ERP ve Kurumsal Yazılım 5 dk okuma

Knowing Your Digital Customer Through CDP: When Physical Contact Fades, Data Steps Forward

Consider a mid-sized apparel retail chain in Izmir with 332 employees and eight physical stores. They launched an e-commerce channel in late 2019 and orders started coming in. But when the general manager looked at the buyer list, none of the faces were familiar. A customer who had been shopping in-store for five years registered on the website with a different email address — so the system treated her as a complete stranger. No loyalty points. No purchase history. No segment data. The company digitalized its channel but lost its customer in the process. Closing this gap is precisely why Customer Data Platforms — CDPs — have entered the strategic conversation for retailers across Turkey.But here is a misconception worth correcting before anything else: CDP is not a solution to a data collection problem. It is a solution to a data unification problem. The data is already there. Point-of-sale records exist. Web logs exist. Email campaign open rates exist. CRM notes exist. The problem is that none of these systems talk to each other, and each platform records the same customer under a different identifier. That is the actual function of a CDP: pulling together the customer traces scattered across disconnected systems into a single unified profile, then keeping that profile accessible simultaneously to marketing, sales and service channels. Treat vendor claims about it being ‘revolutionary’ with appropriate skepticism. A CDP is the technological expression of a decision to manage data properly. Without that decision, the platform simply does not work.Turkey’s retail sector began taking its e-commerce channel seriously in earnest through 2019. As more consumers moved online for purchasing decisions, retailers found that the face the cashier recognized in-store became an anonymous IP address on the digital channel. The practical value of CDP in this context comes from its identity resolution mechanism: different identifiers — email address, phone number, browser cookie, login credential — are matched using probabilistic or rule-based methods to build a single profile. In theory this sounds powerful. In practice, the quality of that matching is directly tied to the firm’s data collection discipline. If the same customer’s phone number appears with a leading zero in one system and with a country code in another, no algorithm closes that gap reliably.Back to the Izmir chain. When management began evaluating a CDP solution, they ran an internal data audit first — a step many Turkish SMEs skip because it feels slow. What they found was telling: the e-commerce platform’s customer database and the POS loyalty card database used different field formats for nearly every key identifier. Phone numbers were inconsistent. Email addresses had mixed capitalisation. Roughly 55 percent of records — a figure determined through the firm’s own internal analysis, not a vendor benchmark — could not be reliably matched between the two systems. This is precisely the condition under which a CDP delivers far less than promised. No platform compensates for structural data chaos. If data cleaning and standardisation work has not been done before the platform is implemented, the unified profile remains theoretical and the customer remains anonymous.So under what conditions does a CDP genuinely earn its place? Based on direct field experience, the following three indicators matter: customer data accumulates across multiple touchpoints recorded in separate systems; inconsistent messaging reaches the same customer through different channels and is beginning to damage brand perception; and calculating customer lifetime value requires manual data consolidation. If at least two of these conditions apply, a CDP evaluation is rational. For a single-channel SME with a simple customer data structure whose POS and CRM already share data, a CDP is an oversized tool. A building materials wholesaler in Ankara with around 420 employees went through exactly this evaluation last year and concluded that a straightforward API connection between their existing CRM and their new e-commerce platform met the majority of their needs — at a fraction of the cost and in far less time. Choosing the right-sized tool is worth considerably more than choosing the most expensive one.There is one practical constraint specific to Turkey’s SME environment that rarely appears in vendor materials: CDP licensing and implementation costs are denominated in foreign currency. Dollar and euro rate volatility turns the annual licence fee into an unpredictable budget line. A retailer that calculated its project budget during a relatively calm exchange rate period in late 2019 may find the same budget suddenly insufficient with early 2020 rate movements. Foreign currency buffer must be built into any CDP project plan. Domestic software alternatives offer a meaningful advantage on this dimension; however, they have not yet reached the identity resolution depth that international platforms provide. This gap is unlikely to close in the near term, and it shapes the decision calculus for cost-sensitive SMEs.The question I would put to any retail manager considering this investment is straightforward: is your digital channel doing what your cashier does? The cashier greets the customer by name, remembers what she bought last season, suggests the right size without being asked. The digital equivalent is a personalised product recommendation, a segment-specific campaign, an email that arrives at the right moment with the right offer. A CDP is the name of the system that makes this connection possible across channels. But before you buy the system, honestly assess whether you have the discipline to collect, label and standardise your data consistently. If a significant share of your CRM records cannot be reliably matched to each other today, start there. A CDP does not build that foundation. It assumes the foundation already exists.

This article was originally published in Turkish by Gökhan MERCANOĞLU on February 17, 2020. The English edition has been reviewed and edited by the author.

Gökhan MERCANOĞLU

Gökhan MERCANOĞLU

Teknoloji Danışmanı & Yazar

ERP, CRM, otomasyon, yapay zekâ ve kurumsal teknoloji stratejisi üzerine yazan bağımsız teknoloji danışmanı.

ERP ve Kurumsal Yazılım — Tüm Yazılar ERP ve Kurumsal Yazılım kategorisindeki yazıları gör →