Data essential for the common good
Published on 18/06/2020 in Inspire
With the growth in digital transformation, collecting and analyzing data helps companies adapt their processes, plan new strategies, enhance the customer experience and improve the world of tomorrow. Why are data so important for companies?
Frédéric Lhostte: “All businesses have huge amounts of data. The objective is to use data to help these businesses become smarter. Data is used to produce a result. Internally, data helps a business to be more efficient and solve problems. Externally, data can create new business models and improve customer experience or solve the problems of tomorrow such as the circular economy.
You simply have to define a strategic objective to determine in advance what the data will be used for, something that is often underrated in this type of project. Data allows us to be better at understanding the past, so that we don’t make the same mistakes again, and better at anticipating the future in the interests of businesses and general well-being.”
Your own big data rapport?Register here
Can businesses manage without data today?
Frédéric: “I don’t think so. All businesses are affected, regardless of which sector they are in. In an age of digital transformation, if a company does not invest in data collection, data quality and data analysis, it will not be ready for the future. Today, everyone agrees that digital transformation is necessary, and that involves (full or partial) digitization. So, the question is: ‘How can we acquire the skills to do this as efficiently as possible?’”
How can a business manage this data collection and analysis?
Frédéric: “First and foremost, it’s a decision that must come from the highest strategic level. Individuals must be dedicated to this process internally with the objective of making rapid and large-scale use of this data in order to unlock its value.”
- Control quality and availability of data
- Adjust the organization following automation, analysis and AI
- Assign priorities for different projects
“Each business must also find the right balance between investing in internal skills and using external companies that have the appropriate expertise. How do you retain control over the analysis of your data while benefiting from the power of the specialist company ecosystem?”
49% of companies say that the consequences of the digital transformation are clearly visible in their sector. In 2015, this was only 15%.
In practical terms, what are the various sources of data available to companies?
Frédéric: “Internally, a business already has a large amount of data as part of its activities (for example, loyalty cards and the EDI systems in the retail sector). Next, new data can be added, for example, through IoT solutions (such as sensors for measuring machine performance in the industrial sector).”
“Externally, data can also be bought or obtained as open data, which means that it is freely available, such as the National Institute of Statistics (Statbel) data or the weather data feeds. Each company has a huge amount of data and can collect even more if it wishes. The important thing to remember is that the relevance of the collected data depends on how it will be used and analyzed.”
Do you have any examples?
Frédéric: “We work with JCDecaux. The physical advertising displays have to be completely redesigned and data can help the company reinvent itself. Advertisers prefer internet-targeted digital campaigns and are moving away from campaigns on physical connected advertising screens located in shopping centers. How is it a game changer? JCDecaux had its own data but it lacked audience data per advertising screen.
This is what we have been able to achieve together through new data on the presence of people in the vicinity of screens using IoT sensors, general data from our mobile network (data external to JCDecaux) and open data on social classes per municipality from Statbel.
By combining these parameters in a statistical and anonymous way, we’re able to define the main audience characteristics of the people passing in front of the screens and at what time. This gives a model for targeting the broadcasting of advertisements more precisely. In this case, data helps to create a new business model for physical advertising displays.”
“In the retail sector, we work with the Maasmechelen outlet. The solution implemented allows us to establish (see below regarding privacy) how many people visit the outlet, where they come from and which social class they belong to. This helps the company to maximize the value of its retail space.”
“Lastly, we have worked with the SNCB to assess the number of people passing through stations in real time in order to eliminate bottlenecks and predict traffic levels based on a Machine Learning model. These are just a few examples of the numerous applications.”
In the public health sector specifically, we can use our data to simulate and model the spread of an epidemic such as the coronavirus.
Does data collection only have a commercial purpose?
Frédéric: “More recently, we have offered our skills as part of the Data against COVID-19 task force set up by the Ministry of Health and the Ministry of the Digital Agenda with the three mobile operators and a number of data science companies and epidemiologists. The objective was to offer guidance, in strict compliance with the privacy regulations, on effectively reducing population movement between various municipalities, drastically reducing journeys of more than 40km, and optimizing the distribution of medical equipment. This project is a perfect example of how Big Data can be used to improve our quality of life.”
“We could also mention our data solutions for improving mobility on the roads via our BeMobile subsidiary, and the occasions when our Real Time Crowd Management solutions are requested by the security police at large events such as the Tour de France in Brussels.”
Four dynamic data models
- Descriptive model
Thanks to computational power, analytical information helps us take decisions by focusing on data from the past and predictions for the future. Historical data allows us to describe the past. If we take the example of queues at the security checks in an airport, by accumulating past data, we can tell what the busiest times are.
- Diagnostic model
If we enrich this model with other data sources, we can look for explanations. In our example, we can combine peak-time data with take-off data.
- Predictive model
This predicts future peak times based on historical data using artificial intelligence (AI).
- Prescriptive model
This model integrates AI-related predictions into companies’ operational processes so that the right decisions can be taken automatically and proactively. It will then be possible to predict which times and days the security companies, for example, will need additional staff. So data can help businesses or public authorities take the right decisions based on concrete, solid information.
What about the confidentiality of all this data?
Frédéric: “We comply fully with GDPR. All our solutions are consistent with the privacy framework. Data security is also essential, so the data is anonymized, aggregated and secured.”
We realize the crucial importance of data for business today, but what is the future of data collection and analysis?
Frédéric: “We are truly on an exponential curve that needs to be channeled. There are increasing numbers of connected objects that directly transmit significant volumes of data which we didn't have before. Artificial intelligence is developing very fast and the growth prospects for data analysis are truly vast. All these factors will help businesses become smarter and more effective while constantly reinventing themselves so that they can better meet their customer expectations. In general, I am convinced that these new insights will also be useful to help meet the challenges of today’s society - which we call Data for Good.”
Frédéric Lhostte, Head of IoT and Analytics at Proximus, is a commercial engineer with a passion for innovation, new trends and the impact of new technologies on daily life. He joined Proximus in 1998 as ADSL project manager. He then worked in operations and in the strategy department. He has been a member of the IoT & Analytics department since 2010.
Proximus AnalyticsDownload the brochure
Stay ahead with the Proximus AcceleratorsFind an IT-partner
Read more about smart entrepreneurship in a digital worldGo to One magazine