Advanced Lead Scoring

[Anonymous]

Conditional use case: We are in Real Estate so a lead looking to buy or sell a home for $300k in Shasta County CA (Avg home value ~$300k) fits our target persona, but a different lead in San Francisco County (Avg home value ~$1.3m) looking for/selling a $300k home would not be a fit. So we need the ability to combine two demographic scores with conditional logic to give true demographic scoring.

Dynamic use case: We use a timeline to move, buy, or sell to gauge readiness and urgency in our leads. So a lead that is 6 months + to make a purchase or list a home is getting a low score today but in 6 months they should have a higher score because they are now under the gun to act. How would we dynamically modify the score to show the change in readiness?s?

[Justin_Laberge1]

There's quite a few ways you can tackle this issue. Lets look at the first case here:

• You could use the individuals location field (would need to be county level) to get demographic information such as average house price per county in a field and then use lead scoring based on that. It could be weighted or you could simply assign a value to each income level. Thus if you have something like a total of 150 to AQL you could set the highest to be worth something like 25 to 50 to really set them apart if that's what you value heavily. This will require information from an external data source and some additional development to make work but would pay for itself in your case.

Similarly the question of readiness VS time will be dependent on your current situation but here's an idea:

• Create a lead scoring programs that set a wait step (maybe a better way) of a certain amount of time since the person has done some kind of activity that would indicate they interested to buy in the future. Create programs that would have a useful interval say 1 - 3 - 6 months, each would increase their score.
• Another cool thing you could do is add an interesting moment campaigns to the time campaign to allow sales people to see quickly milestones such as 3-6 months to get a quick view of how old they are.

Let me know if you have any questions, you should look into your overall lead scoring systems and determine whether or not separate scoring measures (beyond demographic and behavioral) would benefit you as well.

[Anonymous]

Thanks for the reply, and for the ideas. This is my first week working with Marketo so, please be patient with me. Are you saying I can have more score than just behavior demographic and lead( being the sum of the first two?)?

[SanfordWhiteman]

You can have as many score fields as you want. Obviously you don't want more than a few of them to be directly actionable (the others being contributing factors that are rolled up to a meta-score).

[Justin_Laberge1]

While you're in the thread Sanford, I wanted to know is the summing feature we use which looks like

##SUM({{lead.Score1}}, {{lead.Score2}},{{lead.Score3}})##

Enabled by default or not? It wasn't clear to me if this is normal marketo functionality.

[SanfordWhiteman]

Nope, those "Munctions" are no longer supported. New instances have to use a webhook for calculations/advanced scripting.

[Justin_Laberge1]

You are the king of Marketo! Thanks.

Do you have a similar example to accomplish the summing using a webhook? If not, no worries since the "Munction" is still doing it's thing.

[SanfordWhiteman]

Well, there's no one language for webhooks, but in our case (JavaScript-based) the webhook payload would be

var totalScore = +{{lead.Score1}} + +{{lead.Score2}} + +{{lead.Score3}}

or

var totalScore = FBMath.sum( +{{lead.Score1}}, +{{lead.Score2}}, +{{lead.Score3}} )

[Dan_Stevens_]

Hey Justin - like you guys, we're still using Munctions for some advanced data transformations in our instance.  Primarily to transform CRM GUIDs to "friendly" values.  Keeping our fingers crossed that this "unsupported" capability remains for customers that had it deployed before they pulled the plug.

Sanford Whiteman - when using a webhook to accomplish the same functionality, are you calling out to a third-party service (like FlowBoost) or making the call within the local instance (using "localhost" in the endpoint URL)?

[SanfordWhiteman]

That's FlowBoost-flavored JS.