As performance marketers, it’s our job to find the valuable clicks and filter out the crap. By tracking specific events through cookies and pixels, we attribute revenue to clicks and keywords. Newer tools like conversion funnels allow us to further identify which clicks lead to subsequent clicks which in turn lead to conversions.
Generally speaking, we use conversion funnels to show the value of early consumer touch points such as display ads, Facebook ads, or generic paid search.
So, for example, even if clicks from Facebook don’t lead to many conversions, we can determine that there is some value in guaranteeing that a user reaching your site fits a particular profile ad guaranteed by Facebook’s incredible demographic targeting.
Early touch points and clicks are often called ‘upper funnel’ clicks. A click not leading directly to a conversion is considered upper funnel if it leads to conversions after the user re-engages with the site via another channel or keyword.
By definition, an analytics system will not register a click in a conversion funnel if the user does not eventually convert. And by all logical optimization best practices, keywords and ads driving clicks not included in the funnels or directly associated with a conversion get bid down and/or removed from the marketing mix due to inefficiency.
But what if this best practice is leaving a significant percentage of potential conversions on the table? By depending on traditional conversions to measure user qualification we’re putting blinders on – grossly limiting the number of potentially qualified users we can identify.
More often than not, there is an action a user commits on a site prior to converting – for example in the retail environment, a user places an item in a shopping cart prior to purchasing it – which is not tracked as a traditional conversion by most marketing teams as a conversion event.
This data (users committing actions on a site but not converting) is the foundation for remarketing – which has been widely embraced as a strong performing channel…so why don’t other channels with more precise bid management capabilities leverage the same upper funnel qualification to assume potential value?
It just takes one extra step to understand the value of such a click and incorporate it into bid management practices.
To make this happen:
- Define a new conversion for the upper funnel event (ex: entering shopping cart)
- Define a conversion funnel for the shopping cart
Conversion funnels identify drop-off rate between sequential screens.
While they do not provide insight into the sources, keywords, or users progressing through the conversion process, they do provide a normalized conversion rate between the start and finish of a goal or action such as maneuvering through the shopping cart and making a purchase.
With a conversion indicator for the upper funnel action and a CPA – for the purpose of this example let’s say the CPA is $10 – and a 30.31% conversion rate (as shown above) between shopping carts and purchases, the theoretical cost per order is: $10 / 30.31% or $32.99.
Take is one step further by calculating a theoretical Return on Ad Spend (ROAS) by associating a channel/category average order value (AOV) and then divided by the calculated theoretical CPA. So if the AOV for similar keywords is $100 then the theoretical ROAS is $100 / $32.99 or 3.03.
The process is straight forward – all it takes is a little setup and all of the sudden we are associating potential revenue with non-converting events, providing insight into revenue forecasted to arrive as users move downstream.
By establishing a conversion event in a tool like AdWords we can further leverage conversion funnels to identify even more upper funnel terms, or multi-channel conversion funnels to identify those early touch points which driving users to the shopping cart (for which we now can associate revenue).
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.