Data Challenges: Deploying To Serve Performance

The implementation of a data-centric business model makes it possible to realize the value creation opportunities generated by data on the different links of the value chain. The 2010s saw many companies take on the great challenge of Big Data. Because the financial potential was more visible and the ever-increasing number of digital channels quickly offered a multitude of new data, the exploitation of customer data for marketing uses became essential.

It is difficult to dispute this logic: you need to know your customers to meet their needs better and thus create value. Predictive marketing, ultra-personalization of the customer experience and loyalty programs, and monetization of anonymized customer data are all levers that make it possible to transform a prospect into a customer, maximize the value of a customer and, more generally, generate revenue: growth and new sources of income.

If, for some, the results are slow to come despite the investments made, the question is no longer whether it is relevant to invest in customer data and its Marketing operations but rather to know which use cases are the most profitable and value creators and how to industrialize them (organization, skills, operating methods, industrialization process, etc.). However, the data strategy of companies must today go well beyond simple Marketing applications to infuse the entire value chain.

Big Data requires “think big”, particularly in terms of investment; its potential cannot be limited to customer data nor be the exclusive playground of a marketing or digital director to generate ROI. Data must make it possible to modernize, optimize and improve the performance of all the components of the company’s value chain. To fully exploit the potential of data, the challenge for the company is to transform its operating model to become data-centric. In this respect, Carrefour offers a beautiful illustration of a company that has been able to transform itself to become data-centric under the leadership of Alexandre Bompard.

Through its partnerships (Google in the lead), its investments (€2.8 billion by 2022) and the acquisition of data skills (creation of a data lab, etc.), Carrefour is beginning to disseminate a data culture and deploy use cases for numerous activities – for example optimization of supplies to e-commerce sites using predictive algorithms, monetization of customer data via the Carrefour Média entity (turnover €50M in 2019), personalization of promotions using historical data ‘purchases…

In the industry, Air Liquide is a pioneer in digital and data-centric transformation. Supported by sensitive top management and knowledge about digital and data levers, data is used at all levels of the group and is integrated into all its major strategic projects. The deployment of “Smart Innovation Centers” on a global scale now allows remote control of production units, predictive maintenance and adaptation of production to demand.

This change in scale of digitalization enables Air Liquide to proudly display a sharply increased operating margin for the 2019 financial year (+90 basis points vs. 2018). The implementation of this data-centric business model makes it possible to realize the value creation opportunities generated by data on the different links of the value chain. Here are some examples.

Also Read: 6 Key Challenges For Successful Data Science Projects

Manage Company Strategy Using Real-Time Data

Let’s start with a topic that resonates with current affairs and strategic and operational management using real-time data. In a period of the Covid-19 crisis, the duration of which remains indefinite, it is more essential than ever to avoid the uncertainty of events and the environment and to adopt an agile approach to adjust its short-term strategy (opening of stores, resource management, etc.) and medium term (new offers and products etc.). To do this, having a real-time vision of all the parameters of your environment (competitive, regulatory, political, etc.) is essential to make the right decisions in terms of investment, strategic plan and operational management…

Today, our ability to aggregate and gather data in real-time should allow companies to react and decide with many more elements and parameters in times of uncertainty. It is in this context that the most advanced companies in different sectors have relied on real-time data platforms to react to the COVID-19 crisis:

• In the automobile industry, a car manufacturer used real-time modeling of its environment and demand to relocate its production and its inventory policy (between different models and countries – according to anticipated COVID impacts)
• In the insurance sector, certain players have integrated real-time and external data to adjust their risk estimation and, ultimately, their prices and recovery processes in an unprecedented situation where algorithms based only on history are used. revealed themselves to be outdated
• Finally, Shell has relied for 50 years on a scenario methodology to estimate developments in its political and economic environment and its market. Recently, these scenarios have been monitored and refined using real-time data.
• Improve operational performance
• Optimize processes using process mining

Both a strength and a liability of a large company, internal processes are challenging to analyze and attempts at optimization often find themselves confronted with resistance to change and unforeseen obstacles; in particular, how to precisely quantify the time spent on a process among a dozen others?

Faced with this complexity tinged with subjectivity, data provides its solution: process mining. What are we talking about? Process mining consists of identifying and mapping all the steps of a process – the logs of transactional IT applications, in all of its variants, that is to say, with a level of accuracy and completeness that it would be impossible for a human to transcribe alone.

This utterly transparent vision of the process, therefore, makes it possible to carry out a diagnosis: identify inefficiencies and redundancies but also points for optimization and automation of the methods. Beyond diagnosis, process mining makes it possible to intervene directly in existing processes and tools by anticipating the “next steps” of the process and suggesting them to operational staff. The potential is, therefore, vast since it can be applied to all of the company’s processes.

For Deutsche Telekom, process mining made it possible, for example, to optimize procure-to-pay: not only did the optimization of purchasing processes and their real-time management make it possible to achieve €10 million in savings for the purchasing department in 2019 – notably by reducing duplicate payments and late payment penalties – but the automation of processes and the time saved for teams on their execution also created the opportunity to identify €12M in additional savings. Coupled with RPA, process mining makes it possible to reach a new stage in terms of automation for real-time corrections in the company’s key processes.

Reduce Operating Costs Thanks To Predictive Maintenance

The development of IoT (Internet of Things) technologies and the progressive equipment of production facilities and tools are paving the way for numerous optimizations in management and costs: robotization, automation of production, predictive maintenance, etc. By deploying its digital tool Predity Vision, a clever mix of data, AI and IoT, Engie clearly understood the savings potential offered by predictive maintenance. With more than 4 million maintenance interventions per year, it was essential to optimize the travel of its traveling technicians as much as possible.

Today, the management tool anticipates the maintenance needs of its approximately 15,000 sites and can even intervene remotely for certain operations. Beyond reducing the cost of maintenance operations for Engie, this tool also proved particularly relevant during the COVID-19 crisis, making it possible to carry out 800 remote checks on sensitive customers (hospitals, EHPADs, etc.). ). Kone is also an excellent example of transformation through data: for almost five years, predictive maintenance has, unsurprisingly, been at the heart of the brand’s quality of service strategy, with convincing results: 50% less physical intervention and 80% customer re-engagement.

But by transforming these 250,000 elevators into connected objects, the company now has the ambition to monetize services such as the personalization of messages on the screens on behalf of third parties (trustee or advertising), the detection of users to position the elevator at the excellent floor, or even, in these times of COVID, the call and opening of the elevator by smartphone. In some hospitals, elevators communicate with robot assistants for meal distribution in order to facilitate their movements.

Define Your Pricing Strategy And Manage Your Margins Using Data

By grouping, structuring and analyzing all the data that influence pricing, particularly competitive environment data, a retailer can obtain new visibility on its multi-channel price positioning and on ways to optimize its margins. It is with these means and objectives that we were able to support an international retailer and inform its pricing policy thanks to a more detailed knowledge of its price positioning on a national and local scale.

The synthesis of this data made it possible to define the benchmark price indices for the brand compared to its competitors. By specifying its price positioning and reducing the dispersion of its price indices, the company discovers a potential increase in annual turnover of €50 million [1% turnover]. This is only an overview of the different levers for data valorization; many other avenues deserve to be explored.

These are internal optimizations that are not visible but which allow companies to gain in competitiveness compared to their competitors significantly. These are also levers that require profound organizational changes to be fully exploited: the entire organization must undergo its digital transformation and become data-centric. Unsilocating data involves:

• Technological investments integrated into a robust IT ecosystem.
• The acquisition of new skills.
• Acculturation of all employees and associated governance.

Data is, therefore, a challenge for General Management (for those who still doubted it): it makes it possible to create value by generating additional revenue, building customer satisfaction and optimizing costs and processes. It becomes a key asset for the competitiveness of companies, in the same way as the customer portfolio or know-how. Unlocking its full potential requires a technical, cultural and organizational transformation of the company to bring data to all parts of the company – beyond the Marketing and Digital departments.

Also Read: The Biggest Challenges Around Big Data