A manual process is robust, fault tolerant, and easy to adjust to new circumstances or changing business requirements.
A Python script is fixed, and they day it breaks down due to some slight change in scraped data, you have to pull in a developer to fix it. For a company without dedicated staff to support such automation, it is a high risk.
Ad-hoc automation can turn workflows into Rube Goldberg machines. A Python script here, an Excel macro there - soon nobody knows whats going on and nobody dares change anything.
Developers often think only one step ahead: It was manual, now it is automated with a script. Win! A business have to consider that larger picture - future support, maintainability, risk etc.
It may be lucrative for the friend to take the script and turn it into a product. but then they'll need to market, sell, document, optimize, and do a million other tasks.
There has always been a dead zone between "automated script that makes me work a bit more efficiently" and "product that can make everyone with this job work a bit more efficiently". Fortunately, new technology is narrowing this gap.
I worked with him to develop a Python script do basically automate half his job in a few hours that could have saved the company tens of thousands of dollars. The company didn't want to dedicate/hire an actual programmer to maintain it, so they kept paying employees to manually scrape the web.
The crazy part is the next summer he interned at a different company and literally the exact same thing happened. He re-made the original script, saved tons of work, and the new company still didn't end up using it after he left.