Impression Assisted Conversions
- Impression Assisted Conversions
- How Impression Assisted Conversions Work
- Adjusting for Assisted Conversions
- Ratio of Impression Assisted to Actual Conversions
- Cost per Impression Assisted Conversion
- More Information
*A conversion is when a user performs a desired action (being tracked) on your website. This could be buying a product, downloading a file, clicking a link or any other desired action. Typically, conversions are linked either directly or indirectly to revenue.
Quick Description = The no. of conversions for which an account component (keyword, campaign, etc) provided an Impression/s at some point in the user’s journey to conversion
Referred to as = Impr. Assisted conv.
A user may see several of your ads before coming to your site and converting; this metric counts the times when this occurred for each campaign, ad group, keyword ad, placement, etc.
There is a more tentative (or less solid) connection between Impression assisted conversions than with Click assisted conversions. Certainly, the ad was shown, but whether the user either saw or noticed it is an unknown variable.
As with most data analysis, there is often the need for a narrative, or story, that can be told with / by the data (or person analysing it). This metric can assist in creating such a narrative but is not as compelling as other attribution metrics.
A number between 0 and infinity represents this metric.
As mentioned, this metric should be weighted as less valuable than Click Assisted Conversions (we describe this mathematically later in this guide). This is because we cannot be sure the user saw the ad or whether they simply could have seen the ad. If we assume that a user searches 4 times and sees 3 of our ads before clicking one of them, then they go away and come back a week later click another ad and end up converting…
There is a good argument to be made that having seen multiple ads on two occasions in the lead up to selecting our business to finally make a purchase; that these impressions assisted us in converting that user.
PPC Hero did a fantastic post on this topic, which you can read by following the link. In this article they describe, based on some information from Google, how to adjust CPA (Cost Per Acquisition) with assisted conversion data.
Lets assume a basic CPA of $20 based on $200 spend and 10 conversions ($200 / 10 = $20 CPA).
Lets assume we have the following assisted conversions:
- 7 Click Assisted
- 4 Impression Assisted
In this next section, the values we use are based on what Google have said but these values really should be tailored to your business. You can try varying the values and testing to see what works for you.
- We are going to use a value of 20% (0.2) for Click Assisted Conversions, which means we believe that we are valuing a Click Assisted Conversion at 20% the value of a conversion.
- We are going to use a value of 5% (0.05) for Impression Assisted Conversions, which means we believe that we are valuing an Impression Assisted Conversion at 5% the value of a conversion.
So when we apply this, we get the following values:
7 x 0.2 = 1.4
4 x 0.05 = 0.2
Now we add these figures together along with the number of conversions (10):
10 + 1.4 + 0.2 = 11.6
So, in order to translate this into adjusted CPA we simply divide the cost ($200) by this new value (21.6):
$200 / 11.6 = $17.24 (Adjusted) CPA
Now we have an original CPA of $20 and a new CPA of $17.24, meaning that we have created $2.76 of wiggle room for our CPA target or 13.79% ($2.76 / $20). We can translate this into higher CPC’s by multiplying CPC’s by 1.1379 (to break even).
If you apply this formula to your account and adjust CPCs accordingly you should see an increase in conversions while actual CPA stays at or below the target. If you see an increase in CPA but not conversions, then this has not worked.
The variables are the 0.2 and 0.05 values; try tweaking these to get the desired results. An alternative is to use a mathematical model to generate these values; we cover this in more detail in our guide on Impression Assisted Conversion Value.
This ratio isn’t a metric we can see or add as a column within AdWords, so you will need to export data into Excel and calculate this for yourself. Simply dividing the Impression assisted conversions by the total number of conversions can do this:
Impression assisted conversions / Total number of conversions = Ratio
So if we had 1000 conversions and 500 Impression assisted conversions, we would have a ratio of:
500 / 1000 = 0.5 = 50%
What this means is that 50% of our conversions were assisted by clicks of other ads, so in this example we can confidently say that the assisted conversions are making a material difference to the volume of conversions.
If you had a low rate like 10% or less, then we could assume a shorter path to conversion and would attribute less value to keywords and ads that assist.
Again, this isn’t an AdWords metric but one that you can calculate by dividing the cost by the Impression assisted conversions to get an idea of how much these cost. For example, if we have a campaign with a cost of $400 and 100 Impression assisted conversions; we would have a cost per
Impression assisted conversion of:
$400 / 100 = $4 per Impression assisted conversion
Using this and conversion values, we can assess the profitability of Impression assisted conversions.
‘Impression Assisted Conversions’ is a single metric that makes up a suite of metrics relating to attribution. However, attribution is in itself a sub-component of both SEO and SEM and is also used across all of marketing (not just digital). Digital marketing is unique in that its attribution models are rich with data and far exceed offline marketing in their ability to provide precise insights into performance.
The image above is a simplification of a journey to conversions, this represents a user seeing and / or clicking multiple ads over a period of time. They can use multiple devices; see ads on Search, Shopping and Display networks and finally converting by buying products.
If we removed one of those ads or devices, the user may never convert on our site, they will probably still convert but elsewhere. Hence, if we look at only the last ad that converted, we miss the wider story. Attribution models look to disseminate value across the user journey in a number of different ways.
We cover these methodologies in more detail in a guide to attribution.
If you liked this guide on Impression Assisted Conversions, you may also be interested in the following guides related to this subject:
- ROAS (Return on Ad Spend)
- ROI (Return on Investment)
- Converted Clicks
- All Conversions
- Converted Click Rate
- Conversion Tracking
- Conversion Rate
- Cross Device Conversions
- Cost Per Converted Click
- Avg. CPC (Cost Per Click)
- Revenue Per Conversion
- Conversion Optimiser