Here’s how to get loads more customers, free, using Google Website Optimizer.
Split testing software provides a powerful technique for increasing your website’s conversion rate (that’s its ability to turn visitors into customers). It is used by many of the web’s most powerful companies, including Amazon and Google.
Here’s our essential guide to increasing your conversion rate—it contains 108 simple techniques for growing your business.
But hold your horses, what does Google Website Optimizer do?
If you had two possible headlines for your page and you couldn’t decide which to use, you could run an A/B split test in which…
half of your visitors would be shown Headline A
the other half would be shown Headline B.
You could then count up which headline brought you the most orders.
Google Website Optimizer lets you carry out tests like this. However, tests like this often take several weeks to finish.
The real power of Google Website Optimizer comes from its ability to carry out loads of these tests at once!
For example, while you are testing which headline to use, you could also test many other “page elements”—such as text, images, prices, offers, buttons, etc.—all at the same time. Each of your visitors will see a different combination of these elements, then Google Website Optimizer will work out, on average, which of the elements performed the best. This can help you to put together a high-converting “super-page”.
For example, if we were using Google Website Optimizer on this page (the one you’re currently looking at), we could test the following…
…all at the same time!!! And Google Website Optimizer will tell you which version of each page element, on average, brought in the most customers!
Powerful, isn’t it?
This type of testing is what’s called “multivariate testing” (which just means that you’re testing many variables at once).
If you’re really interested in the details, you might like to know that there are several types of multivariate testing. Google Website Optimizer uses what’s called “full factorial” multivariate testing. Another type you might come across is “Taguchi method”. Each type has its pros and cons.
Don’t let these details put you off getting started, though. Any testing is better than none.
If you read all of this page…
…you’ll know more about conversion rate optimization (CRO) than 99% of web marketers.
And you’ll also see a lot of pictures of squirrels … for reasons that will never really become apparent.
Here’s one now…
When we say “conversion rate”, we mean the percentage of your visitors that end up reaching a given goal during the time period in question. Maybe this badly-drawn picture helps…
Typical goals include making a purchase, submitting an inquiry form, or signing up for a free newsletter.
Why you need to increase your conversion rate
There are three reasons why you need to make conversion rates your number one priority for the forthcoming year:
1. Loads of room for improvement. Most websites are losing bucket loads of money each day because they do an atrocious job of selling to their visitors.
2. Pay-per-click will keep getting more competitive. And increasing your bids is not the answer.
3. Split testing software is now highly affordable. Split testing software allows you to test changes to your website – and will tell you which changes brought in the most customers. It used to cost several thousand dollars per month. Now it’s almost free (and, in the case of Google Website Optimizer, actually free).
Unfortunately though, Google Website Optimizer doesn’t tell you what to test. That’s what we’re experts at. And this website is a “taster” of what we do.
If we doubled your website’s conversion rate, it means your cost-per-acquisition would halve.
But when your conversion rate does increase, we recommend you don’t just sit back and enjoy the profits, (tempting as that might be). Instead, we recommend you take advantage of the fact you can now afford to pay twice as much per visitor. This means…
i. You can pay about twice as much per click on AdWords, which can bring a disproportionate number of additional visitors.
ii. You can start advertising in media that had previously been too expensive for you. For example:
Full page magazine ads
iii. Your affiliates can earn twice as much as before – so the super-affiliates leave your competitors and join you.
– so, in summary, your business grows much more than you might expect.
iv. As the number of orders skyrockets, your company gets greater bargaining power with its suppliers, so its cost-per-unit-sold tends to fall – so the company becomes more efficient because of economies of scale. This means the boost to your net profit is deceptively high.
Why a doubling in conversion rate is achievable for many companies
What’s your conversion rate at the moment? 5%? 10%? Don’t know? A 10% conversion rate means that of every 10 visitors to your site, 9 walk away empty-handed. Do you really believe you couldn’t get that number down to 8 out of 10?
Look at it another way: to double your conversion rate, you just need to increase the conversion rate of…
your adverts by 25%…
…and your homepage by 25%……
…and your product page by 25%…
…and your shopping cart by 25%…
And to increase your homepage’s conversion rate by 25%, you would just have to make a 2.5% improvement to ten of the following things:
your company’s tagline, your headline, your introductory text, your offer, your guarantee, your picture, your readability, your usability, your navigation, your call to action, your products, your pricing, your offers, your premium, your testimonials, your “call to action”, your site layout, your returns policy, etc…
Does that sound more achievable?
So in summary, once you’ve finished reading this article, you need to clear your desk and start working on increasing your site’s conversion rate. And hope your competitors aren’t reading this too.
Will these techniques work for YOUR website?
Yes! Yes! Yes!
We have applied these techniques to almost all kinds of website:
business-to-business (B2B) and business-to-consumer (B2C),
in diverse industries: finance, health, travel, technology, leisure and food,
of all sizes, from start-up to corporate,
selling physical goods, services and information,
merchants, affiliates, publishers and e-commerce sites.
Here we go! These are some of the things we’d do if we were working on your site.
A word of warning: don’t be daunted by this list – if you did everything on it, you’d probably be the best marketer in the world! In reality, just one of these 108 things could be enough to double your conversion rate. The most important thing is to do something – now!
So, let’s get started!…
First you’ll need to choose which split testing software—or multivariate testing software—to use:
1. Time split tests (also known as before-and-after tests) – These stink. We only put them in the list so we could mention that. If your orders go up and down week by week (and whose don’t?) time-split tests tend to lead to wrong decisions.
2. A/B split tests – There are many occasions where a simple A/B split test is all that’s needed.
3. Multivariate testing – See the differences between all the multivariate testing solutions at…
By the way, you might have heard of “Taguchi method” or “Taguchi testing”. Taguchi is just a specific type of multivariate test.
Split testing software becomes much more powerful when you use it alongside other tools. Here are the other tools and techniques you need in your marketing arsenal…
4. Google Adwords – is often (wrongly, in our opinion) used to run split tests by creating two identical ads with different destination URL’s.
5. Live Chat – few tools tell you anything about your “non-customers” – that is, the visitors who arrived at your site then left for whatever reason. You rarely get feedback from these people. They are unlikely to phone up, but they might just be persuaded to use a Live Chat feature.
6. Web analytics – You can learn loads from your web analytics package. At the most basic level, the “site overlay” feature tells you where people click, where they don’t click and where they leave your site. Some people say “there are piles of gold waiting for you in your log files” and they are right, in a vague over-poetic kind of way.
7. Usability tests – these can be carried out on pretty much anyone you can get your hands on. They are gold dust, literally*.
* not literally.
If we could have just one testing tool, it would be usability testing. Web analytics tells you what visitors are doing, but usability testing tells you why. No other tool provides so many headslapping “I can’t believe I didn’t think of that” moments.
9. Eyetracking – shows you which things people see but don’t click. And which things people don’t click because they don’t see. Got that? Most eyetracking is carried out using custom hardware, so you need to get a company to do it for you.
10. “Poor-man’s eyetracking” – About 5 seconds into each usability test, ask the person what they have looked at so far. They usually find it easy to tell you.
Crazy Egg is a service that allows you to see the parts of your page that your visitors click on.
ClickTale is similar, but allows you to view Flash movies of your visitors’ browsing sessions. You can see how far people scroll down your pages, how they interact with your forms, and many other aspects of their visit.
12. Customer surveys – Your customers know why they ordered. And why they nearly didn’t. Ask them about it. There are loads of survey services available. Drupal has a survey module that we can plug into your website.
13. Co-opetition – short for “cooperative competition”, this is a technique by which you sell your competitors’ products from your website (usually via an affiliate program). Co-opetition can teach you a lot about your competitors’ conversion rates. And if your visitors prefer your competitors’ products, this is an easy way to find out!
Here are some tips to getting into the right mindset:
14. Stop having debates with your colleagues about who likes what. If in doubt, test. Your mantra should be “let the customers decide”. Like a pair of oversized underpants, you’ll find it strangely liberating.
15. Start to think of your business as a constantly-shifting experiment. Don’t just test your favorite ideas. Carry out tests “just to see what happens”. For example, what would happen if you lowered your prices by 30%? Or increased them by 30%? It’s the only way you’ll learn what matters to your customers.
16. Learn your other new mantra: When people in your company object to the changes you’re making, remind them that this is just an exploratory experiment to “learn what happens”, not a long-term decision. Let this become your mantra: “It’s just an experiment”, “It’s just an experiment”.
17. Copy what works for others (within limits). In particular, copy companies that appear to be tracking and testing. You can spot them because they are using the techniques in this list.
18. Copy the techniques that have been developed by people who have been testing for decades: that is, copy direct response advertisers. The internet may be new, but your visitors aren’t. For about a hundred years, direct response advertisers have been running split tests to find out what works. It’s easy to spot their ads in magazines, newspapers and direct mail – they have tracking codes and coupons on the bottom corner. And they often look a bit cluttered.
19. Place bets with your colleagues as to which of your test samples will win. You’ll be amazed at how often you are wrong. Only the top few percent of marketers appreciate that it’s impossible to always spot the winner. Race to become one of them.
20. Make sure you have great people working on this project:This is the most important job in your company. You have three options:
Do it in-house, with your best staff.
Get ‘experts’ in… and do it in-house.
Outsource it to an expert who has a vested interest in making it a big success.
21. Locate (or become) your company’s best salesperson. Your website is your electronic salesperson. It has the advantage of being able to sell to thousands of people at the same time. However, only person-to-person selling will teach you the reactions of prospects to certain types of argument and approach. It is by far the quickest and most effective way of finding out what appeals to your prospects and what doesn’t. The words on your website need to have been tested on real people. No amount of online testing will give you this gut feel. So you have a choice – either become your company’s best salesperson, or seek out the best salesperson and listen to how they sell the product.
22. Don’t test the small stuff. Test big bold changes. This has two advantages
You’ll get the results quicker (it’s a statistics thing).
You’re more likely to get big improvements.
23. Two stages of testing:
(a) Fix all the things that are “broken” (which you’ll discover during your usability tests). This is worth doing first, because it’s the easiest way to make quick improvements.
(b) Testing new ideas that have the potential to significantly grow your business. Do this next.
24. Don’t worry about temporarily lowering your conversion rate. If a test is a failure, you get one bad day of business. If a test is a success, you get a lifetime of success.
25. Don’t end the test too soon! Make sure you have enough data! Some people say you need to test for two weeks. Some people say you need to collect at least 30 orders. Some people use “gut feel”. They are all wrong. The only correct answer is to use the right statistical tool.
For split tests of AdWords ads, use this tool.
All multivariate testing software contains in-built statistical analysis.
These tools tell you whether your results are significant – or whether you haven’t got enough data yet, and they are just due to random “chance”.
26. The best place to start is by identifying the weak links in your marketing funnel. Sketch out a brief overview of your marketing funnel, from advertising all the way through to closing the sale. This will include…
Your sales force
Your product pages
Your checkout pages
Your order confirmation page
Your call center staff
How the package is sent out
27. Test stuff that your usability testees told you to change.(You are going to do usability tests, aren’t you? Promise us!)
Getting your message straight before you start
28. What’s your company’s positioning? In other words, what makes you different from or better than all your competitors? Have you ever tested it against possible alternatives? Draw up a shortlist to test – then your visitors can let you know which is most important to them!
29. Rank the top 5 points you want to communicate to your visitors. You want to make sure that, whatever else your visitors learn from your site, they definitely learn these top 5 points.
30. Consider all the different types of person who might view your site – try to write for all of them. You might find it easier to usecustomer archetypes (sometimes called “personas” or “avatars”) for this. A customer archetype is a single person who is used to represent a certain segment of visitors.
You may choose to use real people as your archetypes (for example, a customer you know well, who is characteristic of a certain segment of visitors)
or you may choose to create fictional characters who embody the characteristics of a certain segment of visitors. Warning: if you choose to use fictional characters, be sure that you’re basing them on an understanding of your real visitors. We don’t want you sitting in an ivory tower, dreaming up people who don’t exist.
31. For each page, make sure you know what all the “visitor intentions” are. For example, some visitors might be looking to make a purchase, some might be looking for customer support and others might be trying to apply for a job with you.
Instead of just guessing their intentions, survey them to find out for definite. The 4Q tool is a free easy-to-implement tool for getting started. Some of our clients choose to create their own exit surveys.
32. Test everything! Seriously. Test everything. That’s it. We’ve finished. We’re going home now.
What’s that? You want more details? OK then…
33. Identify which products bring you the most overall profit,then put them in prime position on the page, by which we mean above the fold (that is, on the upper part of the page so the user doesn’t have to scroll down to see it), preferably on the left-hand side. Mmm, that was explained well.
34. Headlines are extremely important. If your visitor doesn’t like the headline, they won’t read any further. A simple-yet-effective approach is to express your main message in a headline that:
i. is worded in terms of benefit to the customer, not in terms of product features,
ii. suggests that the person will get the results with ease,
iii. is believable (by containing some kind of proof), and
iv. is specific.
35. What you say is more important than how you say it. You’ll get the biggest improvements by changing the core message of your headline, rather than just tweaking the wording.
36. If you don’t know how to describe your product’s features in terms of benefits, carry out this exercise: imagine the customer is looking at your headline and asking “Why should I care about that?” The way you would answer their question is likely to be worded in terms of a benefit.
37. Struggling for a good headline? Adapt headlines from publications such as Cosmopolitan magazine, Reader’s Digest or MSN.com, who use formulaic headlines that have been proven to work again and again. Today MSN has “7 ADHD truths you may not know” as a headline. Replacing “ADHD” with your product name would give an instantly compelling headline.
Headlines are really important: That’s why we used up 4 of our 108 tips on them.
Plus the fact that 108 tips is starting to sound like a lot of hard work.
38. The tagline under your logo will be viewed almost as much as the headline. So make sure it clearly expresses distinct “positioning”; that is, describing what you do and how you fit into the marketplace.
39. Test high and low prices – because customers don’t always seek out the lowest prices. There’s such a thing as “reassuringly expensive”.
40. Test odd-pricing. Odd pricing is prices that end in 9’s and 7’s, which tend to sell better. Would you or I be fooled by that? No, we’re far too smart. But someone’s falling for it, because this phenomenon has been proven over and over again.
41. Test different offers. Here are some examples…
a one-month free trial
pay in installments
buy now, pay later
first one free
we’ll hold your check for 30 days
In general, do whatever you can to get the product into the customer’s hands. If you’re so confident in your product, prove it by taking some of the risk.
42. Divide your product or service into a standard version (for the prospects who are price-sensitive) and a premium version (for the ones who aren’t). This also has the psychological advantage of turning the prospect’s decision into an either/or decision , rather than a yes/no decision.
43. Even more extreme than the above two options, try changing what you sell. For example, are you selling…
the product itself
a catalogue of products
a free report about the product or about the problem
an invitation for a sales person to call
In general, the larger the purchase, the less effective it will be to attempt to sell it in one step.
44. Many of the visitors who leave without ordering will do so because you don’t offer the product or service that they are looking for. The answer is often to start selling what they are looking for, or at least become an affiliate for it.
45. Test different premiums – that is, the bonuses they get if they order. Examples include free reports, gifts and accessories.
46. Add a guarantee or test different ones. Start with the bravest guarantee you dare test. And if it works, test a braver one.
47. Add testimonials from happy customers. In general, a video testimonial is better than a testimonial with an image, which is better than a testimonial with just a name, which is better than an anonymous testimonial.
48. Add testimonials from the media. If you don’t have any, try giving them free stuff in exchange for reviews and feedback.
49. Develop a systematic way for collecting testimonials. Train your sales staff to request a testimonial whenever they receive a compliment. Email your customers asking for testimonials.
50. Test different “calls to action”. The call to action is what you want them to do next. It is often written on the ‘proceed’ button. Test direct ones such as “Buy Now And Get 10% Off” as well as indirect ones such as “Learn More”.
51. Try making the “call to action” button nice and visible. Large brightly-colored buttons often convert better – they seem to draw the reader’s attention.
52. Test different reasons why the visitor should act promptly.For example, “offer ends Wednesday”, or “only 42 tickets left”. Please note, we’re definitely not suggesting you lie to your visitors – your conversion rate depends heavily on credibility and trust. However, if you look, you’ll probably find that your own business already has real reasons why the prospect should reply promptly. If not, you can find ways of rewarding them for doing so.
53. Make the right stuff “pop”. “Pop” just means stand out. There are several ways to do this:
Highlighting important words
Hand-drawn annotation always gets attention. (This is higher-risk, and depends on the image you want to portray.)
54. A single-column layout allows you more control over the order in which your visitors view your site. When a visitor sees your page, make sure the things they see first are the things you want them to see. This is one of the reasons for the effectiveness of those long pages in the style of single-column long sales letters; because they have more control over the order in which the visitor views the page.
55. Where do people look? Eyetracking studies have shown that visitors tend to look first at the upper-left-hand area of the page, then at your headline, then at the left-hand side of the page. So put your best features there.
56. Remove clutter. Imagine that every pixel on your page either increases the conversion rate or decreases it – or just takes up space. If you can get rid of things that aren’t working, you create more space for the things that are.
57. Put all the best stuff “above the fold”. A surprising number of your visitors will not scroll at all, so it’s best to make sure that the most important content is placed “above the fold”.
58. Decide what to feature on your homepage. Write a list of the things that your visitors are looking for. Chances are, their intentions can be divided into categories and sub-categories. Allocate space on the web page according to the popularity (and value) of these categories.
59. On a similar theme, consider having a list of your top-selling items. These lists are popular, because visitors find it reassuring to buy products that others have bought.
60. Test different navigation structures.
61. If you’re confident your visitors are on the most relevant page for their needs, consider removing the navigation bar (or at least moving it somewhere less prominent). In such cases, navigation bars can be a distraction.
62. If your website has a “cool” non-conventional layout, try a conventional layout. Conventions are conventions for a reason – they make it easier for visitors to find what they are looking for.
63. Remove any distracting links that lead to places you don’t want them to go! Does your site contain any gratuitous links that you never really considered your visitors might actually click on?
64. Use a nice large font for your headline.
65. Make the first letter of your body copy a large “drop caps”letter. Drop caps letters are effective in “bridging the gap” between the headline and the body copy.
66. For the same reason, consider having your introductory paragraph in a slightly more prominent font size or appearance.
67. Test different images. The following tend to be most effective:
images of the product
images of the product being used , maybe by a “role model” character
images of the successful outcome of the product
images of happy customers holding the product (that is, a testimonial and product shot all in one).
Attention-grabbing images are great, but only if they help to communicate your sales message (which they rarely do).
68. Test giving your visitors the option to “zoom-in” to see a larger image of the product. (It’s surprising how few e-commerce sites have decent-sized images, isn’t it?)
69. Put captions under your images and test them. For some weird reason, people almost always read the captions under images.
70. Call-outs (that is, text pointing to particular parts of the picture) tend to be effective.
71. Test Violators, which are attention-getting shapes such as starbursts, ovals and banners.
72. If your page is long and requires scrolling, consider having your call to action button repeated several times on the page.
73. If your page requires scrolling, make sure that there are no “false bottoms” – that is, elements of the layout that imply the customer has reached the bottom of the page when they haven’t.
74. Many websites find they get higher conversion rates if their page is set out in the form of a sales letter with a personable one-to-one style of writing. Despite what your feelings might be about such websites, in some markets they often work.
75. There is a long debate about how much copy to include. In general, write as much as it takes to communicate all of your “sales message”, and to overcome all the likely objections. You are aiming to condense as many persuasive arguments and relevant information into as little text as possible. This will usually require more words than most websites currently use.
76. Use simple straightforward language. No reader is too sophisticated for short simple sentences.
77. Fill your body copy with benefits, not just product features.
78. Include all the information that a customer could possibly require in order to make a purchase. Note that it doesn’t all need to be on the main product page.
79. Make sure you address all the common objections that your customers bring up. As preparation for this, you might find it useful to compile a chart of objections and counter-objections, then rank them in order of importance.
80. Test different font sizes to make your text more readable.
81. Test different font colors. For body copy, black on white is usually a safe bet.
82. Near the end of the body copy, consider having a series of bullet points (or better still, ticks) that summarize the major benefits.
83. Rewrite your article for people who skim read. Use sub-heads (that is, headlines dispersed throughout, like where we wrote “Body Copy” above) and bold to make sure the right things “pop”.
84. Consider putting the start of your order form on the product page itself.
85. Audio can be very effective at selling. Xiosoft Audio is an easy way to put audio onto your website.
86. Video can be effective too. Perhaps the easiest approach is to embed YouTube videos.
There’s a service called OnSite Videos, which can be useful. You submit a script to them, then choose one of their actors to read it out. They then send you some code to add to your website, which displays the finished video, which hovers at the bottom of the browser.
You’d be surprised how many people abandon their shopping carts before they reach the checkout. In fact, your web analytics tool will show you exactly how many do.
87. Repeat your offer and main benefits on the first page of your shopping cart or order form. Some customers click on the Buy Now button just to see what the price and shipping cost will be, so you don’t want to miss out on this chance of persuading them.
88. Don’t ask for too much information, which can be tiresome and off-putting for the customer. Do you really need their fax number before they place an order? Even if it loses you a small fraction of their orders? (…which it will do).
89. Having thumbnail photos of the products in your cart can increase the likelihood of them completing the order (presumably because they feel they can’t abandon the GIFs at your checkout?!)
90. Whenever you’re asking for information, that’s the time to provide timely reassurance as to why you need it. For example,
Under the email field, say something like “We hate spam as much as you do” – and consider including the HackerSafe logo.
Under the “Order Now” button, remind them of your guarantee and returns policy.
91. Use Ajax or DHTML to hide away the bits of forms that aren’t needed. Both of these technologies allow sections of the page to be opened or collapsed without reloading the whole page.
92. Replace long dropdown lists with an Ajax alternative, to increase the chance – and the speed – of the customer finding what they are looking for.
93. Show additional ways to order – for example, by phone or by fax. Each customer has a way that they prefer to order. Sometimes the presence of the phone number itself can increase reassurance, even if the people don’t actually phone you.
94. Do you have an “enter your coupon” field on your shopping cart? Test whether this is turning people away. (People often resent ordering when they see that others are getting a better deal).
A nice bit of borrowed credibility
95. Try adding “reassurance logos” such as:
96. Test a different version of your About Us page that shows you as real likable people, not some cold faceless corporation.
97. Consistency of message: Do whatever you can to keep your message consistent all the way from advert through to order placement.
98. Immediately after the customer has ordered – or agreed to anything – they are in a particularly agreeable mood (seasoned sales people refer to this phenomenon as the “yes set” or “yes ladder”). Take advantage of this by making additional offers to them…
99. For example, a good refer-a-friend program placed on the order confirmation page can be very effective. (Speaking of friends, would any of your friends benefit from this article? Send them a link – they’ll love you for it.)
100. The order confirmation page is also a great place from which to sell other products (known as cross-selling).
101. Entry pop-ups and exit pop-ups. Be careful with these – sometimes they work well, sometimes they just irritate users.
102. View your website using different browsers and screen resolutions, to see how your customers see it. Handy tools for doing this are www.browsercam.com and browsershots.org.
103. Minimize your website’s load times (here’s a nice tool for checking your site).
104. Get your site search feature working. Google Mini andGoogle Free Web Search both enable your visitors to search your site using Google. Then use your analytics package to discover what your visitors were searching for. And then consider adding that content to your webpage – or making it more prominent.
105. Make everything clickable. Visitors click on everything, pictures in particular. And if they are clicking on something, it’s because they expect something to happen.
106. If you have advertising on your site, test that. With many advertising programmes (such as Google’s AdSense and Chitika) you can split advertising into channels. You can then test the following and measure which brings in the most revenue
Different sizes of ad
Different shapes of ad
Different positions of ad
Different ad color formats (“blend in” versus “stand out”)
107. Another way of increasing the revenue per visitor is by increasing the average Lifetime Customer Value (LCV) of people who order. This is a whole other area, which will be addressed in our newsletter.
108. Contac Bazinga! Web Design and we’ll take you by the hand through many of these points – and give you a peek into our world of conversion rate testing. It’s free, and it will change your life (for the better).