Thanks for posting an update about the outcomes and your reflections. It sounds like the right lesson that it would be good to consult more widely in the movement before trying similarly risky approaches.
I just wanted to ask a somewhat technical question about the estimate of the amount raised:
We estimated donations attributable to this campaign by looking at donations that (a) occurred after the first news coverage and before January 18th, (b) came from donor who had never donated through our platform before, (c) weren’t attributable to any other source [i.e. they didn’t use a matching code and they didn’t choose one of our other campaign when answering the “where did you first come across FarmKind?” question in the optional post-donation survey]. Our less conservative estimate counted all donations meeting these three requirements. Our conservative estimate only counted donations which either (a) were made through the offset calculator widget, (b) said they heard of us in a news article, or (c) donated a very specific number that very likely came from our calculator e.g. $209. We manually excluded 10 donations for which we have evidence that they weren’t caused by this campaign.
This doesn’t sound like it takes into account that (I guess) there would have been some amount donated through FarmKind without the campaign i.e. donations from new donors in that period not indicated as being linked to a campaign. What effect would subtracting this “background donation rate” have e.g. if it were based on a representative previous recent period of similar length (maybe from October-November or something if Christmas distorts things)? Or is this accounted for in a way I’ve not understood?
Thanks for posting an update about the outcomes and your reflections. It sounds like the right lesson that it would be good to consult more widely in the movement before trying similarly risky approaches.
I just wanted to ask a somewhat technical question about the estimate of the amount raised:
This doesn’t sound like it takes into account that (I guess) there would have been some amount donated through FarmKind without the campaign i.e. donations from new donors in that period not indicated as being linked to a campaign. What effect would subtracting this “background donation rate” have e.g. if it were based on a representative previous recent period of similar length (maybe from October-November or something if Christmas distorts things)? Or is this accounted for in a way I’ve not understood?