Anyone else work in a fundraising office and willing to talk shop? Part of my job is identifying donors for assignment to fundraising staff (i.e. we think that donor will appreciate/benefit from a 1-1 relationship manager, which will in turn lead to greater engagement and larger or more frequent gifts for the hospital). We have a pretty sophisticated prioritization process with scores and models, and a large donor pool to pull from. I would love to combat bias that is baked into our models and wealth scores, and to elevate prospects we might bypass because they don’t score highly on our model (which identifies donors who ‘look’ like our current pool of likely majority white/older population). One idea was to use ACS data to identify majority non-white zip codes in our home state and flag/elevate donors who come to us from those zip codes even if they do not score in the ‘assignable’ range on our model. So for example, we will review all $1k+ donors for assignment if they score a 10 on our model. Can we widen the net and potentially engage a more diverse pool by reviewing all $1k+ donors regardless of score, who are from zip codes with more than 50% non-white population according to ACS? The capacity of our relationship managers is limited - we cannot work 1-1 with all donors at this level, so how do we make sure we aren’t overlooking people? I am not a data scientist and am looking for any and all advice as I think this through, current considerations: how to define a metric we could use to measure change/success (if that is even necessary?), will doing this just add to/reinforce bias somehow?, is this too simplistic or too subjective?, and more… I’d be grateful for anyone who is willing to connect!