Counterbalancing Query

I’m currently designing an experiment in which participants will be rating foods based on liking and willingness to pay.

Participants will see instructions for liking and then rate 48 foods before seeing instructions for willingngess to pay and then rating the 48 foods.
At the moment it looks like this:

The trials have been created so that the images are presented in a random order.
But what I’d now like to do is counterbalance between participants whether they see liking first or WTP.
I considered adding another loop around the Liking and WTP but I think this would mean that the instructions would also be randomised and potentially presented at the wrong point.

Any help greatly appreciated.