Measure traffic attribution source - rely on 'entry' by timestamp?


#1

I’m new to Snowplow, medium-new to marketing analytics, so if the phrasing here is clumsy, that is why. :slight_smile:

We run Snowplow on our platform, and I am trying to find the best solution for measuring traffic sources. I have read about first vs. last touch attribution, but I am still wondering what the standard is for measuring how a user reached my platform.

Is it any refr_source or mkt_source that I can find associated with that user-session?
Or is it better to ONLY look at that user’s entry point? i.e. the very first event that was collected, ORDERing by timestamp?


#2

@goodonya, take a look at this tutorial: Web traffic driven campaign tracking with Snowplow [tutorial]. You might find some answers there.


#3

Thanks @ihor - that particular article didn’t quite get me all the way there, but it did lead to this one: First and last touch attribution models in SQL [tutorial]

Which I think gave me the info I needed.

Thanks!


#4

The good news it that there is no incorrect way to do attribution modelling. The flip side of this is that there’s no correct way to do it either.

Depending on your industry and what you’re measuring (e.g., a signup, a digital purchase, a physical purchase) the methods vary somewhat though and there’s little agreement on what technique is ‘correct’. Often this needs to be informed by what the business is, what they are trying to achieve and what the current state of play looks like.

Google Analytics provides some helpful examples of some attribution models including first click, last click, linear, decay etc but this is by no means exhaustive. If the attribution problem is across multiple media e.g., physical advertising, television etc this tends to get into the field of marketing mix modelling.