Hey Amanda!
The short answer is that you can use a flow step for "Change Owner" with choices based on Random Sample.
How we trigger this is through a Lead Lifecycle Processing program, which basically runs all operational programs and operational smart campaigns within our Marketo instance.
Here's a snapshot of our Lifecycle Processing program tree:
This is the processing core of our Marketo instances—everything based on your lead lifecycle statuses. Basically, every trigger you currently have in your instance of "lead is created" can be better organized into this Lead Lifecycle Processing program. It also helps enormously in organizing your instance so that when you're onboarding someone new, you can just go straight to the lifecycle processing program to know exactly what happens at every stage of a lead's lifecycle.
This runs off of a new field attached to both leads and contacts called "Lifecycle Status." That way you don't run into the problem of lead/contact there.
Each folder contains the smart campaigns associated with the processing steps once a lead transitions to a new Lifecycle Status. The first one (300a, in this case) has the original triggers that being the daisy-chained, requestable smart campaigns.
Two types of triggers to begin the cascading: system proxies that are triggered automatically without your team's human intervention (such as "behavior score changed to be 100 or more"), and process proxies that are triggered based on certain field updates which happen through your team's human intervention (such as a human changing the Lifecycle Status to "3 - SAL" in this case).
These smart campaigns are daisy-chained together—each smart campaign has a trigger of "requestable" and has flow steps of: (1) requesting the primary processing smart campaign, and then (2) requesting the next smart campaign in the process.
Here's the specific smart campaign you're looking for, in this case "320 - SAL: XDR Round Robin":
Random Sample in the same flow step as choices takes whatever percentage you enter (above: 33%), and 33% of lead will land in that choice option. This is out of 100%, so in the above example it's 33% to Jacob Wild, 33% to Edward Unthank, and 34% to Dale Knudsen. (Currently, you can't go into decimals for random sample flow steps, so that's as Round-Robin as you can get.)
Lots of these smart campaigns in Lead Lifecycle Processing actually request smart campaigns in outside programs. If there's only one flow step that happens, I put the flow step in the Lead Lifecycle Processing program directly, such as the example above. When there's more than one flow step associated with a processing step, the triggering smart campaign in the Lifecycle Processing program begins the more complicated operational calculating program elsewhere.
A good example of this is having Lead Processing request the beginning smart campaign for a Traffic Director (which places leads into the most appropriate Engagement Program based on their persona profile, product interests, previous content downloads, and buyer cycle stage). That's a lot of decisions to make, so we start the processing in Lead Lifecycle Processing (for one consolidated view of what's happening), and if we ever want to deactivate the Traffic Director, we just deactivate that triggered smart campaign in Lead Lifecycle Processing.
I'm speaking about this at Summit 2015! Here's an abstract:
The best setups in Marketo are robust and scalable, allowing you to minimize the amount of effort required to create marketing programs. This presentation will walk you through how to architect Marketo for a best-practice setup, using a lead processing program as the primary driver, creating programs with hooks that latch in, and allow you the most flexibility in creating programs quickly and with the appropriate amount of energy. Come see how you can create a well-architected Marketo instance that’s easy to use, easy to report on, and easy to innovate. Never ask yourself “what does that do” again, and never worry about accidentally breaking programs again.
Hope that helps!
Edward Unthank | Founder, Etumos
Just going through old posts and had an update for this one.
"Random sample," when used on TRIGGER campaigns, actually does have exact-math round-robin distribution. The smart campaign retains memory of where the last random sample ended and then picks up from there. If you have a "lead is created" trigger and then have a flow step with choices based on the choice criteria of "random sample," it adds up and resets at 100. For example, these leads all had a "lead is created" trigger, and then went through FIVE SEPARATE triggered smart campaign to randomly assign a number 1 through 10.
"Random sample" is actually only truly random when used in batch smart campaigns.
Edward Unthank | Founder, Etumos
.png)