90% of the data collected by companies are customer-related. Marketing ops optimization raises four essential questions: who are my customers? What are their behaviors and opinions? What types of clients should I focus on? What will they be worth in the future? Marketing analytics bring factual answers through segmentation, profiling, targeting/scoring and customer lifetime value.case study
- Covid-19 Read more
- Actuarial Sciences Read more
- CFO Advisory Read more
- CIO Advisory Read more
- Change Management Read more
- Corporate strategy Read more
- Data Science Read more
- Digital transformation Read more
- Growth & Innovation Read more
- Human Resources Read more
- Marketing & Customer Experience Read more
- Operational Excellence Read more
- Procurement & Sourcing Read more
- Pricing & Revenue Management Read more
Data is rapidly turning into one of the most valuable assets for a company. Using the right tools, it has become possible to find new answers to critical business issues.
Data science is an open field for innovation due to:
- the cultural shift within private and public companies, who increasingly look for ways to capitalize on the value of data;
- the information sources: use of freely accessible sources (Open Data and web-scrapping) or proprietary sources, taking advantage of Big Data infrastructure capabilities
- the use of up-to-date hardware: implementation of scalable databases and advanced statistical models, such as Machine Learning, Forecasting, and Natural Language Processing
- the way to share and present data: reaching new targets via Open Data platforms and using state of the art Data Visualization principles and tools to provide meaningful insights for decision-makers
For Sia Partners, the success of a Data Science development program requires, above all, a pragmatic and targeted identification of concrete, operational and value-creating applications. A successful Data Science project is based on an iteration between technical and business expertise. This dual expertise is the DNA of Sia Partners’ Data Science team.
The team brings together the necessary diversity of profiles and has developed applications for many uses such as:
- Marketing Analytics
- Perception surveys and sentiment analysis
- Fraud and outliers detection
- Forecasting and Predictive Maintenance
- Business Performance and Process Improvement
- Pricing and Revenue Management
Our team has the ability to address complex technical and management subjects. Thanks to Sia Partners‘ international footprint, the Data Science team is wherever companies need it (New York, London, Singapore, Paris…) and our integrated model allows several teams from different countries to work together on the same project for our international clients.
A major gas supplier asked us to help target the core value customers in the market. Using public data and scoring methodologies, we reduced the number of clients to target from 8 million down to 180,000. We also delivered sales lead files for its commercial team.Our demo applications showcase the model enrichment via social data and the performance measure of various non-supervised detection models.
Perception surveys and sentiment analysis
Sentiment analysis provides a better understanding of customer relationship, and insights to increase its value.
Opinion mining, a quickly developing field, makes use of Natural Language Processing techniques to extract subjective meaning from customers’ feedback. Used in combination with web-scrapping tools, this allows teams to follow the evolution of the opinion online, in real time, on large volumes of user-generated content collected from social networks and review aggregators.case study
For a tourism agency, we analyzed the perception of touristic destinations. Using Natural Language Processing of their social media and comments posted by tourists visiting sights, Sia Partners identified the themes and the polarity of the messages, casting light on assets and liabilities.Our demo applications showcase the value hidden in easily reachable data sources. For example, we have developed web tools for transport competitive positioning and insurance reverse engineering.
Fraud and outlier detection
With massive and real-time digital communication, fraud and malfunction detection has become a priority, especially for public and financial entities. This field is based on both supervised and unsupervised models, usually enriched with social and socio-demographical data.case study
A major French player in the retail sector asked us to run a malfunction analysis of its delivery services, based on the available data – time schedules, destination address of the missing packages and geolocalization of their trucks. Our demo applications showcase the value hidden in open data. For example, we have developed web tools for the measure of road dangerosity and the pricing of medical acts.
Forecasting and Predictive Maintenance
By enabling the adjustment of inventory and replenishment to supply and demand, the prediction techniques contribute to improvements of the cost control in the value chain of the companies. Moreover, confronted to more and more time-sensitive phenomena, the increase in the volume of computable data made the prediction techniques more accurate and more relevant.case study
For the sales department of a major water operator, we collected and cleaned data, created a model to forecast the yearly volume of its residential clients’ consumption. We created a web tool for the sales team to run predictive analysis and adjust contracts and maintenance plans.
Business Performance and Process Improvement
In very competitive markets where innovation makes the difference, one must pay attention to your performance while reaching your customers’ expectations. The outbreak of DataViz tools enabled, not only to spread and communicate data, but also to analyze them on the spot, in collaboration with people that have no statistical expertise. With these tools, it is possible to cross sources and geolocalize many phenomena.case study
The French electricity DSO wished to evaluate the quality of its customer’s experience. Sia Partners delivered a cross-canal analysis of the interactions with end customers. This performance review focused both the quality of delivered services and the perception of the target. We built a Dataviz tool to run comparative and behavior analysis.
Pricing and Revenue Management
The development of e-commerce and the increasing competitive pressure in some business sectors allow pricing and optimization key issues to remain competitive. Pricing techniques aim to answer the following questions: What criteria is necessary to set a price? What is the price impact and the price signal? How can the perceived value be optimized? What pricing strategy should be adopted? Revenue management methods optimize the filling and the turnover by adjusting prices in real time.case study
An administration department decided to increase its revenue streams to make up for the decrease in public funding. We analyzed its sales data, segmented its client base, studied the price elasticity price and delivered a new revenue management model.