Best Practices in Conversion Rate Optimization Debunked

By , July 17th, 2013 in Conversion | 21 comments
Why “Proven” CRO Principles Still Need to Be Tested
In this episode of Myth Busters… (Image source)

Have you ever come across a common misconception (like an unwarranted health scare) that was able to take hold with a large number of people? I’m sure you have.

In psychology, these sorts of widespread beliefs can roughly be classified as availability cascades, which are the:

…development of certain kinds of collective beliefs. A novel idea or insight, usually one that seems to explain a complex process in a simple or straightforward manner, gains rapid currency in the popular discourse by its very simplicity and by its apparent insightfulness.

Basically, we tend to take mental shortcuts and believe that certain “facts” are true (or are always true) because it’s easier to process things in black and white than the often far more gray reality.

When it comes to conversion rate optimization (CRO), you and I both know that there is no place for this sort of thinking. While there is plenty of room for intuition in CRO, these gut instincts need to be put to the test, and regularly.

After all, as data-driven as CRO is, it is not down to a science to an extent where certain changes will always produce similar results (hell, even actual science has a hard time with replicating results!). As Michael Aagaard from Content Verve would put it:

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.

Since different pages have different goals, one-size fits all strategy is never applicable.

With that in mind, let’s tackle some commonly held conversion principles that you should be testing instead of blindly following.

1.) Guarantees & Social Proof

Can “universally” good conversion tactics like privacy promises and social proof really fall flat under any circumstance? According to this case study published on ContentVerve, even the no-spam guarantee (a staple in CRO for email related forms) needs to be put to the test.

Take a look at the results of a split-test that contained a note promoting a 100% privacy guarantee vs. a form with no such information:

The funny thing, tests like these are where mental shortcuts can start: just because the results drastically changed in this one instance doesn’t mean that “no-spam guarantees” are dead (that’s a linkbait title for your next CRO piece).

Rather, it was perhaps that wording (remember, it’s about optimizing decisions) that threw people off.

In a follow up test, a change in copy showcases that no-spam guarantees are not in fact dead, but alive and kicking (commence with the counter linkbait headlines):

Huzzah!

A variety of other tests have shown that the same thing can occur for the highly vaunted tactics of utilizing social proof. Though it’s been pointed about before that having “low” social proof can often result in mistrust from web browsers, what if a company does have a large enough following to warrant social proof?

Consider the results that were found via a split-test on the DIYThemes blog, where the sidebar opt-in was split tested to include a basic form and a form with social proof:

A 3rd variant was tested that removed the “(it’s Free)” text from the form was included as well, and this version also included social proof.

Here were the results:

As you can see, Variation 2 (the version without social proof) performed better and actually improved over time, versus the other forms which (marginally) worsened over time and didn’t perform as well as the more basic opt-in form.

2.) Abandoning the 3-Click Rule

Another apparently established rule in UX is the three-click rule, which states that users will generally give up on something if it takes longer than three clicks.

Not only has that rule been debunked time and time again:

…but this persistent belief that you should always reduce the number of clicks to a minimum is something that doesn’t always stand up to scrutinizing via testing.

In a case study published by a Visual Website Optimizer customer, eCommerce tool Vendido showcased how a second page actually helped increase their conversions by 61%:

Moving from a single page to a two-click process should by all rights be a conversion killer, but it wasn’t the case, possibly due to the “cramming” of information on the solo page resulting in a severe case of analysis paralysis for users.

Para Chopras’ explanation of why the company originally went with this page is a perfect summation of how ‘best practices’ can go awry:

Note that it uses the so-called “best-practice” of embedding the signup form in the landing page itself. They had long been using the layout of the original page because as a best practice they presumed that reducing the number of clicks for a registration increased conversion rate… best practices are NOT always true.

3.) Me vs. You

Should I always start my free trial with your software? Or is asking me directly (‘start your free trial”) more effective?

The un-sexy answer, as analyzed in another fantastic overview from ContentVerve, is that it depends. Consider the following test conducted on Unbounce itself:

Ah, so I should be creating buttons that invoke a call to “start my free trial”, right?

Nope. Check out this test published in the same article that changed an account creation form from invoking the 1st person (my account) to an extended, far more neutral command. This small copywriting change had some drastic results:

A longer, more direct, and less personalized button performed quite a bit better than the quicker copy focusing on the self.

So instead of blindly adding “create my [blank]” to every section on your site, test a variety of different options to ensure you get the most ‘bang for your button.’

4.) Dropping the Drop-down Status Quo

Dropdowns are one of those things most designers expect an eCommerce site to have. This is in spite of the evidence that many users don’t like dropdowns and find them confusing.

In an outstanding analysis published on Visual Website Optimizer, the health site BodyEcology completely dropped their drop-down options in favor of a category page with brief product summaries:

Smriti Chawla, VSO’s resident content marketer, describes the big impact on the company’s bottom line due to these changes:

As you can clearly see via the results, the average revenue per conversion has increased significantly and now stands at $143.61, which is great if you compare it to the average revenue per conversion of the Control, which was $100.33.

Among the multiple of navigation mistakes you can make, this was one that you probably didn’t expect, but should pay close attention to, especially if you operate an online store with a variety of products.

5.) Brevity & Conversions

It’s a known fact that you don’t have much time to capture attention with your headline, and it’s also been shown that a large majority of users prefer “straight to the point” headlines over anything subtle or creative.

These proven practices have generally served marketers and web designers well, but as is the running theme in this article, they shouldn’t be accepted as being universally effective when it comes to increasing conversions.

As Michael again shows us, sometimes excess copy can destroy conversions (as one would expect). In this particular instance, the tests were done on a sign-up form:

Vanilla results, right? You would assume that brevity is king in nearly all aspects of increasing conversion rates.

However, the opposite is true in many other cases.

Take, for instance, this case study on Movexa, where a longer headline (in this instance, on a sales page) helped add clarity and ended up improving sales by 89%:

Again, we come back to the point that we aren’t simply optimizing web pages, we are optimizing a potential customer’s decision making process, and if that means adding clarity by adding a few extra words, then so be it.

This debate over brevity obviously carries over into other realms as well. Consider this test on adding more forms that showcased how extra forms did not decrease conversions and actually resulted in more quality responses:

Last but not least, you have long copy pages vs. short copy, which again produce varying results depending on the website, product being sold, and intended audience.

Consider this test conducted by Neil Patel on homepage copy length for his consulting services. Patel pitted a page with 1292 words against a page with 488 words and found that the longer page converted 7.6 percent better than the above the fold variety:

And it didn’t just end there, the leads from the long form version of the page were higher in quality than the leads that came from the new variation.

No surprise that brevity is just like ever other “proven principle” in CRO: it may work extremely well in some instances, but it’s never universal, and should always be tested.

Your Thoughts?

Here are two things I want you to do next:

  1. Leave a comment letting me know what you thought of these tests.
  2. Pick up some conversion inspiration by downloading my free guide on 10 Ways to Convert More Customers (with Psychology).

Thanks for reading!

– Gregory Ciotti


About The Author

Photo of Gregory Ciotti

Gregory Ciotti is the content strategist at Help Scout, the invisible help desk software that turns email support into a fast, personal, and pleasant experience for your customers. Get more research backed content on our blog.
» More blog posts by

Comments

  1. The guarantees and social proof experiment is surprising. I used to use the “Your privacy is safe with us” line all the time on my squeeze pages. Now i know how to boost my conversion rates by changing a few simple words. Thanks for the great advice.

  2. Mark says:

    Yes it was great article by Michael Aagaard, one that should probably printed out and pinned to the wall, and maybe sent to clients.

    The end result, as we all know, is always do your own testing

    • When an article is so good you want to print it out, you just *know* a site is worth subscribing too. ;)

      Totally agree though, everyone should make a point to add ContentVerve into their feed.

  3. David says:

    I like A/B tests reports as much as the next CRO guy, but I cannot help but feel some of them are a little disingenuous.

    The example chosen in your point 2) here is a pretty bad offender : that people would click a CTA on a simple click-through landing page (in spite of the bad copy) seems like a no-brainer — but I would like to see if any changes were made to the form itself that could drive up the conversion rate.

    I will give VSO the benefit of the doubt, but between that kind of shady stuff, the unreported confidence intervals, and the other stuff I see, I find that many CRO practitioners tend not to make much of an effort to appear trustworthy, and sometimes look like they’re just peddling their products based on pseudo-scientific evidence.

    I commend you for criticizing blind reliance on “good practices”, but I am not sure I would trade that for pseudo-scientific mystification. The testing-testing-testing chapel should be held to much higher standards of rigor.

    • Paras Chopra says:

      David,

      I can understand the reasons for skepticism where a lot of things can be just pushed under the “best practices” rug. At Visual Website Optimizer, the only case studies we report are those that are statistically significant we ourselves make sure that the best practices that we share are data and case study based. We will never put out an article / case study which we ourselves don’t believe in.

      -Paras @ Visual Website Optimizer

      • Agree with Dave to some degree. Also I see so many case studies but with no real background to thes tests, silmply the % uplift achieved. How many visitors were used for the case study? Was it statistically significant etc etc

  4. Victor says:

    I’ve seen #3 been debunked, and to be honest, I was quite surprised. Always be testing! Every business services a different sub sect of the population, and these people will have a different average tendency of doing things – so best practices might as well be best tests to conduct when it comes to CRO!

  5. Amod says:

    The takeaway here is: You will never know until you test!

  6. Jon Payne says:

    Great article Gregory!

    Question: Regarding #4 above, do you have a feel for how this might differ depending on the audience? For instance, have you observed that dropdowns hurt conversions for certain markets, but not necessarily in others? I’d be interested in the behavior is the same or different when a site’s audience is more IT-heavy or technical in nature. Certainly “go test it” is the answer but I’d love to hear your experience on this…

  7. Rubi Duong says:

    Thanks for the great advice. Now i know how to boost my conversion rates and i can understand some of this, i need it

  8. Its always suprising to me how often best pracitce fails because some has not taken into account the target demographic or the sites key USPs and bussiness.

    whenever I do any CRO work I always try to become a customer of the bussines first then a optimiser last.

    I think in most of these examples where the best practice has been proven false you could see that for this site and its users that longer headlines helped with clarity, that extra form fields made them think about way they were enquiring and that while mega menus help navigate a large site they should not be used on every site just because that is the latest design trend.

    I have a Bonus one scrolling banner are used nearly everywhere because its best practice to optimise the space you have but time and time again in my test they have killed conversions

  9. Ayub Malayev says:

    I think this is a pretty useless (and potentially harmful) article.

    “Best practices” or the science of CRO should be based on strategy, not the tactics.

    The question isn’t “should we try making it 1 click or 2 clicks” or some other implementation – the question should be “how can we decrease anxiety and cost”, “how can we increase incentive and value”.

    Each page / product / audience combo may respond differently to different tactics – but those tactics all point to the same strategic goals based on psychology.

    As a CRO practitioner, you should memorize the end goals derived from psychology, and figure out how you are going to get there for your particular page / product / audience combo. Which involves a lot of testing.

    • I think this is a pretty useless (and potentially… no, just useless) comment.

      A reoccurring point in this piece WAS that CRO should be based on strategy and user end goals, that’s probably why I included a quote that says…

      “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.”

      …right at the top of the article.

      The examples used were simply meant to highlight the dangers of blindly following “best practices” without understanding the why.

  10. There are no immutable rules, carved in stone.

    We all have to run our own tests, to make sure… and even then, the challenger that seems to beat the control today, may just fail tomorrow.
    The reason is not only in the page elements per se. The reason is in fact, that the traffic (human visitors) may be different today, from tomorrow.
    Different traffic sources will trigger different responses.

    Cheers,
    Steve ✉ Master eMailSmith ✉ Lorenzo
    Chief Editor # eMail Tips Daily Newsletter

  11. Joe says:

    Are you willing to share how much traffic you sent to each of these tests?

    I am seeing a ton of conclusions being made once the “statistical confidence” reaches 95%. Not just with these tests, but everyone who is a/b testing.

    Many times a 95% confidence is not accurate. We are not rolling dice here. These are people we deal with who are emotional and react differently every moment.

    To reach true confidence in test results, you need to have a significant number of visitors and conversions before you can even trust the confidence levels.

  12. Carol says:

    Very interesting contents.
    Conversion optimization is not as easy as it sounds.
    Promodity, a conversion engine, can help you to provide the best contents possible on your existing home page to maximize the conversion! Learn more at bit.ly/13sGaBs2 or follow us at Twitter @promodity

  13. Marino says:

    Thanks, great article!

  14. Tarrasque says:

    #4: BodyEcology has returned to the dropdown menu? I checked with multiple browsers, I got every time the Product megamenu.

  15. Mike says:

    Hi there,

    Really nice article. I think there are some truly great takeaways from this here. One thing I can’t but help but point out, is onsite content. The content you display about your products is also a huge contributing factor to a site’s conversion success. Detailed products descriptions and high quality photographs are of course contributing factors, but i’m talking about taking it one step further and using video. In this day and age, video the king of content. We’ve seen our clients at Treepodia who never once used video, make the switch and seen conversions of up to 88%. I think we’re going to be seeing a lot more of these results in the coming year, therefore I think it’s going to be worth mentioning video in coming best practice guides.

    Cheers :)

    M

  16. Nothing like examples to see what works. You might be interested in the case study of one of our recent conversion campaigns: http://localcitymarketing.com/conversion-rate-optimization-services/#guaranteed-conversions

    In short, what we accomplished in 30 days:
    $407,400 – Gross sales resulting from Engagement Marketing Service
    $5,820 – Cost for All Conversions provided to Customer
    1.43% – Percentage Cost Per Sale for Engagement Marketing Service

  17. @crotoolkit says:

    Thanks for sharing, Gregory. A great post reminding all of us of the importance of testing instead of blindly following.