A manufacturing company’s general manager has been pulling data from the ERP system for months. Sales reports, inventory movements, customer records — all of it is there. But nobody in the organization can bring these pieces together and turn them into a coherent decision-making foundation. The accountant closes the books, the IT department keeps the system running, the sales team builds its own spreadsheets. The question left unanswered: who will read this data, interpret it, and connect it to strategy? This is precisely where the ‘data scientist’ concept is entering the corporate agenda.
The data scientist does not come from a single discipline. Without statistical knowledge, raw data cannot be made meaningful. Without programming skills, working through large data sets becomes impractical. Without business analysis skills, the outputs produced never touch the company’s real problems. The intersection of these three areas produces a profile that is genuinely rare in the market. Universities are not yet training this profile systematically — statistics departments underteach programming, engineering departments underteach business processes, and business schools underteach data processing. Companies end up either waiting a long time to fill this role or hiring the wrong person for it.
The ambiguity in defining the role is itself part of the problem. Some companies conflate the data scientist with a business intelligence (BI) specialist; others position the role as a senior Excel user. The real distinction lies here: a BI specialist reports on existing data, while a data scientist generates questions that have not yet been asked of the data. This difference shows up directly in job postings and in how the position is placed within the organization. A company that defines the role incorrectly cannot attract the right candidate to begin with — and if it does, it loses that person within six months because expectations never align.
Enterprise ERP systems sit at the center of this conversation. SAP, Oracle, or local solutions accumulate years of data; the finance module, production module, and procurement module each operate independently, and the patterns at the intersection of these modules are not visible through standard reporting tools. The data scientist is the person who can surface those cross-module relationships using statistical methods and translate the results into language that managers can act on. This competency is becoming the single most critical human capital barrier standing between a company and genuine ROI from its ERP investment.
The retention problem is more complex than the hiring problem. The data scientist profile struggles to find its place within corporate hierarchies. Reporting into IT cuts the role off from business processes; reporting into business development restricts access to technical infrastructure. An independent analytics function reporting to finance or strategy is theoretically the most effective model, but building that structure requires a level of organizational maturity that most mid-sized Turkish companies have not yet reached. Add to this the compensation expectations: because this profile is scarce, it commands strong negotiating leverage and typically sits outside traditional corporate salary bands.
The most concrete practical obstacle is data quality. Companies that hire a data scientist often discover that a significant portion of the data sitting in their ERP is incomplete, inconsistent, or incorrectly coded. Cleaning up years of accumulated problems takes far longer than any analysis work. The data scientist spends the first months on data cleansing and standardization rather than the strategic analysis the organization was expecting. This erodes the candidate’s motivation and leaves management’s expectations unmet. Hiring for this role without first establishing a data governance foundation is like building a roof on a structure with no walls.
For executives putting the data scientist role on their agenda, three questions are worth asking honestly. Has the company assessed the actual quality of the data in its ERP and other systems? Has the reporting line and decision-making integration for this role been clearly defined within the organization? Has the compensation and career path been structured to reflect market reality? Proceeding with a hire before answering these three questions tends to disappoint both the candidate and the company. When done correctly, however, this role becomes one of the most consequential investments a company can make — turning data it already owns into a genuine competitive advantage.
This article was originally written in Turkish by Gökhan MERCANOĞLU on January 16, 2012 and has been automatically translated into English and other languages using machine translation.