How do you put a price on a data product? In this episode, Lucas and Luna dig into the messy economics of internal data marketplaces, from cost-plus transfer pricing to value-based models, and how a handful of companies are experimenting with dynamic pricing to balance demand and compute costs. They get concrete: the story of a logistics firm that cut its data warehouse bill by 18 percent by charging business units per query, and the debate over whether the 'consumer pays' model kills innovation. Along the way, they look at why data teams are suddenly acting like mini finance departments, and what that means for engineers. Plus, a short note on how listener support keeps the show independent. If you're building or running a data platform, this is the episode for you.

#DataPricing #DataProducts #DataMarketplace #InternalData #CostPerQuery #DataStrategy #DataEngineering #CloudCosts #DataFinance #BusinessTechnology #DataLeadership #DataGovernance #DataMonetization #Analytics #TechBusiness #FexingoBusiness #BusinessPodcast #DataDriven

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna explore how data teams are shifting their cost optimization strategy from cost per query to cost per active user. They discuss the limitations of granular cost metrics, the importance of aligning data costs with business value, and how a fintech company reduced its data spend per active user by 30 percent by refining its data architecture and query patterns. The hosts also cover the challenges of implementing this metric, including data governance, attribution, and the risk of over-optimization. Tune in for a practical look at how data leaders are changing the conversation around data costs.

#DataCosts #CostPerActiveUser #DataAnalytics #DataInfrastructure #DataTeams #Fintech #DataArchitecture #QueryOptimization #DataGovernance #BusinessValue #DataStrategy #CloudComputing #DataEngineering #DataOps #FexingoBusiness #BusinessPodcast #Technology #DataManagement

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The Data Business Podcast, Lucas and Luna explore how data teams are adopting semantic layers to cut query times and unify metrics across the organization. They dive into a real-world case: a mid-sized fintech that reduced dashboard load times by 40 percent after implementing a semantic layer on top of their lakehouse. The conversation covers the difference between semantic layers and traditional BI tools, the role of active metadata, and the pitfalls of over-modeling. Lucas shares practical advice on where to start, including using dbt and LookML, and discusses how semantic layers are becoming the new governance surface. Luna brings up the tension between self-service and control, and they wrap up with a forward-looking question about whether semantic layers will make data engineers obsolete.

#SemanticLayer #DataEngineering #DataAnalytics #DataGovernance #QueryPerformance #Lakehouse #dbt #LookML #MetricsLayer #DataProducts #DataInfrastructure #DataWorkspace #BusinessIntelligence #DataModeling #ActiveMetadata #FexingoBusiness #BusinessPodcast #DataPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The Data Business Podcast, Lucas and Luna explore the rise of cost-per-query as a key metric for data teams. They discuss how attributing cloud costs to specific queries helps engineering teams identify waste, optimize expensive pipelines, and align data spend with business value. Using a real-world example of a media company that slashed its data warehouse bill by 30 percent after implementing cost-per-query tracking, they break down the steps: tagging queries, building a cost allocation model, and creating dashboards that surface anomalies. They also tackle the tricky parts—like handling shared infrastructure and unexpected spikes from ad-hoc analysis. Tune in to learn why cost-per-query might be the missing piece in your data cost optimization strategy, and how to avoid common pitfalls like over-optimizing for cost at the expense of data quality.

#CostPerQuery #DataCostOptimization #CloudDataCosts #DataEngineering #FinOps #DataAnalytics #DataInfrastructure #CostAllocation #DataTeam #CloudSpend #DataWarehouse #QueryOptimization #DataMetrics #TechBusiness #DataManagement #FexingoBusiness #BusinessPodcast #DataPodcast

Keep every episode free: buymeacoffee.com/fexingo

On this episode of The Data Business Podcast, Lucas and Luna dig into a topic most data teams avoid until it bites them: the end-of-life of a data product. They use a real-world case—a mid-size fintech that had to retire a legacy customer-360 pipeline—to walk through the messy middle: when usage metrics drop, when downstream consumers ignore deprecation notices, and when the cost of running a data product quietly exceeds its value. They discuss concrete techniques like setting usage thresholds, running a sunset window, and using data product lifecycle policies that tie to cost and SLA. They also touch on the cultural shift from 'build and forget' to treating data products as living assets that need a retirement plan. If you've ever been surprised by a dashboard nobody looks at anymore, or a pipeline that's still costing you thousands each month, this episode is for you.

#DataProducts #DataLifecycle #DataRetirement #DataGovernance #DataOps #DataCostOptimization #FexingoBusiness #BusinessPodcast #DataEngineering #Analytics #DataStrategy #DataManagement #TechBusiness #Fintech #DataPipeline #DataQuality #DataCulture #DataLeadership

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The Data Business Podcast, Lucas and Luna dive into the emerging practice of 'shifting left' for data quality — embedding automated checks directly into the data pipeline development process. They explore how a mid-sized fintech reduced data incidents by 45 percent by implementing data quality gates in CI/CD, catching issues before they hit production. The conversation covers the key components of these gates, from schema validation to row-level checks, and the cultural shift required to make developers responsible for data quality. They also compare this approach to traditional monitoring and discuss the trade-offs, including the risk of slowing down development. With practical examples and candid insights, this episode offers a fresh perspective on proactive data quality management.

#DataQuality #ShiftLeftTesting #DataEngineering #DataPipelines #DataObservability #CI_CD #DataGovernance #DataProducts #BusinessAndTechnology #FexingoBusiness #BusinessPodcast #DataTeam #DataIncidents #SchemaValidation #RowLevelChecks #DataPipeline #DataCulture #DataOps

Keep every episode free: buymeacoffee.com/fexingo

Data contracts have become a cornerstone of trustworthy data products, but what happens when data flows across organizational boundaries? In this episode, Lucas and Luna explore the rise of data sharing agreements (DSAs)—the legal and technical frameworks that govern how companies exchange data with partners, vendors, and customers. They unpack a real-world case: a mid-sized logistics firm that cut data integration time by 25 percent by standardizing DSAs with its top ten partners. You'll learn how DSAs differ from data contracts, why they matter for everything from GDPR compliance to real-time streaming, and the three clauses every DSA should have. Plus, they discuss the emerging role of automated DSA negotiation and what it means for data teams. If you've ever grappled with 'whose data is it anyway'—this episode is for you.

#DataSharingAgreements #DataContracts #DataGovernance #DataPartnerships #GDPR #DataCompliance #DataEngineering #DataAnalytics #BusinessTechnology #DataStrategy #DataIntegration #LegalTech #DataStreaming #AutomatedNegotiation #LogisticsTech #DataProducts #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The Data Business Podcast, Lucas and Luna explore the quiet shift from giant foundation models to small language models purpose-built for internal data work. They use the example of a mid-sized logistics company that replaced a massive API-based model with a 7-billion-parameter local model for warehouse incident reports, cutting inference cost per request by 92 percent while keeping accuracy steady at 94 percent. They discuss why SLMs are easier to fine-tune on proprietary data, how they fit into data pipelines for classification and extraction, and why data teams should think about model size as an infrastructure decision rather than an AI fashion statement. Lucas brings a concrete cost breakdown: serving a 7-billion-parameter model on a single GPU costs about 15 cents per hour versus several dollars per 1,000 requests for a large commercial API. Luna challenges the premise, asking whether smaller models can really handle complex reasoning. They also touch on the role of open-source weights, quantization, and the practical limits of what an S-L-M can do. If you are running a data team and wondering whether bigger is always better, this episode gives you a clear framework for choosing the right tool.

#SmallLanguageModels #DataEngineering #DataInfrastructure #MachineLearning #ModelDeployment #InferenceCosts #DataPipelines #FineTuning #OpenSourceAI #LLMOps #DataStrategy #CloudCosts #AIAdoption #BusinessTechnology #DataTeam #FexingoBusiness #BusinessPodcast #DataPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The Data Business Podcast, Lucas and Luna explore how a regional airline cut its baggage-mishandling reports by 30 percent by applying data contracts to its real-time streaming pipelines. They discuss the shift from schema-on-read to schema-on-write, the role of schema registries in enforcing contracts at the edge, and why this approach reduces downstream breakage without slowing down producers. The conversation also covers the cultural change needed to get stream owners and consumers to agree on semantics before data flows, and how this practice is becoming a competitive advantage for companies that need trustworthy real-time data. Lucas and Luna share practical advice on starting small with a pilot stream, choosing the right contract fields, and balancing strictness with flexibility. Listeners will come away with a clear understanding of how data contracts for streams differ from those for batch data, and how to implement them in their own teams.

#DataContracts #RealTimeData #StreamingData #DataEngineering #DataGovernance #SchemaRegistry #DataQuality #DataPipelines #BusinessAndTechnology #DataBusiness #FexingoBusiness #BusinessPodcast #DataInfrastructure #Analytics #DataProducts #Kafka #Avro #DataOps

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The Data Business Podcast, Lucas and Luna explore the growing role of reverse ETL in turning analytics into action. They open with a practical example: how a mid-sized e-commerce company used reverse ETL to sync customer segmentation scores from its warehouse into its email platform, cutting campaign preparation time from two days to ten minutes. The conversation covers the core concept—syncing transformed data back into operational tools—and contrasts it with traditional ETL, the rise of the customer data platform, and why this pattern is reshaping how data teams think about their output. They also discuss the technical reality: where reverse ETL fits in the modern data stack, the maturity of tools like Hightouch and Census, and the operational challenges around data ownership and governance. Along the way, they touch on the 'if it was worth it' mindset—asking whether the effort to build these pipelines pays off. The episode is grounded in real-world trade-offs, with a clear takeaway: reverse ETL is not a silver bullet, but it's becoming a key bridge between data teams and the rest of the business.

#ReverseETL #DataEngineering #DataActivation #CustomerSegmentation #Hightouch #Census #DataWarehouse #CDP #MarketingAnalytics #RealTimeData #DataOps #DataGovernance #DataStack #BusinessIntelligence #DataScience #BusinessAndTechnology #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo