In recent months, it’s become increasingly common to stumble across questions related to the role of the data scientist among LinkedIn, online conferences and newspapers. “How much does a good data scientist cost? How quickly do companies return on their investment in such a figure? When is it time to bring a Chief of Data into a line of management?”. This comes as more and more companies are realizing that their approach to data has remained the same over the past few years, while the market around them has rapidly evolved where some companies are now truly able to use Artificial Intelligence and Big Data to perform analytics and improve business processes.
Investing in a data scientist is the most concrete idea that companies can have, but with less than 50k RAL it’s hard to bring a professional with a few years of experience to the team. Assuming you can find one. The best ones are in great demand and several companies in Italy, such as the largest insurance companies, some emblazoned consulting firms or the rampant fintech companies coming back from large investment rounds can make more attractive offers in terms of money, welfare and perceived prestige.
Before talking about the strategic importance of using a data scientist and understanding the scope of the investment, it is necessary to clarify the substantial differences between two often confused figures, the data analyst and the data scientist. Given the different points of contact and common activities, in the eyes of many, these two roles are almost overlapping. But there are fundamental differences between these two professions, particularly in how they approach data.
Data Scientist vs Data Analyst
To put it simply, while a Data Scientist is responsible for making predictions based on both proven and experimental patterns, a Data Analyst is responsible for extracting meaningful information from the data at the company’s disposal. The role of the latter does not concern forecasting activities, the Data Analyst solves business-related problems but does not offer a vision of future scenarios.
The two roles also have different working methods: the Data Scientist is engaged in managing fragmented information coming from different data sources, while the Data Analyst often deals with a single dataset at a time. The tools used also differ: the Data Scientist uses Machine Learning to work on large amounts of data, while the Data Analyst prefers products for data manipulation and report creation.
What also differs are their interlocutors, since the Data Scientist interacts with figures from different functions within the company, while the Data Analyst’s referents are more frequently Data Architects, Database operators, Database Administrators and other analysts. The different levels of complexity, responsibility and strategic value for the company between the two roles also have an impact on the salary of the two figures, according to a Job Research Data, Data Scientists would earn on average almost twice as much as Data Analysts.
Data Analyst and Companies
Returning to the Italian landscape of small and medium-sized companies with tens of millions in revenue and the growing awareness of having to rely more and more on Data Science to find the best solutions, what are the problems these companies face in selecting a Data Scientist?
- Internal HR is not accustomed to dealing with data scientists, so they often make mistakes in their assessments (the market is full of boasters claiming to be experts) to avoid relying on a headhunter who would weigh 20% of the RAL on their hiring.
- Hiring the best Data Scientists is becoming increasingly complex, especially due to the competitive nature of the offer.
- The Data Scientist often needs a business ecosystem that allows them to work at their best, so they may need a good Data Engineer on the team who can identify the data needed in the business IT ecosystem and can standardize it in the most useful way for analysis. To be able to unload the models wisely built by the new scientist on the ground, it is necessary to add a Data Analyst and perhaps a Strategist, able to translate the mathematical output into business actions and to interface with the managers of the other divisions. As it is easy to understand, having to acquire or train all these figures, the investment grows considerably and predicting the ROI becomes increasingly nebulous.
Data Scientist, why rely on an external team
Many of the companies that have turned to a specialized team like Premoneo’s have done so because they have understood that their pricing activities needed a structured process and that data science was the right solution to use, but that they could not or would not create the necessary team internally to ground that end-to-end process.
- Analysis of the IT ecosystem and data quality.
- Standardization of data useful for analysis and implementation of data exchange flows
- Reception of business needs, dialoguing with the various business functions involved
- Study and construction of mathematical models and AI algorithms (if applicable)
- Definition of the new process and development of the most suitable software solution, integrated with the IT ecosystem
- On-going analysis to verify results and refine the engine
All this at the “price” of a single Data Scientist, and with a guaranteed ROI within 3 quarters.
For companies that want to approach Data Science, the advice is therefore not to give up internal resources, but the most economically advantageous solution is to proceed with a more gradual approach, hiring a less experienced figure who can grow alongside the support of Premoneo or integrate a team, if already present, with the skills brought by our team.