They didn't, not with sufficient detail. Hence the need for the sonar on the submersibles, probably need to do some undersea mapping to pinpoint the wreck. I only know it is "near" Islas del Rosario. It's in water 950 meters deep, so the average concerned citizen can't exactly strap on a wetsuit and go poking around the ocean floor for it.
This site [1] from 2015 claimed it was in 300 meter waters, ~25km from Cartagena. Some of the infographics suggest it is NW of Isla del Tesoro. Maybe best to look at nautical charts of the area, and plot a course from west of Isla Rosario to northwest of Isla del Tesoro, along the part of the sea floor that is ~900m deep (course runs roughly NE). Map/scan that with the sonar on the submersibles, focus on major "outcroppings" that are possibly the San Jose (or the other two nearby shipwrecks). For a second pass, go straight to those outcroppings with a higher-resolution scan, plus visual cameras (use the released photos as reference for visual confirmation). Once pinpointed, send in the bigger cargo-carrier sub with robotic arms, excavator shovel, etc. and start scooping up the treasure.
Split it 60/40 with HSBC, and your team still gets to keep ~$4.7 BILLION. That's a better unicorn offramp than building yet another BS SaaS cloud app.
[1] https://blog.geogarage.com/2015/12/spanish-galleon-may-conta...
You could set up monitoring of Colombian Naval resources.
You could position your team/startup to be the best contract to work the government retrieval job. Using knowledge from the winning bid, you extract part of the treasure during sanctioned reconnaissance.
Using some legal nonsense, you stake an organization to claim partial rights to the treasure and hope discovery during the legal nonsense reveals the location
You marry a daughter of Poseidon and ask for the treasure as dowry.