How can we prevent Marketo from adding scores from multiple records that are being merged?
We are deduping our database and merging records together. This unfortunately is creating inflated MQL's because the scores are being added together.
How can we undo the merging of the score leads and how can we prevent this moving forward?
If you merge leads in Marketo, you have the possibility, on merge, to set manually a value for fields, rather that letting Marketo setting it. Just use the "Custom" column in the merge assistant.
Second thought : the lead score standard field may not be available in that screen...
In this case, you would have to copy the score in another score field before merging, do the merge as written above, and then copy back the score to the lead score field after merge.
Tip : to filter leads that were merged in a smart list there is no "lead was merged" filter. YOu can use the "Lead was created" filter with a "Min Number of Time = 2" constraint
Thanks for the insight, Greg.
We used DemandTools via Salesforce to do all of our merging. It sounds like this will make it virtually impossible to control how Marketo merges Score fields. We will have to think of some other way to prevent Marketo from merging these scores.
Do you have any suggestions on how to undo historical score merges?
Hi again Hoi.
Marketo score fields cannot be set from SFDC. But you can have them set from inside Marketo with a smart campaign. The best way is :
Hi! We're dealing with the same problem -- tons of "fake" MQLs getting sent to Sales because of this Marketo behavior.
I'm in the process of merging duplicate Leads and Contacts (so a Lead-to-Contact merge) with Ringlead (which doesn't let you export any of the duplicates that it identifies in its data cleanse reports). The Contact is naturally always the master, and we are fine with having the Contact's lead score be the score that carries over to the final, merged record.
The workaround I devised was to generate this smart list in Marketo to isolate the duplicate Leads. In Ringlead, we're only merging records that have an exact match on email address -- so I assumed that this smart list in Marketo would be 90-95% the same as what we're seeing there:
From here, I ran a single flow action to change lead score for all Leads to "0". Thus, when the Leads and Contacts get merged and their scores get added, there's no change.
So far it's working exactly as it should! Now, I'm not sure how to effect this workaround for Lead-Lead merges -- the fact that we're merging different object types allowed me to isolate 1/2 of each duplicate pair.
This smart list will only give the lead-lead duplicates, not the lead-contacts duplicates. Is this what you are looking for ?
I can only assume this list pulled all leads with a duplicate match on email address (regardless of record type) because all of the leads I then went onto merge with duplicate contacts in RingLead had a score of 0, which was the desired outcome.
The root cause of this issue is that Marketo's merge feature sums the Leads score so the demographic score gets added multiple times.
Presumably you'd want the behavioral score to be summed.
Please write your local Marketo product manager and ask them to fix this.
Or Mike Allen, the VP.
They tell me "One day" How about next Thursday?
This is a fundamental flaw in the application.
Zoe, of course you can export the results in RingLead, we just hid how to do it really well.
See that tiny little floppy drive (what is that anyway?) beneath the View Lead-to-Lead report? That's it.
Save report to csv.
The very best practice for optimizing your data is to focus on getting the data created properly at the point of creation.
A ounce of prevention is worth a pound of cure, right?
Whatever it costs you to get the data right when it is created, it will cost you 10x (if you are lucky) to clean it it up later.
Don't do that. Clean up your legacy data and put the best preventative measures in place that you can
Don't forget to normalize your data before deduping it, you will see as much as a +600% performance improvement from batch dedupe applications, as well as considerable improvement in prevention.