10 Call-to-Action Case Studies w/ Takeaways & Examples from Real Button Tests

10 call-to-action case studiesWhen it comes to conversion rate optimization, your call-to-action buttons are the ultimate low-hanging fruit. Even minor tweaks to button design and copy can have major impact on conversions.

Here are 10 examples from A/B tests I’ve conducted, where I’ve experimented with everything from copy and button design over to the placement of the button.

 

1. 90% increase in CTR by changing one word in the button copy

Oli from Unbounce.com and I recently ran a split test on a PPC landing page that pitches a free 30-day trial of the Unbounce.com landing page platform.

The only thing we did was to tweak one word in the copy – we changed the possessive determiner “You” to “My”. After running the test for three weeks, the treatment button copy, “Start my free 30 day trial” had increased the click through rate to the payment page by 90%.

90% increase in CTR by changing one word in the CTA

Take-Aways
The biggest change on the page doesn’t necessarily result in the biggest lift.

Conversion rate optimization really isn’t about optimizing web pages – it’s about optimizing decisions – and the page itself is a means to an end but not an end in itself.

Mission critical elements like call-to-action buttons represent a small change on the page, however, they have major impact on the decisions of your prospects.

So, when you approach a CRO project, try not to think of how to optimize the page itself – instead focus on the goal of the page and how to accelerate the decision-making process of the potential customers.

Bonus case study:

One might be inclined to label the above-mentioned test a fluke, because it seems so out of proportion that one word could have such a dramatic effect. But I’ve performed the same test on several other sites and consistently seen dramatic lifts by simply changing the possessive determiner “Your” to “My” in the CTA copy.

Here’s an example of how changing the possessive determiner from, “My” to, “Your” decreased conversion by 24.91% on a forum payment page.

Bonus case study 1

2. 35.81% increase in conversion by changing button color on product landing pages

I’ve anonymized the client here, but we’re talking about a major European e-commerce site that sells hand-painted porcelain.

In this case, we were able to increase sales via product pages (not just CTR) by 35.81% by changing the color of the call-to-action button.

35.81% increase by tweaking button color

Take-Aways

The color of your CTA button has major impact on your potential customers’ decisions – and thereby also your conversion rate.

Think of the button design/color as a visual cue that helps your prospects hone in on the button. In other words it answers the question, “Where should I click?”

The trick is to make your button stand out from the rest of the page in such a way that it’s easy to spot and identify as a clickable button.

Unfortunately, there are no set rules for which colors work, and rules like, “Never use red – it’s a stop color” or, “Green is always best – it’s a positive color” are plain stupid. I’ve seen plenty of tests were red buttons have performed well, and if you have a mostly green website, a green button is most likely going to be very difficult to spot.

It’s all about finding out what works on your specific website, and the only way of doing that is through rigorous testing.

3. 14.79% increase in conversion by adding value to the CTA copy

This is a test I recently conducted on MatchOffice.com – an international commercial real estate portal through which businesses can find offices for rent.

Once a prospect finds a relevant office, they have to click to the main CTA in order to get more information on the office via e-mail. This means that clicking the CTA is the main conversion goal, and every extra click potentially means money in the bank.

By changing the button copy from, “Order Information and Prices” to, “Get information and Prices” we increased conversions by 14.79%.

14.79% Increase in Conversion by Changing one Word in the CTA copy

Take-Aways

The copy you use in your buttons has major impact on your prospects’ decisions. Button color and design are important visual cues that tell the prospect where to click. But in the last critical moment, the copy itself is what impacts the prospect’s final decision. In other words your CTA copy answers the question, “Why should I click this button?”

The more value you can convey via your button copy, the more conversions you’ll get. The word “Order” emphasizes what you have to do – instead of what you’re going to get. Whereas, “Get” conveys value as it emphasizes what you’re going to get – rather than what you have to do to get it.

CTAs that begin with “Buy”, “Order”, “Click”, “Sign-up”, etc. inherently focus on what you have to part with. Coming up with alternatives that start with, “Get” helps you focus on answering the prospects’ number one question, “What’s in it for me?”

Bonus case study:

Just in case you think this test is a fluke, here’s an example from a Danish sister website where exactly the same exercise resulted in a lift of 38.26%. Yes, this technique also works in other languages than English.

Bonus case study 2

4. 213.16% increase in conversion by making the copy more relevant

The client here is Fitness World, a major chain of gyms in Scandinavia. The example here is taken from a PPC landing page, where the goal is to get potential customers to click through to the payment page where they can select a gym and sign up for a membership.

In this case changing the CTA copy from, “Get Membership” to, “Find Your Gym & Get Membership” increased click through to the payment page by 213.16%.

213% lift in conversion by making the CTA copy more relevant

This is the litteral translation of the Danish CTA copy.

Take-Aways

The control version is already pretty good because it conveys value and focuses on what you’re going to get – not what you have to do to get it. Nevertheless, it is very generic, “Get membership” could pretty much apply to any situation that has something to do with a membership.

I did a little research and found out that the location is a very important factor, when deciding on a membership. So, in this case I could make the call-to-action more relevant to the specific conversion scenario and increase conversions by adding “Find gym” (Step 1 in the checkout flow features a complete list of gym locations).

5. 18.01% decrease in conversion by tweaking the font color

This is an example from a test I ran on the home page of a Danish portal through which you can buy and sell used cell phones.

Here I hypothesized that I could make the button stand out more and increase CTR by changing the font color in a green button from black to yellow. What a backfire! Changing the font color actually decreased click through by 18.01%.

18.01% Decrease in Conversions by Tweaking the Font Color

Take-Aways

This simple case study illustrates that even minor tweaks to your button design can directly impact conversion. Moreover, it emphasizes the importance of testing whether your optimization efforts are in fact optimizing the performance of your website.

6. 31.03% increase in conversion on a payment page by tweaking the button copy

Here’s an example from a test I ran on the payment page for WriteWork.com – a subscription-based education website for college and university students.

In this case tweaking the button copy and adding the benefit, “Get started”, increased conversion by 31.03%. This is the very last step in the conversion funnel, and every single conversion means money in the bank.

31.03% Increase in Conversion on a Payment Page by Tweaking the Button Copy

Take-Aways

Adding a clear benefit to the call-to-action increases both the value and relevance conveyed by the copy.

Customer analyses of WriteWork’s target audience have shown that potential customers most often signup to WriteWork.com when they are in a hurry to get started with their writing process. And previous tests I’ve conducted on WriteWork.com have confirmed that adding a bit of urgency to the CTAs increase CTR on this particular website.

Therefor the, “Get started” part represents a tangible benefit to the potential customers (note that urgency doesn’t automatically have a positive impact).

7. 10.56% decrease in conversion by making the button bigger

Here’s another example from WriteWork.com. This is actually a follow-up experiment I conducted on the payment page right after the above-mentioned copy test.

The keen observer would have noticed that t adding the extra copy also increased the size of the button. So one might rightly hypothesize that the increase in size also had an effect on conversions going up.

In order to find out what effect the button size had, I ran a follow-up experiment with Treatment B where I increased button size but used the control copy in both variants.

I was quite surprised to find that Treatment B actually had a negative effect on conversions and reduced sales by 10.56%.

10.56% Decrease in Conversion by Making the Button Bigger

Take-Aways

When it comes to button size, bigger isn’t automatically better…

8. 17.18% increase in conversion by adding a text link CTA to product overview pages

This example is taken from a Danish e-commerce site that sells car care products. The site features bundles consisting of a variety of different products.

The product overview pages showing the bundle packages only feature an, “Add To Basket” button. This means that potential customers are being asked to add to product to the basket before they really know what the offer consists of

Here adding a text link CTA with the copy, “View Bundle” increased conversion to sales by 17.18%.

17.18% Increase in Conversion by Adding a Text Link CTA to Product Overview Pages

Take-Aways

Ask for the right action at the right time. It’s natural that prospects need to have a look at what the offer consists of, before they are ready to make the decision to add it to the basket.

Many websites feature super aggressive, “BUY NOW OR DIE” CTA copy by default. In most cases however, it makes sense to find out where in the decision-making process the prospect is, so you can ask for an action that’s timely rather than pushy.

9. 12.29% decrease in conversion by removing a design detail

Here’s an example from a test I recently ran on a B2B website in order to isolate the effect of a green arrow on an orange button (as well as to end a heated discussion with a very emotion-driven designer).

In this case removing the green arrow detail from the orange button decreased conversion by 12.29%.

9. 12.29% Decrease in Conversion by Removing a Design Detail

Take-Aways

Small design details like a green arrow can help make your button stand out and draw your prospects’ attention to it thus increasing conversion.

10. How a variant with the CTA below the fold generated a conversion lift of 304%

So we all know the golden rule that your call-to-action should always be positioned above the fold. Well, let’s bust that myth right away with this example from a test I ran on a B2C landing page.

In this case, the a variant with the CTA placed way below the fold at the bottom of a very long landing page significantly outperformed a variant with the CTA at the top of the page above the fold. The increase was 304%.

call to action under the fold

There are several other things going on in the treatment. So the whole lift can’t be ascribed entirely to moving the CTA below the fold. However, the fact remains that the treatment with the CTA way below the fold outperformed the control variant – something that simply shouldn’t be possible if you subscribe to the best practice rule that the CTA should ALWAYS be above the fold in order to convert.

Take-Aways

Above the fold is necessarily the best position for your call-to-action. You should place your CTA where it best compliments the decision-making process of your prospects.

I’ve seen many cases where the CTA has performed well above the fold, just like I’ve seen cases similar to this one, where having the CTA below the fold has gotten the best results. My general observation, from a wide range of landing page tests, is that there is a correlation between the complexity of the product/offer and the optimal placement of the CTA.

If the product/offer is complex, and the prospect has to digest a lot of information in order to make an informed decision, positioning the CTA lower on the page generally works best.

Vice versa, if the product/offer is very simple, and the prospect hardly has to do any thinking in order to make an informed decision, positioning the CTA above the fold generally works best.

CTA-placement-graph

Comments

  1. I love the way you make split testing so interesting. This post like most others by you gave me some funny ideas for tweeking things here and there for better conversion. “Tak Michael” ;) God påske

    • Michael Aagaard says:

      Thank you very much Christian! I’m glad you got some inspiration from my case studies!

      God påske til dig også ;-)

      - Michael

  2. One of best articles of yours. Really useful, have passed around the office, twitter and posted on my foreign blog as well. Designers will have to take notes.

    • Michael Aagaard says:

      Hi Sandy – Thank you very much for the kind words and for sharing my work!

      Plenty more case studies coming up ;-)

      - Michael

  3. Great article, just goes to show you the art of persuasion and how language and colour is so important in the use of selling! Sad thing is that a lot of designers and “marketing” people don’t really spend time chatting to their customers to understand what turns them on so to speak!

  4. Came here after having seen Max Minzer sharing the post on G+

    Really awesome tips Michael. Often companies are so engrossed with the other aspects of marketing that they ignore such facets which can make the biggest difference to sales.

    Thanks for sharing :)

    • Michael Aagaard says:

      Hi Collin – Thanks for the kind words!

      You are completely right – many businesses loose track of the small mission critical elements that have huge impact on the decisions and actions of potential customers.

      - Michael

  5. I’m blown away by the “90% increase in conversion” – great test! That word “My” actually changed completely the viewer’s perception, giving him a sense of possesion.

  6. Thanks for this articles – always nice to have articles like this as resource when designing or explaining to clients what a cta button is.

    Also as an addition to your statement that the cta color is irrelevant – the experience that President Obama held says it all. Specific choice for the copy and using a red button has paid off well :)

    Source: http://blog.optimizely.com/2010/11/29/how-obama-raised-60-million-by-running-a-simple-experiment/

    • Michael Aagaard says:

      Hi Sven – Thanks for your comment.

      But it seems you misunderstood me completely. Nowhere did I state that “CTA color is irrelevant” in fact my point is the exact opposite: CTA color has major impact on conversions! That’s the main point that case study 2 illustrates ;-)

      - Michael

    • Its not that the color of the button is irrelevant, but the “psychology” behind the colors has proved to be somehow irrelevant in these tests i.e. red is stop, which means lower conversions etc…

      • Michael Aagaard says:

        Hi Yassin – Now I understand. Yes I agree, in my experience the psychology behind colors seems to have very little impact on conversion.

        - Michael

  7. Nice set of tests! I was particularly impressed by the “your” vs “my” tests – I’ll have to try these on my site…

    • Michael Aagaard says:

      Thank you Luiz – I’d love to hear about your results!

      I’m especially interested in conducting further research on this particular topic. So please let me know what happens, if you decide to test the if you to test the “You” vs. “My” angle.

      - Michael

  8. Great post and lots of insights !

    Would love to get a sneak peek behind your customer analysis/hypothesis generation process for CRO.

    Make your day great
    yassin

  9. Very informative article on CTA, Love it ! As a Web Development Company It’s Always an Important Issue for Us, Thanks Again for the Resourceful Post.

    • Michael Aagaard says:

      Hi Russell – Thanks for reading the article! I’m glad I could be of inspiration to you.

      - Michael

  10. Awesome case studies Michael, well done!

    I had some interesting results myself this week-end with an optin popup: by changing one line of text on the optin title I doubled my cold optin rates from 8% to 15%.

    What tool sdo you use to perform your split tests? Any particular recommendations?

    All the best,
    Bertrand

    • Michael Aagaard says:

      Hi Bertrand – I’m happy to hear that you got such awesome results by tweaking your opt in copy!

      Opt in copy is one of my favorite test areas ;-)

      I use Visual Website Optimizer for all my tests.

      - Michael

      • Hi Michael,

        I came back to ask about what tool(s) you were using and I see you already answered. My biggest challenge seems to be what should I test as a change. So your ideas on My vs Your will give me a good start. Thank you.

  11. Pretty interesting cases. I think you missed one important thing – number of users that participated in that “test”. Without that info is pretty much hard to tell if that changes were good or bad.

    • Michael Aagaard says:

      Hi Piotr – Thanks for the comment.

      I’d love to be able to disclose all data on each test. Unfortunately,most clients aren’t very happy about showing their data to competitors. I include as much data as the client will let me get away with.

      However, I can tell you that the sample sizes for the case studies mentioned in this article are between 30,0000 visitors – most of the cases are based on a sample of 2,000+ visitors, a statistical confidence level of 98%, and a standard variation of 1% or less.

      The statistical confidence level should give you quite a clear indication of whether the results are trustworthy or not. Moreover, whether the changes were good or bad is can be concluded by checking if the test results were negative or positive ;-)

      If you learn more about how I validate tests, check out this video: http://contentverve.com/how-to-determine-statistical-significance-ab-test/

      - Michael

  12. Thank you very much! Now I now, that I need to do some changes on my website. Michael, I can’t see any link in your comment above.

    • Michael Aagaard says:

      Hi there – Glad I could help!

      Which comment and link are you referring to?

      - Michael

      • Your really helped:) I mean you replay to Piotr. You wrote “check out this video:” and there is no link. It’s a video about how you validate your tests.

  13. Hey Michael,

    There’s so much awesome content on this site I don’t even know where to begin. These case studies on split-testing are pure gold. “Amateurs guess…marketers test.”

    Cheers!

    JR

  14. Very informative article!

    Bruce,

  15. Insightful! Thanks for sharing!

  16. Hi Michael,

    This was VERY interesting. I liked that fact that you had similar results independent of language.

    Thank you,

    Ken

    • Michael Aagaard says:

      Hi Ken – Thank you very much!

      I’m glad you liked the article and noticed the linguistic details!

      - Michael

  17. Hey Michael,

    Thank you very much for your post ! Learned some great and interesting things from it

  18. Hurrah! After all I got a weblog from where I be capable of truly take
    helpful data regarding my study and knowledge.

  19. The in formation in this article is a real eye-opener for me. I had no idea how much difference these tweaks can make to conversions. In future, I shall come to this site for insights before moving blindly forward with my copy.

  20. Excellent post – thank you for sharing your findings!

  21. Very useful post, I want to test some thinkgs to increase my affiliate site conversion.

    MICHAEL maybe you can check my site – http://www.kredituabc.lv and give me some advices? I changed design of site 1 week ago, but cant se any increases in conversion rate.

    • Michael Aagaard says:

      Thanks for the comment Aigars!

      Unfortunately I don’t have time to go over your site right now.

      - Michael

  22. Ok, maybe you can give me answer later :) when you have time for this

  23. Michael – this is fantastic info!

    What are your thoughts on multiple CTA’s within a single page? I see more and more sites using multiple CTA’s, often spread throughout a page so that one button is always onscreen as visitors scroll down a page.

    - Chris

    • Michael Aagaard says:

      Hi Chris – Thanks for the kind words!

      On a long landing page, having multiple CTAs is a good idea. But I’d be careful with going overboard – you don’t want prospects to feel like your CTAs are chasing them around ;-)

      - Michael

  24. Hejsa Michael.

    Faldt lige over din test her – jeg bruger i øjeblikket Optimizely, og er lidt ny i det med at splitteste, men synes det er en god artikel du har bakset sammen, og det er faktisk overraskende hvor lidt der skal til for en bedre konvertering. Jeg har gemt din artikel sammen med denne: http://blog.kissmetrics.com/100-conversion-optimization-case-studies/ som jeg også mener er hammer god hvis man vil dykke endnu mere ned i konverteringsoptimering.

    • Michael Aagaard says:

      Hej Peter – tak for de pæne ord om artiklen.

      Yes, nogle gange er det overraskende så lidt, der skal til. Andre gange er det overraskende hvor meget, der skal til ;-)
      Men det finder man ud af, ved at teste, hvorvidt ens optimeringer rent faktisk virker.

      - Michael

  25. Thanks for busting the “above-the-fold” myth. Now I have a link to direct my clients when the next discussion on this topic arises since I do a lot of affiliate landing pages.

    Very informative post overall. Kudos!

    • Michael Aagaard says:

      Thanks man – Yeah, I’m all for getting that myth out of the way once and for all ;-)

      - Michael

  26. JOHN DUGGAN says:

    Mike, many thanks for doing the research and then sharing it saving us lots of work. I have shared this around a couple of my social profiles as it is valuable information and proof on why we need to test and how the marketing end of web content and calls to action is about personalising the experience for your valuable customer or lead.

    • Michael Aagaard says:

      Thank you very much for the kind words John! I-m really happy to hear that you found my research useful!

      - Michael

  27. The best known tool today for measuring in-page conversion and identifying your in-page weaknesses is ClickTale.

  28. Dear Michael,

    Thanks for your post. I have been on your website for a while.
    I am not sure why my website is not ranking is it Design, Content ? or what ?

    Thanks in advance

    Mark

  29. I was doing some research on CTAs as we have a client who has an under performing CTA and I came across your very informative article. I have shared this with the team and on my social network as I do think that so many people easily forget about the power of that little button!

    • Michael Aagaard says:

      Hi Suzanne – Thanks for sharing the article!

      Yes, that little button is one of the most important elements on any website.

      - Michael

  30. Thanks for the great article. Do you ever test or have suggestions about copy that encourages a prospect to call, i.e. Talk to Us, Talk to a Live Person, etc.?

    • Michael Aagaard says:

      Hi Aileen – Thanks for the comment and kind words!

      Yes, I recently tested “Talk to an Expert” against “Get Free Counseling”, and “Talk to an Expert” did a little better in the test

      - Michael

  31. This was so so helpful. Thank you for sharing your knowledge. I know this is hundreds of hours of work, and many more hours just to document your findings. I’m doing more web design and I want to be mindful of the little things that make a huge impact. Really excited about these goods!

  32. It seems every little change can create a great result. Can’t wait to bring it into play. Thank you for your great post!

  33. Excellent article!! A lot of the examples intuitively made sense to me, great explanation. I didn’t expect the above the fold/below the fold one to convert so much differently, but your explanation was excellent.

    Really glad i discovered this site, keep up the awesome work!

  34. Calvin Parker says:

    Do you have the actual data behind the conversion numbers? Whenever I see posts like these, I always take it with a grain of salt since you can always do fun stuff with percentages.

    Any chance you can publish the real data of hits and actions taken for each of these?

    • Michael Aagaard says:

      Hi Calvin – Thanks for the comment, I understand your frustration.

      Here’s the thing; most clients aren’t too keen on publishing the full data for a given case study. In most cases it’s a question of either publishing some data or no data at all. Whenever a client let’s me disclose full data I’m overjoyed, because that makes my case studies a lot more solid – unfortunately that rarely happens.

      The numbers you see here are based A/B tests (not periodical comparisons) and illustrate the relative difference in performance between the control and the treatment. The statistical confidence level will give you a good idea of how reliable the numbers are. So, in short, no funny stuff going on here.

      But under all circumstances, you should always take statistics with a grain of salt. A relative difference of e.g. 21% doesn’t mean that the treatment will perform exactly 21% better than the control for ever and ever.

      If you want to learn more about how I validate tests results, check out this video: http://contentverve.com/how-to-determine-statistical-significance-ab-test/

      - Michael

  35. Nice, thank you for putting all of these experiments in one, gives good food for thought and play! I liked the point about last experiment, probably having things in one column at the top with product image was more important than having direct CTA

  36. Your post is very inspiring. Thank you very much for sharing the stories. I have translated this post to Chinese and put it in my blog with stating out the author and original address. I hope you don’t mind me to do so to share your article in this way with more people in my country.

    I learnt a lot from your site. Thank you very much!
    Flora

  37. Very useful and informative post. I guess I should add privacy policy page to my site now. Thanks.

  38. Selene Gama says:

    Great information! I will definitely check my website and try to use your studies to make it better! Thanks!

  39. I am 99.98% blown away by the precision which you use in your reporting. Surely the margin of error or your results is so narrow because your results have a confidence interval of 1/100 of 1%. Thus, it’s completely appropriate to imply to your readers that the confidence in your results is within down to the .01% level.

  40. Super article Michael! Lots of great info. Better start testing right away.

  41. I am converting in aditc your content.
    Tks to share

Speak Your Mind

*