In this 1996 game by Jean Vanaise, published by Piatnik & Söhne and reviewed at The Game Cabinet, the best strategy may not be immediately obvious.
The way to proceed in this nicely-constructed example is clear, right? Load the Silk on Blue, the Spice on Gold and the Tea on Red and with the high point cards you have in those colors, do your best to get all three of those ships in. Right?
But consider what happens if you do this. When you play those "3" cards, everyone knows which ships you prefer and are unlikely to help them along. In fact, they may become automatic candidates for "-3" treatment. Also, once played, those "3" cards become unavailable until a deck reshuffle occurs and they may not return to you. For these reasons and others, I would suggest a different strategy.
Suppose that instead of allocating as above, we load our 3 Silk into the Red ship, the 2 Spice into the Green ship and 1 Tea into any of the remaining ships. (It doesn't matter greatly as it is only 1 good after all.) Now, the cards to play are our Red and Green ones. This makes players less certain about your interests and less likely to subject your ship to the "-3" backtrack move.
I suggest that one should never play a "3" card except in very special circumstances, namely, that you can use it to bring one of your preferred ships all the way in to Europe. As the player who brings the ship in improves the price of the commodity that he is carrying, this is nearly always a good idea.