Challenging an Industry Assumption to Change the Economics of Growth
A SaaS company believed county expansion required more payroll. A two-week experiment changed the economics of growth entirely.
The Situation
A subscription software platform providing digital title-plant data to title companies across Texas had operated successfully for more than a decade with approximately ten employees. Expanding into new counties meant reviewing, interpreting, and indexing thousands of historical documents, many dating back to the nineteenth century, with poor scans, handwritten notes, and inconsistent formats. The accepted belief was straightforward: more counties require more documents, more documents require more data-entry staff, and more staff reduces the profitability of expansion.
What We Found
Before discussing any technology, we spent time understanding how the business actually operated: where information entered the system, which expenses grew with each new county, and where employees spent most of their time. Document indexing, at five to ten minutes of concentrated effort per document, was the largest component of the largest expense. The bottleneck was never acquiring county data. It was the manual effort required to make that data usable, and the client remained skeptical that AI could reliably process such difficult historical records.
What Changed
Rather than debate whether AI would work, we proposed a simple two-week experiment using Azure Document Intelligence and Azure AI services against a subset of real documents and existing map data, to see whether the underlying assumption still held. The proof of concept succeeded immediately, including on handwritten documents that had historically been difficult to process, and required surprisingly little tuning. The client had an internal developer integrate the approach into the existing platform, and the solution was operational within approximately three months.
What Followed
Employee productivity increased from roughly 35 to 50 documents per person per day to approximately 125 to 175 documents per day, with improved accuracy. The proof of concept cost approximately five thousand dollars to build. By comparison, adding even two additional employees for county expansion would have represented more than one hundred thousand dollars a year in payroll. The client gained confidence that county expansion could occur without proportional increases in staffing. A reasonable assumption, built on years of experience, turned out not to be true once it was actually tested.
Share this case study
Facing a similar situation?
Start with a conversation. We help organizations understand what is actually happening before committing to a path.
Start a Conversation