A retail chain’s general manager describes receiving daily sales reports every morning — branch revenues, stock levels, return rates, all neatly prepared. Yet when asked what the company actually does with this data, the answer is less clear. Reports get read, discussed in meetings, then filed away. Actual decisions still rely heavily on experience and intuition. This picture is familiar to many mid-sized businesses operating in Turkey today.
‘Big data’ has become one of the most discussed topics in the technology world over recent months. The term describes data sets whose volume, velocity, and variety exceed the comfortable processing capacity of conventional database tools. But that technical definition carries little meaning for management teams. The real issue is this: businesses already generate substantial amounts of data — from ERP systems, accounting software, point-of-sale terminals, and email traffic. The problem is not a shortage of data; it is not knowing what to do with it.
Getting the conceptual framework right matters. Data volume alone carries no inherent value; value emerges when data meets the right question. This is why management teams need to answer three foundational questions before asking ‘what can big data do for us?’ Those questions are: Which data are we collecting? Who is analyzing it? Which decisions does it feed? A business that cannot answer these three questions will find that even the most sophisticated analytics infrastructure runs in circles. Conversely, a small or medium-sized enterprise that has clarified these questions can gain a meaningful competitive edge with the tools already at hand.
‘Which data are we collecting?’ is more complex than it appears. Most businesses collect whatever their systems happen to produce — the data that is easiest to reach. But strategically valuable data is not always the same thing. Systematic records of customer complaints, tracking of supplier delivery times, outcomes of sales conversations — these tend to end up in spreadsheets or paper notes, never entering a centralized structure. The decision about what to collect is, at its core, a prioritization decision: which business question do we want to answer, and where does the data to answer it live?
‘Who is analyzing it?’ is an organizational question. In most Turkish businesses, data analysis is delegated either to the IT department or to the accounting unit. Yet extracting meaning from data requires profiles that understand both business processes and analytical methods simultaneously. As business intelligence tools become more accessible, the technical barrier is falling — but analytical thinking remains a scarce skill. Businesses that recognize this gap are moving toward developing internal data analysts or engaging outside consultants. The key point is that analysis must not be left without an owner; results need to land on a manager’s desk and trigger action.
‘Which decisions does it feed?’ is the most critical question of all. When analysis is disconnected from decision-making, it becomes a decorative activity. For every analytical output, management teams should ask: which decision does this change? If the answer is ‘none,’ the resources spent on that analysis are wasted. On the other hand, building a data-supported framework for recurring decisions — pricing, supplier selection, store layout, campaign timing — is both feasible and profitable. The ROI calculation follows from here: the difference between the cost of analysis and the value of the improved decision it enables.
Among the practical challenges, data quality ranks first. When data is pulled from multiple systems, inconsistencies emerge — the same customer appearing under different names in different records, stock figures in the ERP failing to match a physical warehouse count. These inconsistencies make analysis results misleading. Data cleaning and standardization is the most time-consuming phase of any analytics project, and it is routinely underestimated. Skipping it looks attractive when speed is the priority, but analysis built on dirty data is more dangerous than an intuitive decision — because it creates a false sense of confidence.
If you are asking where to begin as a manager, the answer does not lie in a large infrastructure investment. Start by mapping your existing data sources: ERP, accounting software, sales system, customer records. Then identify your three most critical recurring decisions and ask whether the data to support those decisions is actually in your hands. If collecting the missing data requires a process change, test that change at a small scale first. The big data journey does not begin with a large budget. It begins with the right questions.
This article was originally written in Turkish by Gökhan MERCANOĞLU on January 10, 2011 and has been automatically translated into English and other languages using machine translation.