49 Split-Tests That (Almost) Always Boost Conversions

The average adult receives 147 emails a day! What are you doing to stand out in the inbox? We can help... generate more opens, clicks, and sales using our 7-Step email formula. Click here to download!

January 08,2016 by
Justin Rondeau

At DigitalMarketer we’ve conducted 1000’s of split tests and are constantly looking for new ways to move the needle on all our properties.

In this updated post, we’ll show you what we’ve spent a lot of time, money, and resources to learn.

In February 2014 we started with 43 elements that almost always worked to boost conversions. Now — almost 2 years later — we dumped 12 tests that ran their course, and added 17 tests and examples to start 2016 with better conversions.

Instead of sweating the details you should launch these as your control, monitor, and then iterate. And while you may not get the same results we did (these won’t work for everyone, but they sure did for us), you’ll have a set of solid starting points.

Here are 48 marketing elements we’ve tested time and time again, across several industries that just tend to work.

Page Layout

Page layout is incredibly important and if you’re lacking any or all of these elements, you’re going to want to fix that immediately.

1. Clear Visual Hierarchy

Okay this is a little ambiguous, but I hope this will clear things up. Simply put, you need a scannable page. If a person can’t identify

  1. who you are
  2. what you do and
  3. what to do next

… then you’re hosed.

You don’t want your elements competing with each other, so keep it simple. In general people identify images first, move to large text, then to the ‘action area’ or your CTA. Make sure to highlight the most important areas so they stand out immediately.

Can you figure out what Square wants you to do on this page?  That’s great visual hierarchy…

Split Testing - Great Visual Hierarchy

2. The Fold Still Matters

The “fold” is the area of a web page that is viewable without the visitor scrolling down.  And yes, it’s still important to put your most valuable information, calls-to-action, etc. above the fold.

There are really two camps on this issue, but simply put the fold absolutely matters. Sure, the concept of the fold has changed (it is way more elastic) due to the variety of screen sizes and devices that consumer your content. However, your message needs to remain top of mind.

Here’s a perfect example – yes, people will click and interact below the fold but notice there is a much higher click density above the fold.

split-test-conversions-img1

The ‘Download Now’ button above the fold received a total of 3687 versus the below the fold button that received only 528 clicks. That’s a 598% difference in clicks. So yeah, the fold still matters.

3. CTA That Stands Out

Whether your CTA stands out will make or break your page. There are overt ways to do this such as using arrows or contrasting colors as well as more subtle ways of doing this such as using a person’s eye-line or slight accent colors.

Either way, you need to make sure a page visitor knows which action on the page is the correct one to take.

For example when we ran this test on Survival Life we saw a 10% increase in clicks on the ‘Join FPA’ link in the navigation while maintaining Family Protection Agency membership purchases between the two variants.

split-test-conversions-img2

split-test-conversions-img2-2

In other words by increasing the sales page visit volume we were able to get more raw sales when we changed the color of the text to red.

4. Visual Cues

These are the more overt elements that push a visitor’s attention to a specific portion of your site. The classic example here is the arrow. If you point out a particular location, people will look there.

Here’s an example of a heatmap that highlights this concept.

No arrow near the CTA:

split-test-conversions-img3

Arrows near the CTA:

split-test-conversions-img4

This is not an A/B test (the offers are different and the pages are different), but illustrates the potential increase in click density if you were to guide your visitors to the CTA.

Subtle arrows draw the eye toward back to the opt–in form on your page, and they can be surprisingly effective.

split-test-conversions-img5

Notice the subtle arrow that directs your eye toward the email form?

We’ve seen as much as a 12% bump in conversions using this tactic. Be careful with this one though; you want to make sure that the content or link you’re drawing attention to is VERY strong. If you send a prospect to an order form too soon, they may not be ready to convert and you may lose the sale.

5. Avoid Sliders

This isn’t anything new! Sliders and animations disrupt eye movement and compete with your CTA (even if your CTA is just all content clicks). The more specialized you get; the more likely your slider is going to hurt conversions.

We tested this out on Survival Life to see if a content slider actually increased content consumption or just caused distractions. Though it wasn’t a huge win, we saw that when we removed the slider there was a 4.32% increase in clicks on the content links (the posts displayed in the slider window).

6

Sliders are a lazy solution to a complex problem. When you use a slider you avoid making the tough prioritization decisions that ultimately sacrifice conversions in the process

If you are stuck with using a slider, I have a few tips for optimizing your slider. Obviously it is best to get rid of it, but I know we all have political and technical constraints. Here are some tips:

  • Make sure your image file size is compressed to reduce load time
  • Turn off auto-slide
  • If you can’t turn off auto-slide, increase the time between slide panes

6. Well Constructed Headlines

Not everyone can be a headline–writing mastermind, but that’s okay. In fact, creativity can sometimes backfire. These fill–in–the–blank headlines outperform more “creative” headlines by up to 44%:

How to _______

Who Else Wants _______?

_______ in 3 Simple Steps! (numbered headlines are great)

Also, a headline with a negative slant can give your conversions a 20% bump. That’s because, unfortunately, good news doesn’t sell. Frank Kern calls this the “rubberneck effect.”

7. Involvement Devices

A quiz or poll that’s relevant to your sales message can increase engagement on your sales page. When we tested the marketing quiz below, we got a 24% higher conversion rate.

The key is to keep it relevant and include the words “see results below.” Then, of course, place the results in a space that encourages visitors to read the rest of your sales copy.

split-test-conversions-img6

Involvement devices, especially relevant polls and quizzes build engagement.

8. Triggered Pops

No matter your personal feelings about pop ups – the fact is they work! Triggered pops by their very nature are disruptive; the real trick is trying not to be offensive.

We started using triggered pops upon exit intent and we got 2,689 more leads in just 14 days. After a while we thought that there might be a better way to get an offer in front of someone.

Instead of trying a last ditch effort with visitors who are about to leave, but we should try to engage with people who are invested in our content. We tried out a triggered pop based on how far the visitor scrolls down the page and saw a 300% increase in opt-ins.

split-test-conversions-img31

Though you may not see the results we did, we know you will get a healthy bump if you try this out. 

Images

A picture is worth more than a thousand words, it could actually be the difference between a positive or negative ROI on your next campaign. We’ve run tons of image tests, and here’s what we’ve learned.

9. Use REAL People

It’s important to clarify ‘Real’ here – you want to use the face of your company or whoever is most connected with that content, e.g., the author, webinar host, customer service rep, etc…

split-test-conversions-img7

People will see through fake images and they actually hurt conversions. In nearly all of our tests on lead gen forms for webinars when we include a picture of the instructor we see a big lift in opt-in rates.

10. Avoid [Un-Edited] Stock Photos

 We all don’t have dedicated designers and photo labs, so we do have to use stock photos from time to time. The real trick is to do some basic editing on these photos to make them yours!

Across the board we’ve seen an increase in conversions when we use custom images (and an edited stock photo is just that: a custom image :) )

If you’re not a Photoshop wizard, just use something like Canva or Pixlr Express to make changes.

11. Use Image Captions

This small bit of copy is actually some of the most read copy on your page! An element’s proximity dictates how much exposure it gets and captions are close to eye-catching images!

split-test-conversions-img8

When you write your caption, make sure to add a text CTA to try and get the conversion right there.

12. Images Point To CTAs

This is a subtler tactic, but we are HARDWIRED to look where other people are looking. If you have an image of a person on the page, make sure their gaze is in the direction of your main CTA.

upload-3

13. Images With Minimal Copy

Not only are images with copy on them bad for SEO they render terribly and give your page a cheap look. Let an image be just that, an image! This will reduce load times, make your site look cleaner, and won’t cause any message mishaps if your image is broken.

split-test-conversions-img9

(Protip: use HTML buttons for this same reason)

Videos

Videos are an extremely powerful type of rich media that can help convert new and returning visitors Here’s what we’ve learned from testing our videos.

14. Don’t Use Video Fakeouts

When this was originally posted this we saw a lot of success using video fakeouts.

Well… things have changed!

If you have an element on your page that looks like a video but doesn’t function like a video, then you are going to confuse your visitors.

Sure they’ll click the element, but they won’t get the desired end result of watching the video!

In fact, we’ve found that there is no real statistical significance between having a javascript message, making the image load the opt-in form, or not making the image clickable at all.

Here are the results from just one of the tests we ran on a landing page:

split-test-conversions-img10-1

15. VSL Basics

DigitalMarketer has made a heavy shift in our use of VSLs. However, the major difference in our sales letters and VSLs, is that we’ve replaced all the body copy with a video…

…which does all the selling for us.

split-test-conversions-img11

Here are the three basic components of a good video sales page.

16. Video Length

The length of your video is definitely dependent upon the size of the offer, e,g., a shorter video can get the job down for a low dollar offer and visa versa for a high ticket item.

That said, we’ve tested tons of VSLs over the years and found the sweet spot to be between 12 and 24 minutes.

17. Remove Video Controls On VSLs

split-test-conversions-img12

Remove these controls from your VSL’s, but not your on page videos!

With sales videos, it’s important to control the message. You don’t really want your prospect to have the ability to fast forward or skip around.  This is the power of the Video Sales Letter, you’re able to make your sales arguments in the order you want to make them.

It may sound strange, but you want your visitors to have virtually NO control. Set your videos to autoplay and axe the video control bar.

18. Pause Video When Page Isn’t Active

Similar to the autoplay functions, make sure that you hold on to your audience’s attention by not letting them browse other content while your video plays in the background!

We tried this out on a few of our acquisition funnels and we saw a spike in conversions. Your VSL is only a powerful sales mechanism if you can hold your prospect’s attention – don’t let them tune you out while they read the latest clickbait article to pass the time.

19. Time Triggered CTA

split-test-conversions-img13

This buy button was invisible until the exact right moment, then POOF!

When it comes to variables that whisper, rarely do you find one that DOUBLES conversions, but the magic buy button does. When the order button isn’t visible when the prospect first arrives, your video sales page looks like regular content… not a sales page.

When your VSL reaches its call to action, shazam! Your buy button magically appears. But prospects have kept an open mind throughout your video.

20. Autoplay

Much like the video control buttons, a click–to–play video reduces the control you have over your message. More often than not, click-to-play allows the prospect to procrastinate watching the video… and they never actually watch it.

Facebook has also helped set a new standard (in fact even YouTube does this on embedded posts). The video will play on mute and the visitor has to take the extra action to hear the content. The video draws in the attention, and if the visitor is intrigued they will consume the content.

Interestingly enough it is how you launch the AUDIO on your videos that will make or break your campaign.

Think about it, you’re in an office and are taking a well-deserved break and digging through some articles when BAM your speakers start blaring some ad’s message. You don’t want your coworkers to hear that you are taking a break and quickly close the window!

Now had this happened and you were able to put headphones on or check your speaker volume the reaction would have been much different.

While testing audio we’ve seen anything from small bumps (+2%), to a huge drop off (-40%). Overall we’ve found it best to mute audio to avoid any major problems.

21. Trigger Long Form Sales Letter at the Same Time as Your CTA Button

Just having a ‘CTA’ button that appears in your VSL isn’t enough. Some people just like to read. We’ve seen a healthy conversion bump on pages where we load our full sales letter below the video to appease the more skeptical buyers. Here’s what I mean:

split-test-conversions-img14

Sales Pages

Your sales page has to be one of your most convincing pages! Don’t slip up on these easy-to-forget page elements and lose out on sales.

22. Fonts Matter

San serif fonts increase readability (and conversions)

split-test-conversions-img15

While the typeface you use on your sales page may seem trivial, we’ve seen up to a 30% higher conversion rate by simply using a san serif font. For those of you not familiar with typographer terminology, that means a font without those little “feet” at the bottom.

That means OUT with Times New Roman, and IN with Arial. We’ve found that Arial with a size 12pt font (or larger) gets the best results.

Remember, readability is everything. If a prospect has trouble reading your message, you’re screwed.

23. Reiterate CTAs

It’s not enough to have a CTA at the top and bottom of the page. You really need to pepper it in to keep the offer top of mind! The last thing you want is a customer who is ready to convert that can’t take that action immediately!

Notice on this sales page, we only have a CTA at the top and bottom of the page. Obviously it makes sense to have the CTA at the top – as that area gets the most attention (just look at the intensity on the scroll map).

There were times here where it would make sense to reiterate the CTA, but we only did this at the bottom where very few people actually arrive.

split-test-conversions-img16

The hotter areas (or the section in red) indicate that there should be another CTA in that region. More people are scrolling there than to the bottom of the page so why wait?

24. Testimonials & Social Proof

If you can fit a success story into your order form, maybe along the margin, it will boost your conversions by up to 33%. Social proof helps put prospects at ease by validating their decision to buy.

Some of you may think social proof could be a distraction, but when done correctly it will really help boost conversions. Here’s a great example:

split-test-conversions-img17

split-test-conversions-img17-2

When we added social proof to this sales page we saw a 6% increase in sales! This was only a slight lift, but we only made a slight change. If you are missing any kind of social proof (mostly really great testimonials from real customers) you are leaving money on the table.

25. Visual Deliverables (That Are Consistent With The Offer)

These can have a HUGE impact (+145%), if done correctly. Poor product images, however, are worse than none at all. We recommend you hire a professional to do your product images. Poor quality images will make you look bush–league.

Here’s an example of a nice, professional product shot:

split-test-conversions-img18

Oh – make sure that your product image is consistent with your offer. If you are offering a checklist or one-pager don’t use an ebook as your image!

(NOTE: Want DigitalMarketer’s proven 8-Step Optimization Process for turning existing traffic into more leads and more sales? Get our Optimization & Testing Specialist training and certification by clicking here.)

optimization-email-banner-2

26. Strong Ad Scent 

This is just common sense and one of those ‘best practices’ that shouldn’t even be questioned. Make sure the landing page your paid traffic lands on is consistent with the copy and imagery in the ad they clicked.

I know this seems simple enough, but this is a huge mistake people still make despite its incredible impact on conversions.

Do This:

split-test-conversions-img19

The person who clicked this ad has no question about whether the page they are on is the right page. The imagery is the same and the content on the page was represented in the as.

Not This:

split-test-conversions-img20

The ad doesn’t really set an expectation, and the landing page is espousing some other ‘business speak’ mantra. The disconnect between the ad and the page is brutal on conversions.

Checkout Pages

Your visitor has jumped through all your hoops, now you just need to reel them in! A visitor in your cart is at their highest level of anxiety, here are a few must haves that more often than not boost conversions in the cart.

27. Trust Seals

As you can imagine, many prospects are concerned with online security. Satisfaction guarantees and trust seals, even the inexpensive ones, boost conversions by 28%. Consumers just expect to see them on your order page.

split-test-conversions-img21

Trust seals are an expected part of a secure order form.

28. Price Front & Center

Admittedly we still need to make this change! That said, we’ve seen 1000’s of visitor recordings in the cart that proved to us that you need your price to be immediately seen before a visitor will take any action in your cart.

This single gif is a reflection of the behavior of thousands of people when they hit our carts.  They scroll down immediately — looking for that price.

29

Don’t make the mistake we made with your checkout page, make the price viewable above the fold.

29. Visual Deliverables

Our old cart did this much better than our updated one and we did see a dip in conversions when we removed the product image from the checkout page. The first goal of your cart is reassurance and the best way to do this is with a product shot, product title, and price.

Here’s what I mean:

split-test-conversions-img22

Thankfully a few of our high ticket items still have the product shot in the cart – and keep an eye on our cats in 2016 as we try some new ways to do this even better.

30. Linked Policies

31. No Navigation

Simply put you shouldn’t give your visitor an easy out once they are on the checkout page. There are a lot of external variables that can hinder a sale and you don’t want your own site to be one of those!

This whole ‘No Navigation’ rule is by no means applicable to every page! We tried testing this on a sales letter including traffic of people who click ‘Join Now’ on the Digital Marketer homepage and saw some VERY interesting results.

upload-2

When we removed the navigation from the sales page we saw a 20% decrease in DM Lab sales. Remember traffic source and page type matter when you are considering adding or removing navigation.

We suggest removing navigation on your checkout pages and landing pages while leaving it everywhere else!

Landing Pages

We all love a good landing page – but what makes a GREAT landing page? Here are a few things for you to try out now to get some last minute boosts to your campaigns.

32. Landing Page Basics

Here are a few of the most important factors to check on your squeeze page.

split-test-conversions-img24

  • Above the fold design – Keep your opt-in form above the fold, i.e. in the area where visitors can see it WITHOUT having to scroll. If you go below the fold use CTA buttons to reiterate your main CTA!
  • Attention-grabbing headline – Use a proven headline formula to grab attention.
  • Ultra–specific, benefit rich bullet points – Don’t list features; list what those features will do for people.
  • Definite call-to-action – Say something like “Get free instant access!” right inside your order button.

33. Product Images

 This one’s simple, including a product image on your order form boosts conversions by 22%. The image reassures customers about whatever product they’re buying.

split-test-conversions-img25

Use a professional product image whenever possible.

Pro Tip: See that progress bar at the top? That puts minds at ease, sets expectations, and boosts conversions by 15%.

34. Clear CTA 

This is simple – people can’t click what they don’t see. Make your CTA stand out and make sure the copy is indicative of the intended action.

35. Short vs. Long Form

Short form works great for opt-ins and low-ticket/impulse purchases. However, short form pages won’t cut it for your core offerings, high ticket, or purchases that require some thought.

Here’s a perfect example – we tried a short form layout for our $1 trial versus our standard sales letter. The short form layout increased clicks to checkout, but the clicks were actually WAY less qualified.

split-test-conversions-img26

In this test, the short form landing page increased clicks by 17% but really hurt sales. The longer-form sales letter led to more qualified clicks and increased sales by 116%.

36. Reiterate CTA below the fold

Similar to your sales page, you NEED to make sure that you reiterate your CTA.

We add a button below the fold, but are testing some new concepts that we’ll be able to report on next year!

37. Ditch Y-Axis Jump

38. Reports vs. Videos

This may surprise you, but our split–tests have shown that “special reports” make much more attractive lead magnets than videos.

Simply put, written reports have a higher perceived value. These consistently get us 20% higher opt–in rates than video content, which has become commonplace in the tech, forex, and marketing spaces.

Even better, presenting the exact same content in a hand–drawn format got us a 128% increase in opt–ins over a video lead magnet. Hand drawn illustrations and mind maps are golden right now.

upload-1

If there’s a way to reimagine your content as a hand drawn flow-chart or checklist, do it!

39. Form Length Coincides With Offer

Our split tests show there is no ‘magic’ form field number. You need to balance the value of your content with your longer-term business goals. If you only need an email at this stage of the funnel only ask for an email!

However, if you are giving away more valuable content you want to ask for more information! We found that some of our content didn’t convert as well with a single field opt-in because visitors were skeptical of the content’s value!

Here’s an interesting chart created by Unbounce’s co-founder Oli Gardner.

split-test-conversions-img27

Notice the two big drop off points from 2 to 3 fields and from 3 to 4 fields. If you don’t need that extra bit of information, don’t ask for it! This will make sure you are getting enough conversions!

That said, if you need at minimum 4 form fields, you might as well ask for more information because the conversion rate doesn’t really change until you hit an 8th field. It’s also important to note the lift at the end. These fields are ‘high-intent’ forms, e.g., enterprise B2B lead-gen form like this HubSpot form:

split-test-conversions-img28

Our tests have fallen in line with the findings at Unbounce’s research. As of right now we see a pretty big dip when we go from 3-4 fields. However, we will be testing even longer fields on our more premium content to see if they actually hurt conversions!

Email

We send a lot of email AND we’re always testing. Here’s what we’ve learned lately that will give your next email send the bump you need.

40. Proven Subject Lines

The purpose of an email subject line boils down to one thing and one thing ONLY: Getting the message opened.

Use these proven headline strategies to help boost your open rates:

  • Oddly Specific Numbers– Example, “Why He Paid Google $5,129,346.21”
  • Question Marks– Example, “Google Made Me Slap Proof?”
  • Percentages (%)– Example, “99% of People Dieting Need This”
  • Add [Video]– Example, “Make Your Lead Magnet Sexy [Video]”
  • Include a (free report)– Example, “Seven Deadly SEO Mistakes (free report)”
  • RE:– Example, “Re: Watch Me Build Your Membership”
  • Use Personal Pronouns– Example, “You Need to See This”

In addition, confusing, shocking, or just plain weird subject lines can boost open rates as well. Here are a few examples of random subject lines that, for whatever reason, got great open rates:

  • Photos Enclosed Do Not Bend
  • Ryan Deiss Retires?
  • Forced Continuity Dishonest?
  • Why I Love Hotel Bars
  • Chuckle–head Does $3.8M First Two Years Online

When all else fails, go negative.

  • Things Are Bad… (Maybe)
  • Don’t take it personally, but…
  • I Hate Technology
  • One Word… “Crazy”

(RELATED: [PDF DOWNLOAD] DigitalMarketer’s 101 Best Subject Lines of 2015)

41. Pain = Gain

When we began to highlight pain in our subject lines our conversion rates trended upward. This is a great strategy, but similar to Unicode symbols use this sparingly.

No one wants a negative Nancy hitting their inbox all the time. Here are two subject lines for the same content:

  • Little _______ = big sales [QUIZ]
  • This is why your prospects aren’t buying

The pain point subject line (the second one) increased CTR by 18%!

42. Shorter Subject Lines!

We’ve found that we get a bump in open rates when we use shorter subject lines instead of longer ones. When we say short we mean around 35 characters.

A short, punchy subject line gets straight to the point and leaves you plenty of room for this next conversion booster: the second subject line!

43. The Second Subject Line

The second subject line is the ‘sub headline’ of the email world. This is a great way give more clarity while using a short open inducing subject line.

split-test-conversions-img29

We’ve found that when we use a second subject line we see around a 3% bump in opens. This isn’t something you test, this is something you should do…right now! 

44. Images Increase CTR

Images linked to video are great for CTRs, but don’t forget the underlined link below.

According to our split testing, your emails should include an image whenever possible. That said, don’t overdo it and don’t format the entire email as one big image. BUT a couple of small images won’t hurt deliverability.

split-test-conversions-img30

If there’s a video included in your offer, we recommend using a screen shot of that video (play button visible) in your email. That image should be link #2. Then, underneath that image, include a blue, underlined text link. This has increased our CTRs by 120%.

45. CSS Buttons Get Higher CTR

STOP using images for buttons in your emails – they take longer to load, won’t show up if images are turned off, and depending on the amount of content in the email could actually hurt deliverability (based on the image to text ratio).

For example when we started using this in our emails we saw a healthy 22% lift in CTR and 38% lift in EPC. This was an easy modification that directly impacted engagement and sales!

Split Tests - Email CSS Buttons

46. Single Step Registration Forms

We’ve used a mult-step opt-in form across the board for the last few years. The multi-step form us very useful – you get a single CTA that doesn’t take up a lot of real estate on your page. I get it forms are bulky, but lately we’ve seen a conversion boost by keeping that form right on the page.

Here’s what we used to do (and still have a few landing pages out there with this format)

46a

We tested this against a single-step opt-in form that only asked for emails.

46b

This variation consistently out performed the multi-step variation.

47. Unicode Symbols!

split-test-conversions-img32

These are a great way to get attention in the inbox. Unicode symbols are a pattern interrupt that grab the attention of your reader. Your subject line has one goal (and only one goal) to get the recipient to OPEN the email.

We’ve found from testing Unicode symbols that they do give a nice bump to open rates (generally in the high single digits).

Word of caution: don’t use Unicode symbols in EVERY email – the practice will become stale. Pepper it in from time to time on promotions you are dying to get more eyeballs on!

48. Email Elements That Just Don’t Matter

Whenever you test you’ll find (more often than not) that there are tons of elements that just don’t move the needle. Here are a few that we have tested (and retested often) where each test flatlined:

  • Long vs. Short Emails
  • Color changes to banner ads
  • Using pictures of the face of a company versus pictures of the product in banner ads.
  • Having more CTAs in an email versus less CTAs.

It’s your turn to give us #49!

Let the DM Team and 7,944 other members in the DM Engage Facebook Group know which split test always works for your business.

Not a DM Lab Member? Learn more here.

(Connect with us at any time on Facebook or Twitter, too!)


Justin Rondeau

About Justin Rondeau

Justin Rondeau is the Director of Optimization at DigitalMarketer and runs all of the optimization efforts and split tests at DM and is active among our other properties. A top-rated domestic and international speaker, Rondeau has spent his entire career working on optimization campaigns and has helped train some of the leading optimization teams at Fortune 500 companies. Rondeau has run hundreds of tests for both B2B and eCommerce brands and has has analyzed 3,000+ tests across virtually every industry. Connect with Justin on Twitter.

View all Posts by Justin Rondeau