Campaign Classification

[#1]

Hi,

I have a multi-level campaign classification set up with the following features,

1. Tracking code format is something like: [region][channel][campaign][publisher][date] ..etc

2. RegEx has been used on the Rule Builder

3. Manual import of a classification file on a regular basis for mapping the values

4.  Processing rules, marketing channel processing rules have also been set up accurately

Now, we are considering code obfuscation by using numbers instead of literals. We will make changes relevantly across our settings.

Can anyone please let me know the level of impact this will have on our data collection and things to consider so we have an overview of all scenarios? Any insights will be much appreciated!

Thanks,

Sharanya

[ssharanya29 · Accepted solution]

@ssharanya29 - For classification rules, any data you've already collected and mapped should not be impacted. However, for any new values (those with numbers instead of alpha characters), you'll have to add classifications. If the numbers are predictable, you could update your classification rules to continue the auto-classification process. If they are not, you'll have to rely on the manual (or automated) uploads to classify them for you.

For Marketing Channel Processing Rules, assuming you have rules like Affiliate Marketing: tracking code begins with "aff", you'll have to update these to recognize the new identifiers, and I would suggest adding this (as an OR) before you release the new schema. Otherwise, your channels won't be properly identified until the new requirements are added.

Finally, unless you like a good fire drill, I suggest building this out and fully testing your various scenarios in a non-production report suite so you understand how it will all work out before it becomes a production issue.

[ssharanya29 · Accepted solution]

Hi @evolytics_brian, this is really helpful, thank you!

What you said about adding the new conditions as an "OR" in Marketing Channels and testing out on a Dev report suite sounds right.