How we measured advertising promises with regex and why the first export was off by half

The authors share their experience in collecting data to analyze franchise sales pages. Since no ready-made datasets existed for this niche, the team developed a custom script using regular expressions to identify key features on web pages. However, the initial data collection proved inaccurate: manual verification revealed that the script was off by more than double on several metrics. The article details why regular expressions can find formal text matches while missing the actual meaning. The authors analyze the limitations of this approach and propose strategies for optimizing data parsing when moving away from regex entirely is not feasible. This article is useful for data specialists and developers who encounter challenges with content interpretation during automated web scraping and data analysis.