Experts and exhibitors share lessons on data quality, AI and practical application.
- Successful deployment of data and AI starts with clear goals, reliable data quality and careful implementation.
- Without proper data management structure and disclosure, AI solutions often remain stuck in theory.
- New legislation such as the Data Act strengthens the position of users and encourages data sharing between platforms.
- Exhibitors at the Data Expo 2025 emphasized the importance of cooperation between technology and domain knowledge and used creative stands to show how data can be made practical and tangible.
Data is the new gold' has become a cliché in recent years. Yet this statement is incorrect, apart from the fact that 'data' is strictly speaking a plural form. In fact, if you want to make such a comparison, you should talk about 'data are rough diamonds.' Because the value of data does not automatically run out of the 'ore' when you heat it like gold.
While data collections can form the basis for wonderful successes, they must first be skillfully sorted by quality - that is, by the number of carats - and then processed before they yield valuable insights. And then the results need to be properly publicized so that an organization can enthusiastically use them and show them off. That was evident from several conversations with exhibitors on the show floor of Data Expo 2025, which was packed with interested parties for two days.
It is not said that you cannot earn from making data available" - Michiel van Dijk of ACM
Data migration as a revenue model
Data quality and data management are essential if you as an organization want to make the deployment of AI a success. "Garbage in is garbage out," says Lianne Verhoeven of Enable U at the booth. "So if you don't know if you can trust the data you're using, you may well get an answer that may not be of any use to you at all." According to Verhoeven, it is also important to pay close attention not only to data quality but also to how the data is accessed. "You actually want to set up your data management so that the moment you are no longer happy with your technology provider, you can pick things up and switch to another provider."
The Data Act that took effect Sept. 12 helps with that, says Michiel van Dijk of the Consumer & Market Authority. The new legislation regulates that a user of smart devices or associated services must have access to the data generated by them. "You are also allowed to decide who else may use the data," he says. So when switching to another manufacturer or provider, the user must also be able to take that data with him. "That could be a leased car or the machinery in your factory, but also a cloud service." There are producers who find that threatening, but it doesn't have to be at all, according to Van Dijk. "Nor is it saying that you can't earn from making that data available. You can then charge a reasonable amount for that. So there are also companies that choose to set up a cooperation around that data exchange. That way they turn the obligation into a sustainable and customer-oriented business model."
To questions like 'I want something with AI,' I always answer, 'What do you want with AI then?'" -Lianne Verhoeven of Enable U
For success first back to the source
Technology and data are very important, but the starting point of any data project should be the question: What do I actually want to achieve? Verhoeven of Enable U: "I get a lot of questions about AI like 'I want something with AI,' but then I always answer, 'Then what do you want with AI?'"
Sem Lemmers of Cmotions recognizes those questions and emphasizes that the business strategy should be leading when deploying data and AI. "Reason from your company's objectives. So what do we want to solve?" Only then comes the question of what technology you can deploy for that and what data are needed to do so.
This often requires linking different data sources together, and that is a specialty of Dataddo, explains Jurjah Slota at the company's booth. "In a no-code way, organizations can connect everything they use in terms of sources to a data platform, such as CRM applications, databases, cloud services, social media and sources on-prem. We now have 350 connectors and that number continues to grow."
If you don't give people the opportunity to properly test the value of an application, they form subjective opinions about it." -Vincent Hoogsteder of Mozaik
Ensure adoption
A new AI strategy must also have a place in the organization. If employees see the innovation as a threat, it's not going to work, no matter how nicely the technology can support the efficiency of a business process. Many AI projects still fail because of that. Vincent Hoogsteder of Mozaik emphasizes the importance of technicians and experts with domain knowledge working together from the beginning of a project. "What really can't happen nowadays is that engineers are working separately from people with domain knowledge. You have to make it one team to achieve a successful implementation."
Mozaik has compiled a list of 10 pitfalls in an implementation process. One is that the quality of an AI application is measured at the coffee machine, Hoogsteder says jokingly. "If you don't give people the opportunity to properly test the value of an application, they form subjective opinions about it. There's a lot of emotion involved in that. So the first thing we do with clients is to measure the quality of the AI through example cases. That approach gives you numbers to talk about."
Colorful spectacle
In addition to substantive insights, the booths of many exhibitors also offered a surprisingly varied experience with an attraction. For example, there were two fast cars on display on which exhibitors explained their vision of the right approach with data-driven work. There were games to play, prizes to win and treats to score. It was even possible to have a personalized perfume created using a large device with tubes full of colorful liquids. "It's a good way to physically showcase what our company does," said Alessandra Riviere of Optimizely. This company primarily helps large businesses use data to offer their customers a very personalized approach and thus achieve more conversions. "The machine here shows how filling out a form gives you something very personal in return. People find that very interesting, as you can see, and in that way personalization also works on a daily basis."
Thus, on the show floor this year, more than 100 exhibitors told their own stories around the strategic value of data and AI. While the focus varies by organization, they all shared the insight that data only has value if it is managed properly, put into context and actually used by people. They see bridging the gap between technology and application as the common challenge for the coming year.
