I am Korcan Hekimoglu, a senior digital analyst in London. I have spent nine years in marketing and customer analytics, and the settings have varied more than the work: co-founding an education technology startup in Ankara, running six-figure acquisition budgets in San Francisco, three years of agency analytics across a dozen client brands, and now Canon EMEA, where I own measurement strategy and web performance reporting across more than 45 markets.
Underneath the job titles, the work has always been the same question asked with increasing rigour: what is the data telling us, and what should we do about it. The first half of that question is the analytical craft I care most about. At Canon that means the full range, from KPI frameworks and purchase intent tracking to the heavier machinery: customer segmentation with k-means clustering to find the audiences the averages hide, Monte Carlo simulation and scenario modelling for forecasting, causal readouts of A/B tests, and an AI and Search impact index we built to quantify how AI-driven search is reshaping customer behaviour. Analysis like that is only as good as the foundation under it, so a real part of my work lives in the plumbing: a properly structured BigQuery warehouse, modelled with dbt so that definitions live in code rather than in someone's memory, and governed dashboards in Looker and Power BI where a KPI means the same thing in every market that opens it.
The second half of the question, what should we do about it, is where I have changed the most. For most of my career an insight ended with a person: someone reads the analysis, someone decides, someone acts. As AI Lead for our analytics operation, I now spend much of my time building systems that carry part of that weight. On Google Cloud, we configure agents and automated workflows that handle data exploration, generate first-draft narratives from performance data, and summarise insight across markets, with Vertex AI, Gemini, Claude and OpenAI models each doing what they do best, and with governance and validation frameworks underneath so that every automated claim can be traced back to the query that produced it. I think of it as extending the analytics team rather than replacing it: the machines draft and monitor, the analysts judge and decide.
If there is one conviction the nine years have left me with, it is that insight delivery is not the last step of analysis but the point of it. A finding that never lands with the people who can act on it is indistinguishable from a finding that never existed. So I put real effort into the unfashionable end of the pipeline: readouts structured around what changed, why it matters and what to do next, narratives written for the stakeholder rather than the analyst, and self-serve reporting that answers the routine questions so the conversations can be about the interesting ones.
I studied at Bilkent University and later at Imperial College Business School, where I completed my MSc and, in 2025, an executive programme in business analytics. The tools change every few months; I try to keep the statistics underneath them in better shape than the tools. On my own time I build Agent Library, a small project for registering and monitoring AI agents in one place, and I am a regular at cloud and AI practitioner sessions around London, comparing notes with the people building this stuff for real.
Outside of the data, I shoot photography and video, and head for the hills whenever London allows it.
I am always looking for the next interesting challenge. If you have a hard problem that involves data, AI, or ideally both, I would like to hear about it.