Best Practices for List Cleansing/Scrubbing

[Jamie-Carey]

Hey Everyone!

I am looking for your best practices for those weekly smaller lists (under 300 or so) of leads from your events teams. Do you rely on Marketo to clean up State, Country etc via trigger or batch campaigns? Or do you fix all those values beforehand in your file?

Thanks!

[Grégoire_Miche2]

Hi Jamie,

You want to set up and activate some data cleansing campaigns. Triggers are "lead is created" and "data value changes" on the specific field (e.g. country), and flow step will be "change data value" with choices (if country is US, new value is USA).

-Greg

[Jamie-Carey]

I do know *how* to do it, I am wondering what people end up doing. I am curious because running too many trigger campaigns can bog down your instance and create a backlog in the trigger queue preventing more important things from happening.

BUT if you wait to do something like that via batch, then you are sending dirty data to SFDC if you sync all leads right away (especially after important events).

I'm curious as to how people find balance between those two problems.

[Grégoire_Miche2]

Hi Jamie,

Sorry, I had not got the exact sense of you first question

Agreed, performance can be a problem. Yet, we do run these data management smart campaign as triggers, but we make sure that we add filters to the triggers, to limit the leads entering the flows. It will not prevent the trigger to be evaluated, though, but will still limit the leads entering the flows and then reduce the backlogs. For instance, for country cleansing, we will create a smart list "correct countries" and add a "member of smart list IS NOT correct countries" filter to the data value changes trigger.

My understanding is that one of the Project Orion (announced at summit last week) objectives is to very significantly increase Marketo capabilities to handle these efficiently. Marketo is announcing huge perf improvements specifically on these trigger evaluation / processing issues.

-Greg

[Jamie-Carey]

Oh that is helpful! So you have a choice for the change data value for country:

member of smart list> is > smart list: country> is not> list of correct values

(I hope that makes sense)

Yeah, I am really looking for increased processing on trigger and batch campaigns. I've seen some disasters.

[Grégoire_Miche2]

Hi Jamie,

not exactly. In the smart list of the smart campaign, we combine triggers and filters, as is:

-Greg

[Jamie-Carey]

Ah I see, different than what I was picturing but helpful! Thanks!

[Andy_Varshneya1]

I approached the problem similarly but with a slight twist.

I exported every lead in Marketo and used a pivot table to organize them by country and state. I then set the filters in the trigger campaigns to be only the countries that need updating (e.g. "US of A", "America", etc) and had the flow steps to update the country field.

I've also set up a smart list and subscription to send me an email every day of leads that were created in the previous day so that I can look at their countries and update the filters in the trigger campaigns if needed.

[Grégoire_Miche2]

Hi Andy,

that would work but the limit with the list of invalid countries is that it's criteria may change often, as user import new leads, for instance, while the list of valid country can be set once for all.

which is why I prefer to set a list of valid countries and exclude it. I also run a report of leads per country, filtered to show only invalid countries. So, from times to ti'es, I run that report that will give me any invalid country that is not handled by the smart campaign, and I update the smart campaign accordingly.

-Greg

[Andy_Varshneya1]

You make a valid point about the list of invalid countries being ever evolving and thus reactive rather than proactive. However, what I don't understand then is how are you updating the country values to the correct one? Surely you have to check somewhere what the invalid value is and update it accordingly. Perhaps a look at the flow you've set up would help me get on the same page.

Andy