The Power of Trust Logos : The Proof is in the Numbers

Recently at Booktopia we performed an A/B test in order to ascertain whether our trust logos at the top of our site were still necessary. The following is a discussion of the testing process and the subsequent results. My hope is that by sharing these results you can understand the importance of trust logos on your site.

The Hypothesis
At Booktopia we don’t just A/B test things because we can, there is always some reasoning to justify the effort. This particular test was driven with the thought that removing our awards from our header would make it look cleaner and less cluttered. (Ask most graphic designers their opinion of our header and that’s the first thing they say).

A less cluttered header would make our search easier to find meaning more people would be inclined to search our site to find what they are looking for. With Booktopia fast becoming a well known name online our thought was that such visible trust icons were no longer necessary to convert customers.

The Variations
For this test we had the original page (the control) and a single variation with the trust logos removed. We setup the test to run for all users who visited our site, excluding crawlers (eg GoogleBot) splitting users 50/50 between the variations.

Original. With Trust Icons
Original Header

Variation. Without Trust Logos.

The Measurements
In this test, as with all our other testing, we measured a range of “Goals” however the most important for us in order to reach a conclusion were conversion rate, revenue per visitor and the percentage of customers who found and used the search bar.

The Sample Group
Our test ran across 194 817 unique visitors meaning approximately 97 000 people viewed each variation, which we believe was a big enough sample group to reach a conclusion.

The Tool
For those of you with your own sites who are interested in what tool we used to perform this test read on, for those who couldn’t care click here to skip to the results.

For all our A/B testing at Booktopia we use Optimizely. Optimizely is a dedicated A/B testing platform with a range of plans for the biggest sites right down to one man operations looking to have a bit of fun with their site. The most impressive thing about Optimizely is that once initially setup, you do not have to have any technical know-how to perform a test with the majority of testing requiring zero input from your web developers. To setup a test it’s all point a click including the moving, removal and adding of objects to a page. While this is no advert for Optimizely, I highly recommend them!

The Results!
The results were that our conversion rate decreased by 2% when the trust logos were removed and our revenue per order also decreased by 6% without the trust logos. Search use increased by 0.5%.

Our conclusion was to leave the trust logos in place. The fact that search did increase was actually irrelevant as conversion rate is the number metric I use to track our sites performance. Since conversion rate was impacted upon negatively there is no use even thinking about the removal of the logos.

We were actually quite intrigued that even though we are now a well known name in the online space, people still get a sense of comfort from seeing other more trusted brands like Telstra and BRW on our site. This just shows you that all trust logos including awards can have a dramatic effect on your website’s performance. So while they may not always look good they are definitely serving a purpose being there.

Future follow up for us here at Booktopia is to look to see if we can move our awards to another prominent place without effecting conversion rate. This would allow us to benefit from an increase in search while still maintaining our sales.

The Power of Trust Logos : The Proof is in the Numbers

Website Design for SEO, KISS!

Following today’s announcement that we, Booktopia, have won the 2012 SmartCompany Web Award for Best Search Strategy I was inspired to write this blog, to detail what we are doing at Booktopia, from a tech perspective, to be rated number one in Australia. Specifically, since our new site is only a couple of months old, what we did when designing our site to ensure good SEO.

The first thing to understand is that a good SEO strategy always goes hand in hand with a great SEM strategy. For that I have to thank my colleagues, who, for the last 8 years who have tirelessly built up one of the best (and biggest) Google Adword accounts I have ever seen. I will have to leave this for another post, but you’ll just have to trust me, its awesome!

Nevertheless, when it comes to SEO what we have done in terms of website design is nothing fancy. If you’re looking for the latest hacks for SEO, you’re in the wrong place. Why? Because I don’t believe it’s that complicated. When it comes to website design I believe in the old KISS (Keep It Simple, Stupid) principle. Below are my top three points that I believe are important when thinking SEO for website design.

1. Design SEO from the ground up
SEO needs to be thought about when you’re designing a site. I too often see amazing looking website designed and built, and then the team looks around and I hear them say “OK, now lets get to number one on Google”. In the majority of cases this will involve rewriting large parts of the code to make it “SEO friendly”.

So what design strategies did we implent at Booktopia? Having only released our new site a couple of months ago, the following tips are fresh off the press. (Remember I said it’s simple!)

  • No <table>’s, unless you’re presenting data. Use <div>’s for presentation and lists, <ul>’s, for menus, results or any groups of things that are related.
  • Use CSS. Avoid HTML tags for presentation. Eg. <B>, <U>.
  • We avoided Flash. If you’re going down the flash road you’re looking for trouble. I’m not saying it’s impossible to do good SEO on a site which uses flash, but it will be significantly more work.
  • Remember your <alt> and <title> tags. Googlebots can’t read your images (yet), and so need your help to know what the images are about. Using <alt> and <title> tags allows you to do just this. (It doesn’t hurt to also help Google with links). These alt tags can also help you increase your keyword density.
  • Don’t forget your meta tags, nofollow’s and noindex’s. While many people say Google no longer uses the description and keyword meta tags, having them in your source can’t hurt.

2. Design your site for customers, not robots
Google wants to see content that is designed for humans not the Googlebots. Don’t pollute your site with keywords that make no sense to the customer. Google will always rank sites higher that have natural language and they can see through sites with extremely high keyword densities. Furthermore, for customers, these keywords are noise and will lead to a higher bounce rate and consequently a lower quality score, further hurting your SEO.

3. Make sure your site loads quickly
While I can’t tell you how much weighting Google actually gives to page load times, I definitely believe Google does penalise slow loading sites. As judge of the Smart Company awards Jim Stewart puts it, “A faster site in a highly competitive space will always give you an edge for ranking.”

  • At Booktopia, our favourite tools include Google’s Page Speed and Yahoo’s YSlow. Both of these tools can give you great insights into the load times of your pages, and more importantly tangible action items to help you increase your “Page Speed Score” or “YSlow Grade”.
  • Optimise ALL your images. Where possible optimise all your images before uploading them to your site. Adobe Photoshop’s “optimized” versions are not always good enough. Get yourself a lossless image optimisation tool. My tool of choice is ImageOptim (for Mac). Optimising images saves us up to 80% on images supplied to us by publishers, dramatically decreasing the load time of the respective pages.

Page load speed is not only a front end thing. At Booktopia we have spent a lot of time on our in-site search speed. Search and category browse makes up for over 90% of the requests on our site, so it’s important to us that it happens quickly. While I can’t give you the secret sauce to our search, I can tell you that under load (click frenzy load!), on average, a search on our site takes approximately 700ms. As impressive as that sounds, add the fact that we are searching a database of over 11 million products. This is an area of our system we are continually monitoring and working to improve on.

4.  Don’t try the latest SEO hacks
Googling “SEO tricks” will give you 11,600,000 results and counting. But I give you this warning. If it sounds too good to be true or sounds like something that is not in the interest of the customer, avoid it. If Google haven’t caught onto it yet, your time is limited. And when Google does find it, not only will all your hard work be wasted, you’ll probably be penalised. Just talk to all the people that got burnt by Google’s Panada and Penguin releases.

At Booktopia we believe in avoiding the latest SEO fads, while still being innovative and at the forefront of SEO. Remember, Google will always reward sites that have content that customers like.

5. Remember SEO friendly URLs
This is a well known strategy, but nevertheless important. Make sure your URLs include keywords that are related to the page that is being displayed.
At Booktopia, we use category names, product titles and author names to ensure Googlebots know what content they will find if they follow links to our pages. You can’t rely on third parties (people linking to your site) using “good” keywords in their links.

And that’s it. My top three (five, yes I got carried away) points for SEO when designing your site. This is by no means an exhaustive list of Booktopia’s SEO strategy but five points that I think are crucial when designing your next site.

Do you have more tips to share? (Please no hacks!)

Website Design for SEO, KISS!