10 {{ upvoteCount | shortNum }}
10 {{ upvoteCount | shortNum }}

Adwords Ads have higher CTR but the CPC is also higher

by modking4
10 replies
  • PPC/SEM
    • |
I have been testing ads to try and increase my CTR and bring click prices down.

At the moment I am testing 4 ads and conventional wisdom says that I should pause the ads with the lowest CTR and create new ones to try and beat the winner.

However I am being charged a higher CPC for the Ads with a higher CTR which doesn't make sense to me.

Here are stats from the last 7 days from Adwords:

AD 1 - CTR 2.48% - Avg. CPC $0.39
Ad 2 - CTR 2.27% - Avg. CPC $0.39
AD 3 - CTR 2.68% - Avg. CPC $0.41
Ad 4 - CTR 2.51% - Avg. CPC $0.42

Which ads should I pause? The ones with the lowest CTR or the highest CPC.

Does adwords take into account the text from the ads to calculate CPC prices? All the ads are very similar in wording.
#ads #adwords #cpc #ctr #higher
  • Profile picture of the author LucidWebMarketing
    First, do you have enough clicks on each ad? If you are talking about just a handful of clicks for each, that's not enough to have enough confidence that one ad is better. How many is enough? At least 20 and I would wait for more such as 30. That's for each ad, not total.

    Other thing is to test only two ads at a time in a group. You want to give each ad equal chance at everything. Make sure the Rotate setting is on for this. If you have too many ads, especially on lower search volumes, your test of four ads is too granular which means a larger margin of error in analyzing which is actually best.

    Let's assume we be confident of the numbers you gave. Your CTRs are within a small range. There's barely even a 10% difference between the worse and second best ads. They are so close, it's basically a crap shoot.

    But you don't give average ad position which could help somewhat but I get the feeling they are also grouped closely together. The same for the CPCs you give, they are basically the same which is not surprising given the similar CTRs.

    Another thing we don't know is conversion rates. There too you want many for each ad before deciding which is the best overall for converting. I suspect there too they are similar although you may not have enough conversions for each ad to make a sound decision. It's not about getting the lower CPC, it's about ROI.

    What Google and most other PPC services do is compare your CTR to the historical CTR of the keyword at each position. It then calculates a Quality Score from this. There are other variables but CTR is by far the most important. This QS is used to calculate your price based on the QS and bid of the competitor ad below you. It's therefore mostly about getting more people to click. So in a way, your ad text is what is taken into account based on ads searchers clicked on. Higher CTR means they liked that ad better.

    So what I would do is, pause all but the 2.68% ad which is technically your best. Create a new ad to try to beat it. Don't use a similar ad, you know it doesn't work. Think about what your prospects are thinking and looking for and give it to them. Make them want to click your ad, generate an interest and find out more on your page. Then, wait until you get enough clicks and conversions. Repeat the process. Over time, you should get better at it and get higher CTRs and higher conversion rates (the best ads are those where the product of those numbers is higher). This will lead to better ROI and get higher QS which should help with improved position and/or lower costs. Have a great sales page and you will be successful.
    Signature
    Profitable PPC Campaigns Don't Happen By Accident
    {{ DiscussionBoard.errors[9580203].message }}
    • Profile picture of the author berz
      IMO , if you have conversion stats , pause two of the ads that haven't converted yet.

      Then split test the other two. Do one change at a time so that you can measure the results properly.
      {{ DiscussionBoard.errors[9580731].message }}
    • Profile picture of the author sandeepinsan
      Very nice explanation Lucid!

      I'd like to add to check your click type for devices. Clicks from mobile devices have higher CTR and desktop ads so make sure you have enough data before pausing any of these ads. Sometimes the ads with lower CTR has high conversion rate that's matter.

      The CPC isn't only depend on CTR but there any many other factors as well like landing page relevancy, site-links, QS etc.


      Originally Posted by LucidWebMarketing View Post

      First, do you have enough clicks on each ad? If you are talking about just a handful of clicks for each, that's not enough to have enough confidence that one ad is better. How many is enough? At least 20 and I would wait for more such as 30. That's for each ad, not total.

      Other thing is to test only two ads at a time in a group. You want to give each ad equal chance at everything. Make sure the Rotate setting is on for this. If you have too many ads, especially on lower search volumes, your test of four ads is too granular which means a larger margin of error in analyzing which is actually best.

      Let's assume we be confident of the numbers you gave. Your CTRs are within a small range. There's barely even a 10% difference between the worse and second best ads. They are so close, it's basically a crap shoot.

      But you don't give average ad position which could help somewhat but I get the feeling they are also grouped closely together. The same for the CPCs you give, they are basically the same which is not surprising given the similar CTRs.

      Another thing we don't know is conversion rates. There too you want many for each ad before deciding which is the best overall for converting. I suspect there too they are similar although you may not have enough conversions for each ad to make a sound decision. It's not about getting the lower CPC, it's about ROI.

      What Google and most other PPC services do is compare your CTR to the historical CTR of the keyword at each position. It then calculates a Quality Score from this. There are other variables but CTR is by far the most important. This QS is used to calculate your price based on the QS and bid of the competitor ad below you. It's therefore mostly about getting more people to click. So in a way, your ad text is what is taken into account based on ads searchers clicked on. Higher CTR means they liked that ad better.

      So what I would do is, pause all but the 2.68% ad which is technically your best. Create a new ad to try to beat it. Don't use a similar ad, you know it doesn't work. Think about what your prospects are thinking and looking for and give it to them. Make them want to click your ad, generate an interest and find out more on your page. Then, wait until you get enough clicks and conversions. Repeat the process. Over time, you should get better at it and get higher CTRs and higher conversion rates (the best ads are those where the product of those numbers is higher). This will lead to better ROI and get higher QS which should help with improved position and/or lower costs. Have a great sales page and you will be successful.
      {{ DiscussionBoard.errors[9581169].message }}
      • Profile picture of the author dburk
        Hi modking4,

        As Lucid pointed out, there is not enough of a statistical difference in your ads to choose a valid winner, unless you have tens of thousands of clicks it isn't statistically significant. That plus your statement "All the ads are very similar in wording" points to another issue. Your testing methods may need to be adjusted.

        What is the purpose of your testing? Are you trying to find out if similar wording will yield a different result? If so why? The purpose of testing shouldn't be merely to see which ad performs better, it should be to understand why one ad perform better than another. The ultimate goal of testing is to gain an actionable insight. For that to be possible you need to understand why an ad performs better, not simply which ad does. An insight only comes with the understanding of why it works, or why it doesn't work.

        With that concept in mind let's question the goal of your test, shall we?

        Are you testing to find out which value proposition resonates with your targeted audience? Similarly worded ads will not do for that for you. Are you testing headlines, CTAs? First decide what it is that you are testing, setup your test with a very specific purpose and and a clearly defined goal, one that will lead to an understanding of why one approach works better than another.

        If you change you testing methods, to be more as suggested, you are not likely to have similarly worded ads, and you are more likely to see actionable data, data that brings true insights into the mind of your targeted market. Test with a purpose that brings valuable insights.

        I recommend you start by testing value propositions within your ad messages. Do a little research on your niche and try to identify the value propositions that you think your audience might respond to positively. Then write ads to test each of those selected value propositions. When you find a significant difference, retest to confirm your results are repeatable with other other variables being isolated. At this point you will likely have an actionable insight that makes the time and effort worthwhile.

        Rinse and repeat with headlines, and then CTAs. Then use the same core concepts on testing your content, and even your overall strategies. Test with a worthwhile purpose in mind.
        Signature

        * Get Results - Outsource Your PPC Management |
        *         Don Burk Advertising & Marketing - www.donburk.com

        {{ DiscussionBoard.errors[9581977].message }}
      • Profile picture of the author LucidWebMarketing
        Originally Posted by sandeepinsan View Post

        Very nice explanation Lucid!
        Thanks.

        Originally Posted by sandeepinsan View Post

        Clicks from mobile devices have higher CTR
        That's over-generalizing. It may be true for you but not for everyone across the board. You also have to take position into account which greatly affects CTR. I could generalize here and say that positions on mobiles are closer to the top but that is not true for all (but mostly). I also suspect that as more people use mobiles to search that the numbers will regress towards the mean. In a year or two, there may be little difference in results between mobiles and computers.

        Some offers, I would not even use mobiles while for others, I want to optimize for them.
        Signature
        Profitable PPC Campaigns Don't Happen By Accident
        {{ DiscussionBoard.errors[9583495].message }}
  • Profile picture of the author austyn13x
    Originally Posted by modking4 View Post

    I have been testing ads to try and increase my CTR and bring click prices down.

    At the moment I am testing 4 ads and conventional wisdom says that I should pause the ads with the lowest CTR and create new ones to try and beat the winner.

    However I am being charged a higher CPC for the Ads with a higher CTR which doesn't make sense to me.

    Here are stats from the last 7 days from Adwords:

    AD 1 - CTR 2.48% - Avg. CPC $0.39
    Ad 2 - CTR 2.27% - Avg. CPC $0.39
    AD 3 - CTR 2.68% - Avg. CPC $0.41
    Ad 4 - CTR 2.51% - Avg. CPC $0.42

    Which ads should I pause? The ones with the lowest CTR or the highest CPC.

    Does adwords take into account the text from the ads to calculate CPC prices? All the ads are very similar in wording.


    lowest ctr would probably be best
    {{ DiscussionBoard.errors[9585292].message }}
    • Profile picture of the author LucidWebMarketing
      Originally Posted by austyn13x View Post

      lowest ctr would probably be best
      So you are saying that Ad2 at 2.27% CTR is the best? Why? What's your rationale for saying that?

      In other words, you are willing to get less traffic. Most people want more traffic but you want less. I can only guess it's because the CPC is the lowest. If so, why not Ad1 which has the same CPC but higher CTR?

      I'll assume the conversion rates are the same (and they would be). For every 100k impressions, Ad1 would get 2480 visitors and Ad2 only 2270. Whatever the conversion rate is, it's obvious Ad1 would generate more sales. Even 1% would be 25 and 23 sales respectively. In percentage terms, that's almost 9% more sales.
      Signature
      Profitable PPC Campaigns Don't Happen By Accident
      {{ DiscussionBoard.errors[9585834].message }}
      • Profile picture of the author dburk
        Originally Posted by LucidWebMarketing View Post

        So you are saying that Ad2 at 2.27% CTR is the best? Why? What's your rationale for saying that?

        In other words, you are willing to get less traffic. Most people want more traffic but you want less. I can only guess it's because the CPC is the lowest. If so, why not Ad1 which has the same CPC but higher CTR?

        I'll assume the conversion rates are the same (and they would be). For every 100k impressions, Ad1 would get 2480 visitors and Ad2 only 2270. Whatever the conversion rate is, it's obvious Ad1 would generate more sales. Even 1% would be 25 and 23 sales respectively. In percentage terms, that's almost 9% more sales.
        Hi Lucid,

        Looks to be just another drive by link dropper.

        That poster, austyn13x, didn't even bother to write a complete sentence, nor hit the shift key to capitalize the first word, nor take the time to add an extra keystroke to add punctuation. The statement seemed completely uniformed. I think it is safe to assume that austyn13x intended to provide no value to this thread, but to simply get his signature link into the thread.

        I agree that choosing the highest CTR with the lowest CPC as the winner, assuming that data reflects many thousands of impressions. However, it seems the test was not properly conceived based on the OP's description of the test. He doesn't seem to be testing anything worthwhile. I would recommend ignoring the results if this test and form a new test using a worthwhile objective, like testing completely different value propositions.
        Signature

        * Get Results - Outsource Your PPC Management |
        *         Don Burk Advertising & Marketing - www.donburk.com

        {{ DiscussionBoard.errors[9586650].message }}
        • Profile picture of the author optimize
          Hello All,

          I loved your explanation Lucid but have two more questions,

          - Do you have Analytics installed? - I never take a decision on CTR or Clicks alone. I always look where the click lives, behaves and navigates. It might be that all of your clicks for one ad are coming from one place which you will have to restrict.

          - I would also apply the 80/20 Rule. 20 % of your clicks will mean 80 % of your conversions.
          In this case 20 % of your clicks will determine your best performing CTR. Find that and create a 20 % ad group. You have all the stats.

          - On the other hand I would apply the formulas which Lucid mentioned.

          (i) Ad Rank = Max CPC x Quality Score
          (ii) Avg CPC = Ad Rank of Advertiser below you/Quality Score + 0.01
          (iii) QS = Relevancy between Keyword/Ad/URL + User Experience

          Let me know if this helped you somehow.

          Thanks

          Jan
          {{ DiscussionBoard.errors[9590766].message }}

Trending Topics