Hi there Marketo Nation,
I am reviewing some operation programs my predecessors put in place and this one has me confused.
The issue: Inflated lead scoring caused by bots/email scanners.
The fix: Copy/paste an invisible link into communications. If clicked, we know it was a bot or scanner, and not a human.
Questions:
1. Where would my predecessor have created this link? It goes to our homepage "http://bit.ly/2DIQR6Q"
Also here are the campaigns in question. The belong to a Bot Removal folder:
2. Can someone confirm my understanding of "remove Scoring" and "remove SI"?
If the below link is clicked, the system
a. will wait 5 minutes
b. remove that lead from our behavioral scoring program, and
c. request a new campaign.
Smartlist:
Flow:
This new campaign that is requested, Remove SI,
a. removes the lead from another campaign that attributes an Interesting Moment (so no interesting moment will be applied) and
b. requests the third campaign and final campaign in the bot folder, Remove High Value
Smartlist:
Flow:
This is where I am getting tripped up...
Why is this campaign requested only to remove them from the campaign in the flow? Doesn't make sense to me.
What is the logic?
Could it be that its an unfinished campaign?
Smartlist:
Flow:
Any insight would be helpful! Thanks!
Hey Elise,
Q1: when you ask where the link was created, do you mean how did they shorten it, or do you mean where within an email would it be placed? If the former, it's shortened via Bitly. If the latter, we can't answer that without seeing the code for an email - likely it's either manually added to emails, or it's included in the email template.
Q2: Your interpretation of what the campaign is likely trying to do is correct, by my reading. But what the campaign is likely trying to do is probably not going to be what it's actually doing, because I don't think the original designer of this flow fully understood how the remove from flow step works. I like to think of it like an ejector seat - it boots someone out of whichever part of the flow they're up to at that point. It doesn't remove someone from a program entirely, and it doesn't inherently guarantee that the flow actions of the target smart campaign will not be completed. If the intention of these campaigns is to:
this will only work if this campaign is able to "eject" the person from the flow of the target smart campaigns before the score change, interesting moment, or high value actions are committed. It's effectively what we refer to as a "race condition" - where the outcome of the process isn't guaranteed to be consistent because the outcome is dependent on actions being committed in a specific order, and the way that it is built out doesn't inherently guarantee the order (unless there's a wait step on the target smart campaigns that ensures people are held in step one for long enough that all of this processes).
If an analogy helps, it's like having two racers on a hurdle track. This logic effectively only works if Racer A is able to jump their first hurdle before racer B jumps any; but we don't know whether the racers are on the same starting gun (i.e. whether they're being triggered on the same condition) and whether Racer A's hurdle is taller and further away than Racer B's hurdles (i.e., whether their processes can be committed faster, esp. given the wait step included here).
We can't really guarantee this is working accurately without seeing the logic for the other campaigns, but overall it's an unnecessarily complicated and risky way of achieving something that (in my opinion) could be done more simply and accurately within the scoring campaigns themselves - as a filter on the campaign trigger, or simply placing the remove from flow step within the target campaign itself.
RE: your last point of confusion; again, we can't be sure without seeing the logic of the smart campaign it references, but for each other campaign, the only two steps are effectively: remove from flow X, trigger B where B then removes from flow Y, and triggers C, and C then removes from flow Z, and has no further campaigns to trigger - so only the removal from flow step is required. Again - the intention seems rational, but the logic is likely not accurately translating the intention.
Long but hopefully that's helpful
Thank you Grace! I like your analogy of an ejector seat. That makes sense in my mind. I also think you make a good point, that these campaigns were probably onerous and the end result could simply be achieved within the scoring campaign. I'll do one more review of all the campaigns, including the campaigns targeted by the "remove from flow" and see if the "race condition" is applicable.
Very helpful response!
The fix: Copy/paste an invisible link into communications. If clicked, we know it was a bot or scanner, and not a human.
While the click on that hidden link is probably not human, what do you do when these link scanner eventually send the email to the intended recipient's inbox? And when they finally do engage with the email? If you're discarding anyone that clicks that hidden link, you very well could be discarding users who eventually engage with your email.
Yep, I assumed (optimistically) that the Wait was meant to minimize impact to other campaigns that fired after the 5-minute interval. Though deciding that no human clicks within 5m is as irrational as deciding that every human will click within 5m.
As you know, whole thing smells of false positives, we really need to stop looking at clicks that could be automated as meaningful either way.
I agree...we knocked the scoring way low as a result. My aim here is to clean-up this marketo instance I inherited.
They should still score appropriately...but Grace above mentioned, I should chuck the bot programs and simply set up a filter.