandable to humans – such as running text – have to be brought into a format that computers can evaluate. While a human has the gift of decision-making and the ability to draw on experience or even gut feeling when questions are unclear, processors need clearly defined codes for classification.
The main challenge is therefore to identify and eliminate these potential stumbling blocks in advance of the digitalisation project.
Even when this has been done, examples sometimes arise in practice that show how business processes can seriously thwart the project goal. Not too long ago, for instance, a client had the objective of determining which process parameters in production lead to defective parts. A good example of the benefit of data analysis. Unfortunately, the part numbers of the rejected parts were not recorded, nor was the reason for rejection captured. This clearly shows that, even when all the necessary process parameters are available, no conclusion about the causes of quality defects can be reached.
Because too little or poor data does not lead to the goal.
Algorithms do not solve organisational problems
So far, the focus in digitalisation projects in companies has been placed disproportionately on the technical implementation and the algorithms used. The probably most success-critical factor – the human being – is often overlooked in the process.
In addition to the methodological and technical obstacles described above, many workflows in the company have to be examined, processes evaluated and often also changed. In digitalisation projects, you therefore need people from different areas of the company who would not normally work together in this way.
This can be both a curse and a blessing. A blessing in that a completely new and holistic view of the company emerges across departments and divisions. A curse because, in order to implement changes quickly, decision-makers at the highest level are suddenly needed – people who, in our experience, do not see themselves in the trenches of project work.
The set-up of the project group should therefore be carried out in advance by people who have already gained experience in interdisciplinary projects. Also indispensable is the binding support of top management to bring about the necessary organisational changes at short notice, and also to contribute operationally here and there. This also applies to the representatives of employee co-determination.
A good group is therefore made up of the following people:
- Those responsible within the company
- Technologists from production
- Those responsible for quality and process management
- Those responsible for the ERP system
- Those responsible for shop-floor IT
- Those responsible for corporate IT
- Those responsible for IT security
- Management and co-determination
- Middle management
- Top management
- Works council
- An experienced project partner
- with profound technical knowledge
- with experience in the field of digitalisation projects
- with experience in management and process consulting
- who also offers sustainable project management
When you keep in mind the number of areas and people involved, it quickly becomes clear that the external – but above all the internal – project effort should by no means be underestimated.
In particular, because of the need for process changes, possibly across the entire value chain, the speed of implementation – and thus the cost risk – does not depend on the technical realisation, but on the company’s ability to shape change processes.
The commitment of top management is essential for project success, because short-term decisions also have to be implemented promptly.
A good plan is an advantage – the Digital Success Road Map
The problems and challenges described so far are not isolated cases. Even larger companies with well-established project management are not immune to mistakes when dealing with digitalisation projects.
Project success depends largely on the human component. A often tough struggle over responsibilities and competencies is the order of the day. Only the necessary openness with one another and a collegial spirit lead to results.
An approach further developed from software development – the Digital Success Road – takes into account all the success-critical facets of a digitalisation project. Good planning is the be-all and end-all, and the smoothest possible implementation guarantees success. After all, implemented digitalisation projects are often an important driver for achieving strategic corporate goals – they strengthen the resilience of companies in increasingly volatile markets and provide the decisive competitive advantage.
Pure IT consulting without accompanying strategy and process consulting and close project support does not lead to the desired goal.
