Newsletter Archives

“I don’t trust the data” – these four words can stop a data project dead in its tracks. Data validation challenges typically fall into two categories: genuine data inconsistencies and misaligned expectations. Whether it’s discrepancies in your Warehouse Management System (WMS) or unexpected variations in your Order Management System (OMS), getting to the root cause can become a resource-intensive process that pulls in more stakeholders than necessary. That’s why it’s important to have a…

Do you expect your team to be 100% on every task they do, never making a mistake? Of course not. So when they’re using manual reports, calculations, and formulas to report on SLAs, billing, and invoicing, Don’t be surprised when they get it wrong. All the Best, Tucker

Here’s a quick way to query your data warehouse data using AI. Export your data model to PDF Create a project in Claude and add your data model as a document Add any Business Glossary or data dictionary to your project Ask Claude to write a query to get x data It’s still not fully automated, and there is the chance that Claude doesn’t understand the full business context, but it’s a quick and cheap way to experiment with AI. All the Best, Tucker

Nearly 20% of 3PLs still do manual billing. Unsurprisingly the ones that have found a way to automate their billing are more profitable. So if you could automate your billing why wouldn’t you? All the Best, Tucker

Billing is one of the hardest things to do in fulfillment. It sucks for 3PLs and brands. There are so many moving pieces within the warehouse and tech stack that exceptions happen constantly, and using Excel only compounds the number of mistakes, errors, and omissions that can occur for every invoice. This is the area where being data-driven can have the largest ROI in terms of increased billing quality and customer satisfaction. Here’s three you need to do if you want to have data-driven…

Data projects tend towards failure. Not necessarily because of anything anyone did but rather because of the things people don’t do. Here is a brief list of all of the things you should do if you want your data project to fail: Don’t have an executive sponsor. Do have daily check-ins Avoid implementing data quality checks Don’t bother understanding the business process Focus on the technology rather than the outcome Deliver everything at once Take everything people tell you at face value;…

Data is only valuable when it is in motion. Typically, it goes through a couple of stages of motion: ETL/ELT and Integration – This is where data is extracted from source systems, modeled, and combined into one source of truth. Reporting and Visualization – This is where you would use a BI tool like Power BI to visualize and present data to your end users as actionable insights. If you try to skip a stage or take a shortcut, either by integrating your data but not reporting on it or jumping…

Using Excel as a data source for Power BI is risky business. Excel’s volatility—from typos and inconsistent entries to frequent formatting changes—makes it a poor choice for producing reliable data insights. Even minor alterations, like a column header change or sheet rename, can cause errors that take analysts hours to trace and fix. Power BI works best with structured, stable data. When Excel disrupts that stability, the risk of poor data quality and frequent report errors skyrockets. This…

Most companies today understand that data is the backbone of sustainable growth. A common trap many fall into is overemphasizing quick wins or focusing solely on long-term projects. Either approach alone often leads to internal friction, mistrust, and frustration—especially between fast-moving sales and a stretched-thin operations team still relying on outdated Excel reports. The real answer? Build for both immediate value and future scalability: Data models over quick reports: Start by…