I did something like this at my last offsite. I also split the task into two sections, i think we traditionally use impact and cost/difficulty. Then we stack

I did something like this at my last offsite. I also split the task into two sections, i think we traditionally use impact and cost/difficulty. Then we stack ranked ideas using a horizontal wall. Once for impact. Once for difficulty. Then after the fact I correlated the outcome into a matrix. It’s like the difference between having humans estimate volume (hard and error prone) versus estimating length (much more likely to be accurate). The exercise was less painful to implement and the results seemed more accurate.