Easily Increase Sales With Assisted Impressions & Clicks

By , President at Hanapin Marketing

161 SHARES

Google has recently added search funnel data as a column in the AdWords interface. This data allows you to modify bids based on an impact a keyword has on both direct conversions (which have always been counting in the conversions column) and indirectly through a click that eventually led to a conversion through a different keyword or even an ad that got an impression from a user that eventually converted.

 

This is commonly known as attribution modeling (although a very simplistic form of it) and can lead to an overall lift in sales or leads if used correctly. This article is about how to use this data to make bid adjustments quickly and accurately.

 

I would quickly like to give a nod to one of our Google teams here, as they provided some recommendations that became the basis of this post. I don’t think I can name them but I will say that there is a team in New York that works in the Education vertical that will be getting a big pat on the back at Think EDU this year.

 

Here’s how to easily increase sales by adjusting bids based on assisted conversions and clicks:

 

  1. Go to the keywords tab.
  2. Change date range to last 90 days.
  3. Filter all keywords by any keyword with at least $10 in cost.
  4. Download a keyword report and make sure you include the following columns:
  • Campaign
  • Ad Group
  • Keyword
  • Current Max CPC
  • Match Type
  • Cost
  • Conversions (either 1 per click or multi depending on which you use)
  • Cost/Conv (either 1 per click or multi depending on which you use)
  • Impression assisted conversions
  • Click assisted conversions

5.  Open the report in excel and add the following columns (leave data blank for now)

  • Adjusted conversions
  • Adjusted CPA
  • Recommended bid

 

The headers for the spreadsheet will look something like this:

 

Assisted Click and Impression Bid Spreadsheet

 

 

Determining how to weigh impression and click assisted conversions

The default recommendation I received from Google is to count impression assisted conversions as 15% of a true conversion and click assisted conversions as 30%. In my own experience I would say these are weighted a bit high and aim for counting an impression assisted conversion as 5% of a true conversion and a click assisted conversion as 20%.

 

It is going to take some trial and error on your part to determine the perfect mix for your account(s). If you are risk adverse I would start with my numbers and if you really want to test the limits I would start with Google’s. (For the rest of this article I am going to stick with Google’s recommendations any place it is pertinent.)

 

Calculating adjusted conversions and CPL

To calculate your adjusted conversions you simply multiply the number of impression assisted conversions by .15 and add that to the number of click assisted conversions multiplied by .3 and then add that number to the true number of conversions you had. It may sound confusing but let me illustrate how simple it is with an example.

 

Let’s say you had a keyword with:

  • 10 conversions being reporting in the conversions (1 per click) column
  • $100 in cost
  • Thus a normal CPA of  $10 ($100 / 10 = $10)

 

Now let’s say you also had:

  • 5 impression assisted conversions
  • 2 click assisted conversion

 

The math for adjusted conversions would be:

 

5  * .15  = .75

2  * .3  =  .6

 

.75 + .6 + 10 = 11.35 adjusted conversions

 

Lastly, to determine your adjusted CPA you just take your total cost and divide it by your adjusted conversions.

 

$100 / 11.35 =  $8.81 adjusted CPA

 

Below is what an excel doc doing all of this would look like. The top is showing the formula and the bottom is the actual calculations. (I have hidden non-essential columns for the sake of making it easier to read.)

 

Attribution Modeling with Search Funnels

 

 

Determining bid change percentage

Carrying out the example from above, if the ultimate CPA goal is $10, you have just found yourself  $1.19 of wiggle room. That means you can increase bids by 11.9%  (1.19  / a $10 goal CPA) and keep your adjusted CPA under goal.

 

On the opposite side, if an adjusted CPA is still above goal then the keyword should be bid down (or if way too far out of tolerance it should be paused.)

 

How to measure if it is working

The only way this method works is if volume grows and total (real) CPA stays at or below goal. If calculating adjusted CPA and then bidding up just raises cost and CPA without leading to more sales then it is not  a method that you will want to continue using.

 

However, if over time you see that total sales volume increases and CPA stays below goal (even if the keywords that you are bidding up are now above goal) then this tactic is working and you should keep it up!

ACO_endad_Knockout

Twitter Facebook LinkedIn Google+ Email Print More
  • Kevin Chamberlin

    Jeff, nice post this is something I am looking forward to trying.

    • JeffreyAllen

      Glad you found it (potentially) useful!

  • TheCakeEmpire

    Could I ask, Jeff –

    You say that when the CPA Is lower, then you can afford to adjust your CPC up the full percentage of that difference.

    So, this assumes that you have CPA targets at the keyword level.

    Is it mostly the case for you that your CPAs are at the keyword level?

    For example, in my own accounts, I have the CPA at the adgroup or even campaign level, acknowledging that some keywords will have CPA’s lower, and some higher, so that overall, they are coming in at my target.

    Do you recommend having a different CPA for different keywords in your adgroup/campaign that come in at your target CPA for your campaign OR does every keyword in your adgroup/campaign have the same CPA?

    Maybe this is something that is automatically managed by PPC software?

    Sorry – it probably isn’t an appropriate question for this post – but maybe it could inspire a new post?

    (unless the entire question is completely confusing).

  • Ben

    Jeff, Can you clarify your last sentence…If the method is working, shouldn’t the CPL for the keyword effected by the method be below goal?

  • http://twitter.com/GnosisArts Gnosis Media Group

    My head hurts – I think this is fantastic. But I don’t actually have … or I’ve never really figured out – what my CPA goals are/should be. hrmph.

  • Charly Molina

    Hi Jeff

    Why do you multiply cell D4 by 1.2 ? Shouldn’t it be multiply by (1+11,9%) to increase the maximum bid exactly the wiggle room you have?

    • JeffreyAllen

      The reason I did 1.2 was that an increase in bids does not directly correlate with a direct increase in average CPC. The 1.2 I used was based on this account specifically, knowing that I usually have to overbid to reach my goal CPC/CPA based on the variance from goal to actual.

      • Charly Molina

        Thanks!! Good blog!! Love it!

      • Sena

        “1.2 I used was based on this account specifically”

        How to Calculate this 1.2?

  • Jason Ferdando

    Hi Jeff, I’m relatively new to PPC however for one of the accounts I’m working on my manager believes that generics has an impact on brand sales without statistically proving this through attribution modelling. Any andvice on how I can model this assumption?

    • JeffreyAllen

      The search funnel data tells you which keywords, ad groups or campaigns assist conversions in other keywords, ad groups or campaigns. So you can start by looking at that data to see if Brand terms have assists from generic terms. You can then do tests such as investing heavily in generic keywords for a week or two and then not at all for a week or two and see how Brand conversions change.

  • Lianne Murphy

    Hi Jeff, great article. Looking into improving our keywords performance, I have overall 11% CPA for a brand. How is this to be divided by keywords? Thanks