THANK YOU FOR SUBSCRIBING
Good Data Quality - The Road to Enormous Profit
By Davide Cervellin, Head of Analytics, Insights & Data, Global Experiences, Booking.Com
Key Pain Points in Implementing Data Analytics
There is a distinction between technology companies that have data centres since their inception, and conventional ones existing for decades and having reached success without too much data. Technology companies often face difficulties in what I call “analysis paralysis,” a situation where they are so used to data for every business decision they need to make, that when there is no data, they struggle finding other ways to make a solid decision. On the other hand, traditional businesses that have reached success based on the instincts of higher-level executives find it difficult to employ data scientists and trust their insights when they radically differ from their views. In addition, managing data quality remains another significant challenge for modern companies. Until recently, it was part of every analyst’s or data scientist’s job to begin every analysis by cleaning data. Now, this process is changing. A very recent trend that I have observed is the emergence of a new position—the data engineer, responsible for the processes that move data from the production environment to a place that is consumable by other data practitioners, ensuring quality, speed, accuracy, and redundancy.
Data engineers help analysts and data scientists effectively do their work by improving data quality. Data practitioners can now trust the data assets maintained by data engineers without having to double-check it, eliminating the need for clean-ups and massively increase their speed, accuracy, and engagement. The role of data engineers expands toward taking care of taxonomy in a way that is uniquely tracked and appropriately logged. In case of any data issues, data engineers get notified to act quickly and look for fixes, informing all downstream users of possible delays. Data engineers also take care of data lineage, which is a massive task for companies like Booking.com with enormous data. The need for data engineers has grown significantly in recent months, and there seems to be higher demand that supply in the market for these profiles.
Your Approach to Data Analytics
The most significant skill for a data practitioner is the capacity of asking the right question because, too often, they are considered just as people who provide answers.
The most significant skill for a data analyst is the capacity of asking the right question at the right time
Business owners come to them with questions like “I need the following data for this” or “please give me that number.” However, to me, this is not the way the data practitioners should work, or at least, companies don’t get the best talent if this is the kind of questions they would have to deal with. Instead, companies can benefit by considering data analysts as business partners and approaching them to brainstorm about any problem and take their suggestions for the kind of data required to make any decision.
When I interview people, I focus on two main characteristics—assertiveness and business awareness. New graduates from universities often assume that being a good analyst means increasing knowing math or several programming languages. However, these graduates with high technical knowledge should focus more on soft skills because technical skills can be learned more easily than behavioural ones. People with excellent soft skills, ability to manage stakeholders, and having a peer-like conversation with a business owner can be unique and become “one in a billion.” Business-owners should relate to analysts and data scientists in business terms and, for this to work, it is important that the latter has a full understanding of the former’s point of view. If a data practitioner wants to maximise the impact they have in a company and make sure their ideas are heard and implemented, assertiveness and business understanding are foundational skills to master.
To have a team of good analysts, businesses must have processes that ensure the analysts can move from execution into the thinking mode, in accordance with the situation. For example, if a stakeholder has more requests that what can be provided in a given week, for instance every week there are ten questions and the analysts can only look into five of those, it is necessary to have a process where different aspects can be prioritized based on objective criteria (for instance it could impact, feasibility, actionability). This is where most companies fail, not being able to prioritize needs and ending up hiring more analysts. When business is good this may work, but as soon as it slows down, these analysts find themselves out of a job as they are usually seen as costs to the company.
To support the business, data analysts must ensure that the work they do is entirely unbiased and objective and, for this to happen, it is crucial that reporting lines make sure that they can speak their minds freely without any fear of repercussion on their career if they disagree with their managers.
Organizational setup, prioritization, and hiring with the characteristics that I spoke about—communication, commercial awareness, business sense, and technical skills -- are absolute requisites in the data analytics arena.
Advice to the Young Data Analytics Professionals
It depends on their background. If they come from a technical background, I would strongly suggest them to spend time on further improving their soft skills to handle negotiations, understand business matters, and try to speak with business people in a business sense. For example, a young professional who is on the verge of getting promoted to a more senior position in a company may end up struggling if his only strength is technical knowledge. The higher professionals rise, the lesser they deal with technical aspects. They have to focus more on the softer business aspects and conversations like the budget, resource allocation, hiring, talent management, as well as how to design effective organizational setups. In the end, their success goes through a lot of people, so they have to understand them and respect them if they want to have a positive impact on the business.
In my book “Office Of Cards” (available on Amazon), I wrote in detail all the advice I have for young professionals in the data analytics industry. It is based on my own experience over the last 15 years of my career, and it focuses on how to effectively deal with other people in large corporations. I wrote it with the hope that it can save the younger generation from the mistakes that I have made that resulted in frustrations at the beginning of my career.
Is Artificial Intelligence For Everybody?
Byron Flores, Global Head of Digital Finance, Data and Tools -Novartis Technical Operations at Novartis
A Picture of Health: Why we need a unified approach to Covid-safe air travel
Brandon Balcom, Snr. Director, Innovation Business Development, CWT
The evolution and future of Data Analytics
Clemens Baader, Chief Data Officer, International Personal Finance plc
Our safety mindset has broken through AI trust barriers
Caroline Gorski, Group Director-R Data Labs, Rolls-Royce