Lead Gen Vintage Highway Traffic PPC Hero

Grow Leads 17% By Segmenting Search Partners During Bid Changes

Bid changes are a very routine optimization in any account. So routine, in fact, you might have most if not all of it automated. I know in a lot of my accounts I could write out my process for how I perform bid changes to the point that I could replicate that process with automated rules in AdWords or through a bid management system like Acquisio. However, there is something that automated bid rules, and even most people doing manual bid changes don’t account for: the difference in metrics between Google Search and Google Search Partners.

Let me go ahead and qualify this now: this is most applicable/important for accounts that have found being in top ad positions is important for their conversion volume and the return or cost-per-conversion at those positions are within goal for you. Specifically, the account I used for the case study on this generates real estate leads. Their keywords are highly targeted and quality; all exact and modified broad match. Their account structure is great. I ran their account through the WordStream AdWords Grader, and they got a 99%. Therefore, anytime they can be in the top ad position usually results in more conversions still within goal cost-per-conversion. This may not be the case for your account, so be careful!

Why Segment?

When looking at the segmented data for this account, I found the following:

Screen Shot 2014-02-07 at 11.39.50 AM

 

An astounding 88% of our impressions came from Search Partners, but only 17% of our conversions did. This means our average position of 1.4 for the account was mostly due to the impressions coming from Search Partners. The problem with that is we can see Google Search has an average position of 1.9. It’s influencing the aggregated average position much less, but due to it’s high volume of conversions, its’ actually more important. So, when we are looking at a keyword’s average position during a bid change, we’re looking at an aggregated average position, which tells us more about Search Partners than Google Search. This is a problem, since we want to strive for higher average positions on Google Search due to the high conversion volume there.

How To Segment For A Bid Change

  1. Go to the keyword tab. Set your date range for whatever your preference is when pulling bid reports. Download a keyword report for that date range without segmenting anything. Create a second tab on that report.
  2. Go back to your AdWords interface and set-up to download the same report. This time, select to segment “network (with search partners)” and download the file. Paste this data into the second tab of the other report. Delete everything that reads as “search partners” so you only have Google Search data. Then, delete all metrics except the keyword, match type, campaign, ad group, and avg. position data.
  3. Create a column in the first, aggregated data tab next to avg. position that reads “Google Search Avg. Pos”. Use the V-Lookup function to pull in the the Google Search Network avg. position into this column. I’ve included a video of what this looks like, but here are the instructions for this:
    1. Under “Insert”, select “function” and then “Vlookup”.
    2. Select the aggregated keyword column on the first tab for the first argument.
    3. Select the range of data that includes the keyword data and avg. position data on the second, Google Only, tab for the second argument.
    4. The third argument is a number, which represents the number the avg. position column is within that range. You can see me in the video counting this with my cursor!
    5. Just put FALSE for the last argument.
    6. Hit “enter”, and then drag the formula down.


If you have any results that read “N/A”, then that means that keyword didn’t have any Google Search impressions even though you had Search Network impressions.  If you’re interested in knowing more about what the heck that means and how it could happen, check out our own Sam Owen’s article that looks into it here.

Now, do your normal bid optimizations however you normally do them, but you can look at the avg. position for “google only” instead of the search partner one. You can see, sometimes there’s a difference that warrants a difference in your bid change:

Screen Shot 2014-02-07 at 12.52.19 PM

 

If I’d just looked at the aggregated data, I would have thought that 5th keyword was at position 1 and not changed it at all. However, I can see it’s actually at position 1.6 in Google Search, so I will want to increase its bid if my goal is to get to position 1 on Google search. You can also perform this Vlookup for Google Only conversions if that interests you to see.

The Case Study Results

So, here are the results!

This data is for the week before the first time we did a bid change this way:

Screen Shot 2014-02-07 at 12.56.22 PM

 

Here is the data for two weeks later:

Screen Shot 2014-02-07 at 12.57.03 PM

 

While we see an increase in CPA and cost, it’s still well within our goal CPA. Our goal here was to increase conversions, which we did by 17% overall. You can see we grew conversions on Google Search by 17% and on Search Partners by 15%. We grew conversions more on Google Search because of the concentration on its average position in our bid changes!

ppc-hero-adwords-disapprovals_search-ads

  • Asim K

    Excellent post, If you segment your superstar keyword that drive high conversion volume in to their own ad groups/campaigns you can quickly adjust their bids via advanced bid strategies.

    • Amanda West-Bookwalter

      Yes, we call those SKAGs (single keyword ad groups)!

  • Yossi E.

    Nice tactics. I’m not sure though that this is the cause for the growth in number of conversions. There is a 28% increase in spend which can result from a variety of reasons (micro and macro). One of the reasons is the end of the holidays season.
    If the reason was the average position, I’d expect a higher CTR, but the CTR decreased.
    I thought that the main achievement here is the increase in CTR for Google Search Partners, but then I noticed that the number of impressions for 10-16/1 is wrong (apparently copied by accident from Google Search).
    Am I missing something?

    • Amanda West-Bookwalter

      Thanks for pointing that out, Yossi. Yes, that was an accidental copy. Whoops!
      It’s always hard to say 100% what the cause of a change in metrics is. However, this was the only change made in the account, so we can assume this was the cause. But yes, it’s hard to rule out seasonality for sure.

      • Yossi E.

        Hi Amanda, thanks for the reply and the in-depth analysis of the search partners bids optimization. However, I think that the wrong number displayed in the number of impressions for the search partners in the second period (Jan 10th-16th) is greatly affecting the results you’re getting.

        Assuming the number of the total impressions is current, and the search partners account for 83% of all impressions (i’m assuming it’s a typo and the impressions value was copied from the box below) that would move the number of impressions up to about 1,098,000 and the real CTR would be about 0.29%.

        My point is this: if my numbers are accurate, the CTR didn’t increase but rather decrease, the conversion rate decreased and the ad spend went up, the claim that the bid management led to an increase in the number of conversions can’t be supported.

        • Amanda West-Bookwalter

          An increase in CPC = and increase in position = an increase in clicks = an increase in conversions. Even if we have a decrease in CTR or conversion rate at higher positions, we were still able to get more conversions, and we’re still in our goal CPA.

          • Yossi E.

            Not always. Not accurate.
            According to the screenshot, before the optimization process you were on position 1.4 (on average). That means that your possibility to increase traffic by increasing the bids is quite limited. That’s the reason I think there is an additional, more important, cause for the increase in general traffic on this campaign and we don’t know what that reason is.

          • Amanda West-Bookwalter

            Sorry if I didn’t provide enough data here for you to understand the influence of the bid changes. It can be hard sometimes when you’re in your account every day and working so closely with the data to represent it to the public in a way that causes everyone to understand it. But, I can assure you that the explanation provided in my last comment is precisely what happened here. The screenshot I provided is aggregated, anonymized, and not for the same date ranges in the data represented later. I was only using it to showcase how huge of a difference there is between conversions and impressions on the two networks. I understand how that can be confusing for you, and I agree you could say the shift in seasonality could have had an effect on the account. However, we’re well out of the seasonality changes after the holidays now and we’re still seeing the same results with each segmented bid change with an increase in average position on the Google Network. Thanks so much for reading and commenting! Have a lovely week!

  • Paras Joshi

    It’s a useful method and definitely works in most of the cases. in my case I am getting better CTR and conversion rate from search partners, so I will try to do it for search partners :)

    • Amanda West-Bookwalter

      Awesome! Hope it works out!

  • David B.

    Hi Amanda, great post. One question, do you separate search partner kw into a seperate campaign? What is your structure? Do you have one campaign only for KWs with Search Partner and one Campaign only for Google Search?

    • Amanda West-Bookwalter

      Hey David! Unfortunately, Google doesn’t allow you to separate search partners from Google Search at this time.