They tested the counterintuitive claims by performance marketers that 'brand' keywords (containing Ebay) had the highest ROI. They did an experiment and found that, contrary to attribution model, these Ebay keywords resulted in ZERO incremental sales, although they were clicked on by many people that purchased.
This resulted in cut of Ebay marketing budged by $100 mn.
There is now a big literature in economics looking at experiments and natural experiments, generally finding much smaller sales impact of advertising than claimed by industry participants, and genearlly -ive ROI.
Obviously 'brand keywords' has the highest apparent ROI. It is users already looking for your site that click those. It is probably some internal bonus scam that fuel such nonsense - a industry wide marketing department conspiracy.
https://www.nber.org/system/files/working_papers/w20171/w201...
They tested the counterintuitive claims by performance marketers that 'brand' keywords (containing Ebay) had the highest ROI. They did an experiment and found that, contrary to attribution model, these Ebay keywords resulted in ZERO incremental sales, although they were clicked on by many people that purchased.
This resulted in cut of Ebay marketing budged by $100 mn.
There is now a big literature in economics looking at experiments and natural experiments, generally finding much smaller sales impact of advertising than claimed by industry participants, and genearlly -ive ROI.
https://www.nber.org/system/files/working_papers/w20171/w201...
This Freakonomics episode is a nice overview.
https://freakonomics.com/podcast/advertising-part-2/