3 ms·

I appreciate you sharing your information as it came across as an honest attempt to be helpful, however I think you might find it helpful to read this post by A

I appreciate you sharing your information as it came across as an honest attempt to be helpful, however I think you might find it helpful to read this post by Avinash Kaushik [1] and then reread my post. I do digital advertising for a living, and that post is what I recommend to all who are new to the concept of cross-channel attribution as it is a very complex topic from both a data, math, and business standpoint. Easily one of the biggest opportunities and problems in the ad industry at the moment (and has been for some time). [1] http://www.kaushik.net/avinash/multi-channel-attribution-modeling-good-bad-ugly-models/ http://www.kaushik.net/avinash/multi-channel-attribution-mod...

mbesto 11y ago
No worries. I am actually in conversations with a client right now about multi-channel attribution, as I know it's a problem. The reason I responded the way I did was about your comment here: > If FB can provide advertisers with the tools to more definitively say "this is making you money" and not just "oh, we summed up all these random stats and call it Engagement--look at the shiny engagement metric!" I'm confident they would find advertisers even more willing to open their wallets, shift spend away from the GDN and exchanges, and stop questioning or caring about click fraud since it would be a self-correcting issue (ie. high levels of click fraud would in theory result in lower perceived incremental lift if proper analysis was conducted). Engagement metrics to me are much more about brand awareness then it is converting sales, so I didn't understand why you would point to engagement metrics when questioning tools that drive sales. Are you saying that you use conversion based ads on FB and aren't confident that their attribution window provides a highly accurate view of their performance?
- shostack 11y ago
  NP, I can see why that might have been read in a different way. I may have mixed my points up a bit, but basically I've noticed that they seem to lump quite a few things under "engagement" (an autoplayed video that plays for 3 seconds from someone scrolling in their feed is apparently a valid view). I'm using engagement in the broader sense, not the specific definition in the FB platform here, although it is pretty all encompassing there as well in terms of Likes/Comments/Shares (none of which are really the apples to apples at all). My bigger beef is on the view-through side. Attribution windows alone don't dictate attribution weighting. You really need to be measuring incremental lift, otherwise you end up with lots of people seeing an ad and a ton of view-throughs, but a lot of wasted spend since many would have converted without the impression.

mbesto 11y ago

No worries. I am actually in conversations with a client right now about multi-channel attribution, as I know it's a problem. The reason I responded the way I did was about your comment here: > If FB can provide advertisers with the tools to more definitively say "this is making you money" and not just "oh, we summed up all these random stats and call it Engagement--look at the shiny engagement metric!" I'm confident they would find advertisers even more willing to open their wallets, shift spend away from the GDN and exchanges, and stop questioning or caring about click fraud since it would be a self-correcting issue (ie. high levels of click fraud would in theory result in lower perceived incremental lift if proper analysis was conducted). Engagement metrics to me are much more about brand awareness then it is converting sales, so I didn't understand why you would point to engagement metrics when questioning tools that drive sales. Are you saying that you use conversion based ads on FB and aren't confident that their attribution window provides a highly accurate view of their performance?

shostack 11y ago
NP, I can see why that might have been read in a different way. I may have mixed my points up a bit, but basically I've noticed that they seem to lump quite a few things under "engagement" (an autoplayed video that plays for 3 seconds from someone scrolling in their feed is apparently a valid view). I'm using engagement in the broader sense, not the specific definition in the FB platform here, although it is pretty all encompassing there as well in terms of Likes/Comments/Shares (none of which are really the apples to apples at all). My bigger beef is on the view-through side. Attribution windows alone don't dictate attribution weighting. You really need to be measuring incremental lift, otherwise you end up with lots of people seeing an ad and a ton of view-throughs, but a lot of wasted spend since many would have converted without the impression.