Manage Scoring for Data Value Changes

[Anonymous]

Hi all - question on how some of you manage the demographic scoring front.

Obviously I have some initial demographic scoring set-up for when a lead is created based off of a few factors. For purposes of explanations, I'll just work off of company size.

In our initial scoring, company size might breakdown as follows:
1-50 :: +0
51-250 :: +10
251-500 :: +20
501-1000 :: +30
1001+ :: +50

This works great working off of a trigger of lead is created since it is the first time.

Where I'm concerned this is failing me is when data values update. For example, if someone comes back to our site 2 months later, and in the form they fill out they update their company size from 251-500 to 1001+ (they grew, or inputted it wrong the first time... whatever). I want the scoring to update accordingly, but am looking for a way to manage this without having a ton of separate triggers to update this.

The only way I see being able to do this is via triggers of data value updated with a constraint of a previous value. So for one trigger the previous value would have to be set to 51-250, and then the flow steps are a bunch of choices based off of the difference between the point values for the new choice. But then this would require a separate trigger for each initial value. Not to mention all the other aspects that the demographic scoring is based on.

How is this managed at your company, if at all?

Thanks!

[Anonymous · Accepted solution]

Whenever I create any trigger, not just for scoring, I usually have to create two triggers: one fires upon record creation, and the other upon data value change. Unfortunately, in Marketo there is no choice like there is in a SFDC workflow rule that says "upon creation or future edit."

But your case a little bit more complicated because not only do you need to create an additional trigger for data value changes - the way that you've structured your scoring, you necessarily have to imbed - somewhere - the logic of "if this, then that" of your formula.

One way you could try to deal with this is to create a score that's a weighted average of other "absolute" scores (likely using SFDC). So in this case you would have a "company size" score that's set in absolute terms - if bucket is X then score is Y instead of +Y. That will take away the requirement to deal with all of the "previous values" logic, but it instead moves the complexity into the score formula instead.

A totally different way to approach this is to say - so what? What's the harm if the formula is not exactly precise? So what if a lead gets an extra 20 or 30 points based on a data value change?

In most scoring scoring formulas these types of errors are common, and are usually ignored if not undetected.

Because your score is not meant to be a preciseI would do nothing about it until you can show that it's causing a real business problem (ex: reps are routinely served up unacceptable leads) - which I would be very surprised if it did - and then it will be worth spending the time/effort dealing with.

[Anonymous]

Thanks for the reply Grant.

Yeah, I was planning on letting it ride and just monitoring it. It's definitely an exception scoring case, as things like company size and job function won't be changing very often for leads. That said,  I'm still a perfectionist with these kinds of things!

Was just one of those random thoughts I had while deep in the weeds and figured I'd see if anyone else had a better potential solution.