SEO

Adwords Strategies: How to Pick a Winning Ad

Recently, I got into an argument with a respected colleague about how to select a winning ad in Google Adwords.  You would think that it’s an easy decision, since you have just about all the data you could possibly need.  However, we still managed to get into an argument on whether conversion rate mattered more or less than click-through rate.  Here’s how it went down…

“Hey, Brennan, this ad you made that has the price of the service we’re selling in it has a really low click-though rate,” I said.

“Yes it does, but the conversion rate is pretty good,” Brennan (my colleague) said.

I barked back, “Who cares about conversion rates?  Don’t you realize that a high click-through rate equals a low cost-per-click?  Have you read anything about quality score in the last ten years, man?

I happily wallowed in my bubble of superior knowledge for a few minutes until the jabbing spear of obvious intellectual curiosity hurt enough to pay attention to.

What if conversion rate were a more important metric?  I had always ignored it. partially because it was mainly a product of the landing page which had mostly always been out of my control.  Sure I could put in a request to change a line of text or move an image from the left side of the page to the right side of the page, but the .001% difference it would make didn’t really seem to be worth my time when I could just change the copy of an ad and go from a 7% CTR to  15% CTR.  Of course, everyone knows that the higher the CTR, the cheaper the cost per click (CPC), so why would I ever look at conversion rate?

Here’s why: The ad he created had a qualifier in it.  Example:

Hey, We sell Widgets
These widgets are awesome.
Just $5 per widget! ← QUALIFIER!
widgetsrus.com

The ad the was there previously looked something like this:

Hey, you like Widgets?
Sure, widgets are awesome.
We got lots of cheap widgets!
widgetsrus.com

Turns out that the $5 per widget qualifier was filtering out thousands of clicks that we would have otherwise payed Google thousands of dollars for clicks from customers that weren’t even willing to spend $5 for a widget.  When we actually broke down the numbers, the ad without the qualifier had a higher clickthrough rate, but the ad with the qualifier had almost four times the conversion rate.  We were saving a lot of money by running the ad with the qualifier in it and filtering out non-buyers, but were the savings worth more than the potentially sales lost due to fewer people clicking on your ad.  Or were we actually hurting ourselves by paying a higher cost per click for an ad with half the CTR?

The Numbers

After a bit of confused clucking with my colleague, we realized that the answer was staring us right in the face.  Here’s what the numbers looked like…

Original Ad (No Qualifier)   
Clicks: 25468
Impressions: 72398
CTR: 35.18%
Conversions: 28
Average CPC: $0.03
Conversion Rate: 0.11%
Cost per Conv: $3.39

New Ad (Qualifier)
Clicks: 4763
Impressions: 33026
CTR: 14.42%
Conversions: 37
Average CPC: $0.14
Conversion Rate: 0.78%
Cost per Conv: $1.62

The number we needed to look at is Impressions to Conversion, because it is telling us that if an ad has high CTR it is actually cheaper because enough non-converting potential customers will refrain from clicking it.  Furthermore, filtering out these useless clicks allows us to raise our bids which will potentially show the ad more often and make us more money at a lower cost.  The bottom line is CTR is not the holy grail of PPC.  What is?  It depends on your campaign’s goals.  If you want to drive the most traffic at the cheapest cost, then it appears that impressions to conversions is the best indicator of a winning ad.

The big-picture takeaway is that paying attention to just one metric can be useful if you’re in the middle of a combat situation (if you’re an in-house SEM rescuing a campaign from a poorly performing enemy-combatant/agency for example), but only paying attention to that one metric can be very damaging over the long term.  PPC has a lot of moving pieces, and you need to be able to juggle all of them and make sense of all of the data, otherwise you’re just a third-world dictator hiding in a drainage pipe waiting for your inevitable fate, and we all know how that ends.

 

You Might Also Like

Comments are closed.

15 thoughts on “Adwords Strategies: How to Pick a Winning Ad

  1. Agreed. So often people forget that they only have one impression to make – the one that actually converts. CPC may be a number that you don’t mind watching climb if it does translate to a higher percentage of qualified, converting leads. Thanks for the article!

  2. Always a fun discussion. Never underestimate the power of conversion rate. I wrote a blog entry June 2010 with an example of the strength of conversion rates that shows the flexibility you get in a campaign with strong conversions. (

  3. Mike, it seems to me that there is one piece of information that is missing in all of this. How much profit is being made by selling the widget. In example A it seems to take 113 leads for 1 conversion (cost per conversion / CPC) in example 2 it took 11.5 leads avg for each conversion. So the question becomes with the second example if you are driving enough traffic to make a profit? While is higher conversion rate is really desirable its almost a mute point if the item being sold doesn’t generate a high enough profit. All the metrics are important, but no matter how high the conversion rate is, if you don’t get enough traffic to make money it really doesn’t matter. That’s why its called the bottom line. Love the write up! There are really interesting things to look at!

    1. Jose, thanks for the comment. The economics of the sale and conversion rate of the landing page are outside the scope of this article. Basically what we we’re trying to figure out was how to select a winning ad from one that may look like a winning ad but is actually more expensive and less effective. Optimizing landing pages and setting price points is another animal entirely, however it is a good idea to run some tests with your pricing. For example, set up three different identical campaigns each with a different price point and corresponding landing page and see which one ends up making your client more money, while taking costs into account.

  4. Great article I would check out unbounce for testing your landing pages. Going to test out the qualifier principle with one of my clients as well. It’s good to get rid of the tire kickers.

  5. You’re joking – right?

    First – you’re buying the right visitors (who can both buy, and afford to) [Traffic]
    Second – you’re tracking *meaningful and profitable* online transactions [Conversion]
    Third – you know how much a conversion is worth to you [Economics]

    Then, ALL ads are the winners when they create conversions to your target CPA (Cost per Acquision, or cost per conversion, or cost per sale).

    I once had an adwords ad group with over 30 profitable ad variations, all creating conversions to target CPA (think Google Conversion Optimizer).

    1. If you have an ad group with 30 active ads in it then you are definitely not optimizing your account. Some ads will perform better than others and you should never have more than 3 or 4 ads running in any given ad group because:

      A) None of the ads will get enough clicks to give you the data to select a clear winner
      B) You are wasting impressions by showing poorly performing ads that don’t generate clicks
      C) You are probably lowering your quality score by showing ads that are not specific to the keyword you are targeting.

  6. The conversion rate is the only thing a client actually cares about (edit: assumptions made that the campaign is service / product based) as it is information you can provide for them to make business decisions on. Guess what, I can drop my prices by $10 because I’ve saved $X per conversion, that means I can be more competitive! and get a greater % of market share. See this by our SEM manager

    1. That’s not necessarily true. Almost every client I’ve ever worked with wants two things: more sales at a cheaper cost. You could have an ad with a 90% conversion rate that only gets clicked once every three months. Its low CTR could mean that you’re spending three times more per click than if you spent some time optimizing your ad. That is why you can’t look solely at conversion rate. Impressions/conversions is the best all-encompassing metric that will tell you the true cost and effectiveness of an ad.

      1. Click = cost
        Conversion = effectiveness

        If the ad only gets clicked once very 3 months, it’s only being paid for once every 3 months. Even if this means the click costs $3 instead of $1, if that $3 is going to get a sale 9 times out of 10 I think you are going to have a happy customer (assuming there are other ads providing more volume).

        Contrast that with an ad that has a more realistic 5% conversion rate and a $1/click and your cost per acquisition is going to be $20/sale vs. the 90% conversion rate ad at $3/click being $3.33/sale. The 90% converting ad could cost $18/click and still be only $20/sale. If you can bid $18/click instead of $1/click you are probably going to be in a pretty high position in the auction and have a decent click through rate.

        Get an ad to convert 90% of the time, then find a way to get that ad in front of as many people as possible.

  7. Price Qualifiers

    As a rule, would you say that price qualifiers in a Google PPC campaign vastly improve the impressions/conversions rate?

    In your example, the PPC ad with a price qualifier outperformed the ad w/o a price qualifier by over 700 %.

    Since I’m about to sell an ebook (and services) on Excel your information comes at a very good time.

    Thanks in advance for your answer.

    1. I have yet to see an ad where a price qualifier does not save wasted clicks. However there is always an exception. Most of the campaigns I have run have been for consumer products and services which are all typically priced within a range. I would imagine that extremely expensive and extremely cheap products and services would behave differently, as well as business to business products and services.

  8. Hi Mike
    You make a good case for qualifying an ad
    The hard part is knowing whether you will attract the right converting customers rather than simply shifting the clicks to another lowly performing metric
    I’m still going to try qualifying some ads as I guess there is not too much to lose as some of our ads perform badly

  9. Great article, the power of conversion rate tells you what proportion of the traffic people intent to buy and it will enable you to test two different versions of the same page and see which one works better.