With more people getting laid off and the economy taking a downturn, more and more people are starting their own design and development firms. Some of them will do really well. Many will design horrible, horrible websites (sorry guys & gals!).
So what’s the deal with these websites? Why do some perform better than others? Well, as you may know, every website has some goals to achieve. No website is created equal, whether you’re just building a blog, or selling stuff online, you have different goals you need to accomplish. The more a website is able to achieve its goals, the better it is. And the truth of the matter that most websites out there flat out stink. How do I know? Because we have many ways at our disposal to measure the effectiveness of these sites.
Now, if you start your own personal blog and put few hours a month writing in it, and you eventually discover that your website stinks, no big deal. However, if you are a big corporation with armies of designers and developers and yet you still manage to put out a crappy website, well someone is going to be held accountable. So, with that, lets examine which of America’s companies have the worst performing websites:
Website hall of shame
We measure poorly performing ecommerce websites by the number of orders these sites are able to capture for each 100 visitors.
Not that you have a reason to order anything from DavidsBridal or Radioshack.com
Yep, that is right: 0.25% conversion rate – To put that in perspective, a 0.25% conversion rate is worse than junk mail conversion rates.
The good news: there is something you can do about low conversion rates. The bad news: Chances are your competitors are also reading this article and will do something about their conversion rates as well – so stop wasting time and DO SOMETHING!
Back to the basics: What is a conversion rate?
Your basic conversion rate is equal to the total number of visitors your site receives during a specific time period divided by the total number of conversions for that same time period. So, if you receive 100,000 visitors in one month and you get 1000 orders in that month, then your conversion rate is = 1000/100,000 = 1%
So, what kind of conversion rate should you aim for?Are you ready for this?
Think it is impossible? Here are the top 10 converting websites from April of 2009:
So, let say you are the new VP of marketing at Radio Shack and you want to take the company from 0.75% to 10%, is that even possible?
Google Website Optimizer (GWO) to the Rescue
In case you have not been involved on the web in the last 4 years, there is a slight change you might have missed that Google released Google Website Optimizer back in 2006. Google Website Optimizer allows you to test different versions of a page and select the elements/design that persuades more of your visitors to convert.
Google Website Optimizer was not the first online testing tool how it was the first tool that made testing really, really cool! GWO provides two types of testing:
A/B Testing: A/B tests allow you to test two versions (or more) of the page against each other to see which one works best. If you are testing 3 different versions of a page, then you are conducting an A/B/C test and so on. A/B tests are especially useful because they allow you to test major design decisions by placing two or three completely different designs against each other to find out which one resonates best with your visitors. Of course, the downside is you can not pinpoint the elements that persuaded more visitors to convert. Multivariate Testing
Multivariate tests (MVT) allow you to test multiple elements of a page at the same time. So you are able to test different headlines, different images, or different colors on a single page. You can always think of A/B tests as a simplified version of multivariate tests. The beauty of multivariate tests is the agility it provides with testing multiple elements and distinguishing which combination of elements persuaded users to convert.
9 steps to designing your first A/B and multivariate test
Setting up a test in GWO is fairly straight forward:
1. You think about your test scenario: the more thought you put into designing a test experiment, the better off you will be in the long run. Poorly designed experiments take forever to run, do not produce valid results and will confuse you even further. As you are creating a test, answer the following questions:
- What are the business goals are you trying to achieve?
- What is the best strategy to test that goal?
- What page(s) are you going to test? Where should you start?
- home page
- product pages
- checkout process
- category pages
2. Are you going to conduct an A/B or Multivariate test? In most cases: A/B tests are a little easier to conduct, require less time and are used to test major design decisions. A/B tests may be a better options when there are fewer visitors to the tested page. Multivariate tests can be more complex, will test many variations of a page and will take longer to conclude. Multivariate tests are the better option when you are getting enough traffic to the tested page.
3. How many visitors does the test page receive and what is its conversion rate? This will impact how long it will take to run the test. The more visitors, the faster the test will be. The less conversions (conversions meaning the goal you have set for the visitors to complete at the end of the test: conversion page can vary depending on the goal you want to achieve), the longer the test will take.
4. What is the best conversion page for that test? More on that later
- the images,
- the copy,
- the offer,
- the headline
- the layout
- Call to action buttons
- 10 other elements on the page
- None of the above?
- All of the above?
6. Go out for a walk, sleep and come back. Seriously, I am not joking!
7. Think more about your test scenario: I lost count of how many poorly designed experiments I see every day. So, instead of having to apologize later to your boss or come up with lame excuses why an experiment failed, do everyone a favor and re-think the answers to the above questions once, twice, and maybe even 3 times!
8. Setup the test in Google website optimizer: If it’s a static page you are testing, implementing GWO is as easy as 1,2, 3. But if you are running an ecommerce store on a specific custom, open source or hosted platform ,you may run into issues and should Search for specific instructions. For example: Google website optimizer and Magento, etc.
That is it. You are ready to run the test and monitor the results! But that is the easy part of using GWO. The real work is in analyzing the results.
9. Conversion is NOT only for ecommerce stores: Every website has a conversion goal.
- If you run a blog, a conversion might be an RSS subscription, a comment from a reader or a link.
- If you run a content site (such as cnn.com), a conversion might be a page views, subscriptions forwarding a story to a friend, etc.
- If you run a lead generation site, a conversion happens when someone fills a contact form.
10. The notion that setting a test will take 5 minutes is probably by people who never implemented a test on a real website: Sorry guys! But incorporating something as user friendly as Google Website Optimizer into a dynamic ecommerce store will take a few hours the very first time. We usually recommend that you set 8 hours for setting up your first experiment.
11. Do not run tests that will take months to complete: GWO will run for months and even years if you’d like. But there are too many factors that impact your conversion rate. Some you can control, these are the ones you are testing. But there are also external factors that you have limited control over. What if a competitor runs a 50% huge sale? What if the economy tanks? The longer your experiment runs, the more you leave yourself open to external factors. We like to avoid deploying tests that take longer than 4 to 6 weeks to conclude.
12. Testing for the sake of testing is a waste of time: GWO is powerful, but it is just a tool. Testing random things just because you can do so, does not make sense. Do not let the software do the thinking, spend time thinking of what elements you should test.
13. Three Golden rules of conversion rate optimization: Memorize them, write them down, repeat them first thing in the morning and last thing at night: First Rule of Conversion Optimization: the visitor is always right. And visitors do not care how much time, money and tears invested in your website. If they do not like it, they are gone in a second!
14. Second rule of conversion optimization: Designers are great at making the pages look nice, they are rarely good at making the pages convert more.[my partner told me that I will never live this one down! really?]
15. Third rule of conversion optimization: Conversion optimization is a humbling practice: You might think you know for “a fact” how visitors will react to a design but if you have been doing CRO for a long time, you know that what you think and how visitors will react are two separate universes. My 4-year old daughter hates admitting she’s wrong, and she’s still a toddler. Imagine a grown, professional consultant. I am regularly reminded that working in conversion can is a humbling experience. If you are going to do conversion optimization, then you must be ready and aware of what is going on your website.
16. What should I test first? We can spend hours answering this question. The simple answer: There is no rule. Each site is different. Some sites should start testing with the home page. Other sites should start testing in the checkout process.
17. The notion that you should pick something, anything and just starting testing is WRONG. I hear this from speakers and experts all the time. Sorry guys, that is just wrong. If you pick the wrong page to test, you will end up concluding nothing. Your aim with the first test is to pick the area where you can have some level of impact.
18. So, really what should I test first? Well, if you insist, then take a look at your analytics and let the visitors tell you the story. Here are some “general” rules of thumb:
- If your checkout abandonment rate is more than 60% then start with the checkout process
- If your website abandonment rate more than 70%, then consider website navigation and categorization
- If your home page bounce rate is more than 50% then consider starting with the main home page
19. The fourth rule of conversion rate optimization: “general” rules of thumb can be so wrong . There is no way around looking at your analytics.
20. Just start today. Too many ecommerce stores accept low conversion rates as a fact of life. You should not. Here is a quick tip from “9 things you MUST know before you start any conversion rate optimization”
If your conversion rate is less than 40%, then there is room for improvements. So, instead of debating why I picked 40% as the mark, how about you do something! Take a look at your analytics. What pages have high bounce rates? Start making few changes to them and see if you can fix the problem. But be careful, unless you really know what you are doing you might make things a lot worse! A client of ours told us before taking on the project that they randomly added features to the site: if new features caused a major upset (dramatic drop in conversion rates), they removed them, if conversion rate was stable, they’d let them be. This is not a good approach because you aren’t gauging exactly why users would benefit from the feature, where it can be strategically more effective for them, etc. And you’re losing a lot of business because of the drastic changes being made to the site without a careful approach to measuring its impact on that specific page.
21. When it comes to traffic quality and quantity matter!
22. Home page conversion: When a visitor is on the main home page of the site and they do not leave your website, the main home page did its job and kept the visitor on the site
23. Product page conversion: When a visitor to your product page clicks on an “add to cart button,” then a product page conversion takes place.
24. Even catalog pages have a conversion: Catalog pages are funnel pages, they funnel the traffic to the product pages. When a user navigates to the product pages, then a catalog page conversion takes place.
25. Micro Vs. Macro conversion: Home page conversions, catalog pages conversions, and product page conversions are considered micros conversions. Micro conversion can take place at every page of your site. Macro conversions usually take place when a visitor places an order with your website.
26. Micro Vs. Macro testing: Not every test should have the order confirmation as the goal. If you are testing a site wide change such as new navigation, a new header or a new tag line, you should consider using the cart page as the conversion page. If you are testing the first step of the checkout process, then consider setting the 2nd page of the checkout as your conversion page.
27. Run the same test twice with different conversion pages: Running the same test scenario with different conversion pages can be very telling. You might be surprised that a combination that works well for a micro conversion performs poorly for a macro conversion.
28. The fifth rule of conversion rate optimization: Any chump can run the test scenario with different conversion pages, the real art of conversion optimization is analyzing why a combination that works well for a micro conversion performs poorly for a macro conversion. Ask yourself: hat are the lessons you learn, and what impact on the design you can conclude from running the test?
29. Site wide test: rethink your header. Is your header over crowded with information? How much information are you presenting the user with? Too many companies slap too much information into the header, we consistently find that clean and well organized headers work a lot better.
[side note: Seriously, do you see the design on the weekly specials banner? Common Frys.com hire a real designer!]
Oh yeah, using small font and slamming too much information to the header still does not work well:
30. Site wide test: your tag line can make a difference. A good tag line that communicates your value proposition will have an impact on your conversion rate. Test different tag lines to see which resonates best with your customers. A word about tag lines: think of something unique about your business. Rarely does customer service or lower price make for a good headline.
31. Site wide test: The search box. How many of your site users are currently using your search box. On most ecommerce stores, you can expect 15 to 25% of visitors to rely on the search box for navigation. A more prominent search box encourages visitors to use it. Which one of these ecommerce stores has a more prominent placement for the search box?
32. Site wide test: The navigation. Too many ecommerce stores fail to pay attention to navigation. Even worse, too many commerce stores let their technical teams and engineers decide on the site navigation.
33. Site wide: Testimonials: Never underestimate the power of a strong, credible testimonial, but be sure they are concise and on point. Testimonials are easy to figure out when it comes to B2B sites. They are more challenging when it comes to B2C ecommerce stores.
34. Testimonials done incorrectly can actually hurt your credibility: a testimonial from David in Texas stating how great your service can raise suspicions. Or who can ever forget the Espresso Machine reviews that were clearly company inspired?
35. Test placement of testimonials: Main home page, checkout process, maybe even a fan club section on your site.
36. Site wide: More elements to build credibility. Establish your credibility and build trust by using positive media coverage, industry certifications, awards, BBB memberships and standards compliances – anything that a customer would recognize as being characteristic of a reliable organization.
- Test the location of the authority & trust seals
- Test if authority seals help your site in the first place: Contrary to the common wisdom that authority seals will help you increase your conversion rates, we had cases were introducing these seals reduced the conversion rate! Moral of the story? Find what works for your site.
37. Main home page testing: what is the conversion goal? A well designed home page will convince visitors to stay on the site.
38. Main home page testing: The main image: Many ecommerce stores use a large image on their main homepage. This is very valuable real estate on your website, so pay close attention to your main homepage analytics data. What’s your bounce rate for the page? Does this image support your value proposition? How many people are scrolling down to the links?
39. Main home page testing: Getting too creative with main page image can backfire: If you have to explain how this image relates to your products or offering, then there is a good chance many of your visitors are not able to connect with that image. The image could be very clever, but that doesn’t mean that your users are going to use the same analysis to figure how image and company are related. Users land on the site and expect to see something in particular, if they are surprised by an “unrelated” image (in their mind), they will click out. You have a few seconds to convince users that they landed on the right page.
40. Main home page testing: What image works best? No matter how much you love your image, testing will show you that many visitors tend to ignore this image and the messaging you use in it.
41. Main home page testing: Flash vs. Images – In case you have not heard, flash kills conversion rates, and most visitors will ignore it. So, even your fancy flash with different images will most likely not work for your visitor. Do not take our word on it, test the flash against a still image or against text only and compare the results.
42. Main home page testing: Your headline: Do not just place an image on the main page without any messaging or headline. Take the winning image from your test above and try different headlines with it and see what works and what does not. When creating these headlines, consider the following:
43. Create headlines that appeal to the aggressive personality types: these visitors are asking: What can you do for me? Why should I choose you over your competitors?
44. Create headlines that appeal to the spontaneous personality types: these visitors are asking: I know what I am looking for, how can I place my order quickly?
45. Create headlines that appeal to the caring personality types: these visitors are asking: who else ordered from your site? How satisfied were they with your service?
46. Main home page testing: Best sellers, featured products, new arrivals. Typical of many ecommerce sites is to include a list of featured products. While this feature works very well for some sites, it may not be appropriate for ever type of site. Again, beware of generalizing.
- Consider what type of lists will resonate best with your visitors
- top sellers
- new arrivals
- discounted items
- Some other set based on custom logic
- Test the placement of these lists on the site. Should they be located in the middle section of the main home age, left or right navigation?
- When testing with any of these lists, examine the impact on including an item on these lists and the increase in sales that item reports.
47. Main home page testing: your copy. Little attention many times is paid to copy. It an important element to consider testing, because besides the copy’s impact on your search engine ranking, some of your visitors will actually take the time to scroll through the main page and read every word of the copy.
- Install some heatmap software and notice how many of your visitors are actually scrolling all the way down.
- Create visitor focused copy and place it in the middle of the page. Test the main page with copy vs. the main page without copy and see if there is any impact on your conversion rate.
48. Main home page testing: your video spokesman. Okay, I must admit. I hated them back in 2006 when they first started to appear. I’m more neutral towards them at this point in time. DO NOT take the words of the company selling a video solution. Of course they will tell you their software will increase conversion rates. Instead test to see what impact the solution has on your sales.
49. Site Search – if your website offers more than 100 products, then the internal search engine on your website becomes critical. From our experience, anywhere from 15 to 25% of site visitors will rely on site search to find the products they are looking for. Consider the following for search functionality:
- How many of your visitors are using the search function? If you have less than 10% using the search box, then the location of the search box might not be clear or prominent enough
- How are search results displayed? Presentation of results will impact visitors click through
- Allow visitors to filter through results based on feature sets (new arrivals, price, manufacturer, etc)
50. Category pages: How many is too many? Considering Categorization: how many categories, what are the various subcategories, what the different navigational paths users take; are very important question. Too many times ecommerce categorization is left to IT teams to decide. The result is an okay working navigation. Is your navigation really catering to your visitors? Examine the % of visitors who land on your main home page and how many of these visitors actually are able to make it all the way to the product pages. 40% or more indicates a good working navigation
51. Before you change the navigation, examine how minor tweaks can impact your conversion:
- The way navigation is grouped on the site will impact conversion. Test to see if different navigation grouping will impact your conversion
- Even easier is to test the labels use in navigation. Your navigation may be excellent but visitors using different words to describe category. A usability test can do wonders to gauge wording users are familiar with.
52. Category pages: the product display. Category pages are designed as funnel pages. They funnel traffic to product pages. When was the last time you clicked on “add to cart” from a category page? You will most likely navigate to the product page, read product reviews and then maybe add a product to your cart. Testing scenarios:
- Remove those add to cart button in the category pages.
- Product ids /SKUs rarely make sense on the category page. Get rid of them!
- How many products should you display on a product page? Examine your analytics, are most visitors navigating to the 2nd and 3rd page of a category page? Consider displaying more products on the 1st page!
53. Category pages: filtration options. What features are most important to your visitors? How do they shop for a particular product? Adding a filtration options on product pages enhances customer experience and allows visitors to find what they are looking for quickly. So, Test with different filtration options: Price, shipping options, best sellers, color, etc.
54. Oh, I have so many more tips to give away: How can you use GWO to test pricing on your site? how should you present payment options in the checkout process? how playing around with the proceed to checkout button can increase your conversion by 10% and so much more. However, I can’t give everything away in one post. So, if you liked this post and looking for 55 more tips, subscribe to our RSS feed.
55. If you are about to start conversion rate optimization, then these resources are a must read:
Website conversion rate 101: this is a must read collection of blogs for anyone just starting out with conversion optimization.
Personas 101: a complete guide to creating personas for your website
56. We must warn you! Google website optimizer can be extremely addictive. Side effects include random bursts of happiness (why experiments go well), mild to severe cases of depression (when tests go bad), insomnia (thinking of all the possible test scenarios), and mild case of questioning authority. In rare occasions, GWO has been reported to cause some of its users to change careers.
If you enjoyed this post, please consider subscribing to our RSS feed to have future articles delivered to your feed reader.
Or, receive weekly updates by email: