Interest in the field of data analysis is high, and this is leading many professionals to opt for a total or partial career change. Companies, for their part, need to constantly structure their analytics projects to cope with the changes brought about by the information demands of Industry 4.0.
To help you through this process, we’ve written this article based on several occasions when we’ve worked with and trained data analysis teams.
The aim here is to present the conceptual differences between transactional systems (which generate transactional data) and analytical systems (which generate analytical data).
“I’ve seen in Aquarela‘s consultancies that one of the biggest steps companies are facing in climbing the data maturity ladder is getting everyone involved (scientists, analysts, directors) to know the difference between transactional systems and analytical systems. There’s a big communication problem going on in companies. Without this, teams not only work on different pages, but also (I venture to say) on different books.” (Joni Hoppen – Founding Partner of Aquarela)
Transactional data
When we make a bank transfer, the transaction of money from our account to the establishment generates Transactional or Operational Data. They are characterized by short transaction times and a small volume of manipulated data. Their focus is on “writing” information.
The transfer generates a transaction that has a value, an origin, a time limit, etc. These are systems that guarantee the integrity and temporal order of each transaction.
One of the main requirements of transactional systems is performance, i.e., the transaction must take place at the time it was required. We can also imagine them as real-time control systems (on-online systems) or near-real-time control systems.
Below are some examples of these systems:
- Banking systems: every transaction, payment or withdrawal you make generates a record of the action you have taken in a secure way, which is usually distributed across several systems.
- ERP systems: generally, companies that sell products and services keep their management systems running 24×7 in order to receive and integrate the transactions for buying, selling and stocking their products, right up to linking this information to the e-commerce site.
- In the area of digital marketing, we can highlight RD Station, a tool developed by Resultados Digitais that we use to manage our operation’s digital marketing. It is a transactional tool that also performs analytical functions, collecting information in real time on accesses, conversions, e-mails and leads that interact with our websites, forms and social networks.
Analytical Data
Analytical data is information generated from transactional systems. In other words, it is the set of transactions collected for the purposes of specific administrative decisions or even for defining long-term policies. Analytical data is the main input for planning, answering questions such as:
- What are the best-selling products at any given time?
- How do customers in region X behave in relation to customers in region Y?
- What are the factors that most influence the increase in sales during the winter?
The work on analytical data takes place offline and includes the analysis of transactional data grouped according to the type of question asked by the business analysts. Analytical data needs to be structured in analysis datasets (What are Datasets and how to use them). Analytical data is obtained in various ways, but mainly by extracting the databases into files in .CSV or .XLSX format.
Important recommendation: do not perform data analysis on transactional data, otherwise online services will be interrupted.
In this context of analysis, we include the constant use of Artificial Intelligence algorithms, statistics, mathematics and econometric models, depending on the business sector – (14 Sectors for applying Data Analytics). With these tools, we can structure analyses that make it possible to generate insights or new data that are important for increasing business efficiency.
The focus of data analytics is on reading and studying the patterns accumulated in transactional systems. A list of the types of analysis described in this article about descriptive, prescriptive and scenario analysis.
Examples of Analytical Systems
- Business Intelligence (BI) systems: these are systems that can be used to extract transactional data and generate simple, intuitive visualizations for management. There are several options on the market and their use allows companies to access analytics maturity level 3 (DCIM).
- Google Analytics (GA): aggregates a variety of information on the behavior of visitors to the company’s website, such as the number of hits, pages visited, visit time and various other indicators. This tool is very important for marketing and sales teams.
- Aquarela Vorteris: this is a data analysis tool that allows you to insert datasets from various sectors to detect outliers (read here what outliers are and how to treat them), measure the strength of factors in relation to certain results, generate preventive actions in logistics, billing, fraud and equipment maintenance.
Hybrid Systems
The greatest value extracted from analytical data occurs when there is a structured integration of transactional data with the knowledge generated by data analysis. This process can take place manually, by discovering patterns and adjusting the transactional system, or automatically, in the situation where it questions the analytical base before presenting an answer to users. Some examples of this are:
- Medical appointments: in the case of a hospital, the transactional system for scheduling medical appointments receives a request for a new appointment in real time and can query the analytical base in real time to infer the patient’s likelihood of missing this appointment. If, for example, there is a high probability of missing the appointment (informed by Artificial Intelligence tools), the administration could apply a rule so that he/she is obliged to make two confirmations or even choose the date with the lowest chance of missing before the appointment is scheduled – more information here.
- Netflix: This is a classic example of a disruptive business model that combines the transactional data from the delivery of high-quality videos in real time with a set of rules and computational heuristics generated by Artificial Intelligence that guarantee high-level movie recommendations according to the tool’s user profiles. Don’t think it’s strange that all the types of movies you like the most are the ones on the front page of your television.
In the table below, we’ve listed some practical examples and how the type of information (data) is classified:
Analytics projects – Conclusion
As we have seen, it is of great importance that the people involved in data analysis initiatives are aware of the conceptual differences involved, and of the information infrastructure set up for this. Failure to understand these concepts can lead to alignment difficulties, expectations and frustrations among the teams that operate transactional and/or analytical data.
In the end, we see that the big goal, or the path that everyone is seeking in analytics, is to achieve increasingly autonomous hybrid systems to serve their customers, such as the extraordinary case of Netflix, which undoubtedly falls within level 5 of data maturity in the DCIM methodology. To give companies and professionals an idea of the maturity levels of Brazilian companies in terms of analytics, we suggest checking out this survey we carried out:
Research on the quality of business data in Brazil.
The future of analytics is the automation of intelligent behavior, supported by Artificial Intelligence in all sectors where there is information, generating a level of optimization and personalization of services on a large scale, unparalleled in history.
What is Aquarela Advanced Analytics?
Aquarela Analytics is the winner of the CNI Innovation Award in Brazil and a national reference in the application of corporate Artificial Intelligence in the industry and large companies. Through the Vorteris platform and the DCM methodology, it serves important clients such as Embraer (aerospace), Scania, Mercedes-Benz, Randon Group (automotive), SolarBR Coca-Cola (food retail), Hospital das Clínicas (healthcare), NTS-Brasil (oil and gas), Auren,SPIC Brasil (energy), Telefônica Vivo (telecommunications), among others.
Stay tuned following Aquarela’s Linkedin!
Author
Founder – Commercial Director, Msc. Business Information Technology at University of Twente – The Netherlands. Lecturer in the area of Data Science, Data governance and business development for industry 4.0. Responsible for large projects in key industry players in Brazil in the areas of Energy, Telecom, Logistics and Food.
Experience building decision making pipelines for multiple industries using machine learning, econometrics and data analysis. Bachelor’s degree in Economics.
PhD and Master in Finance at Federal University of Santa Catarina – Brazil. Researcher in behavioral finance / economics, and capital markets. Currently Data Scientist applying machine learning strategies in business problems of large organizations in Brazil and abroad.
