Converting means turning objects like Loft Object into editable polygon objects.
This can be done either by pressing ctrl+c, using the button on top of the left sidebar (object has to be selected in object manager) or right click on the object in object manager and select “convert object” from the context menu.
If you use the context menu, you can also use “current state to object” to create a polygon copy of your object and keep the original loft object as backup.
Sometimes this can be helpfull if you need to make further changes on the object later on.
But make sure to organize your objects well and delete all the backup objects before exporting.
Usually i have a work file and an export file, which is a copy of the work file minus all the backup and preview objects.
Also i like to export them “game ready” with all the smoothing, uv mapping (texture and lightmap UV), optimizing and triangulation done.
That way i don’t have to rely on any of the automatic features and have control over every aspect of the model.
As for optimisation just take a look at the resulting object of your loft object after you converted it.
There are 3 objects for your pillar, the pillar itself but also the caps.
Since in your scene there is geometry above and beyond every pillar, these caps will never be seen, but they still might get rendered ingame.
So in this case it just means you can delete these two cap objects.
However if they would be seen ingame, you would merge them with the pillar to a single object and edit the UVs and materials to cause as few drawcalls as possible.
Usually you select the meshes use right click “conect + delete” and then the optimize command (right click in the viewport while in polygon, edge or points mode and select optimize from the context menu, or top menu: Mesh>commands>optimize, or shortcut U~O,U~shift+O).
The optimize command in default configuration will delete all unused points (from deleted faces) and weld points that overlay each other (e.g. from connecting objects like caps to the body).