Small software company
One untested assumption was setting the price of their growth.
A small software company believed growth meant hiring more people to key in documents by hand. We tested that belief, and output per person roughly tripled.
Strategy & Transformation · Operations & Optimization · Engineering & Infrastructure
3×
documents processed per person, per day
The business moment
The company sold data to its customers on a subscription and had done well at it for years. Growing meant turning thousands more documents into usable, indexed data, and the documents were hard to read. Poor scans, handwriting, inconsistent formats, misspelled names and references that pointed to other documents. The industry took it as given that this work needed people, and a lot of them. So the math looked fixed. More growth meant more documents, more data-entry staff and thinner margins.
The company had looked at hiring and at older text-recognition software, and neither one changed the numbers.
The real problem
Before we talked about technology, we spent time understanding how the business actually ran, where the money went and where people spent their hours. The answer was the hand work. Each document took five to ten minutes of concentrated effort to read, interpret and index, and that effort was the real cost of every step of growth.
The belief holding up the whole cost structure was that only a person could read these documents. Nobody had tested that belief against current tools.
What changed
We proposed a simple experiment to settle the question. In two weeks we built an AI-assisted proof of concept and ran it on a sample of the company's real documents. It read each document, picked out the details that mattered and produced structured data ready for the company's platform. It worked from the first runs, on handwritten documents too, and needed little tuning.
The company chose to have its own developer build the approach into the platform, and we trained the developer to run with it. Within about three months it was part of daily work.
The outcome
After the change, each person processed roughly three times as many documents a day. What had looked like a staffing problem became a question of leverage, and the company went ahead with its growth plans, making better use of the team it already had.
The assumption had been reasonable. It rested on years of experience and on earlier attempts across the industry. Reasonable assumptions still deserve a test, because tools change and a constraint that was once real may no longer hold.
Client details anonymized.
Share this case study
Practices