I think this issue is not so much animal advocacy related, but more general about the measurability bias that could creep in when you don't follow the EA principles correctly. I think you should always make a cost effectiveness analysis, but acknowledge all your uncertainties in it. The heuristics you propose should be translated into numbers as well, for which EA already offers methods:
Is it strategic? Give a numerical estimate of how big the impact is when you succeed.
Is it needed? Use counterfactuality. What is the impact of the outcome with and without your intervention?
Is it likely to succeed? Use probabilities to arrive (together with 1 and 2) at Expected Value calculations.
Is it efficient? Compare interventions that are expressed in common units.
I think the underlying issue is that if something is hard to measure, people disregard the impact. This doesn't make sense: even if something is hard to measure, disregarding it is equivalent to saying out loud that your best estimate is that it doesn't have any impact, which is clearly wrong. So put a number on it!
I think this issue is not so much animal advocacy related, but more general about the measurability bias that could creep in when you don't follow the EA principles correctly. I think you should always make a cost effectiveness analysis, but acknowledge all your uncertainties in it. The heuristics you propose should be translated into numbers as well, for which EA already offers methods:
I think the underlying issue is that if something is hard to measure, people disregard the impact. This doesn't make sense: even if something is hard to measure, disregarding it is equivalent to saying out loud that your best estimate is that it doesn't have any impact, which is clearly wrong. So put a number on it!