As my blog has grown in the last few months I wanted to polish up my branding online.
Branding really is everything and allows you to stand out from the crowd especially when you consider how competitive the internet marketing niche is.
As you can see I use a distinctive blend of blue and pink throughout my brand which is certainly having the desired effect-
Does Having A Good Logo Really Matter?
When I first launched the blog I created the logo myself in Photoshop which looked like this-
Not too shabby considering I suck at Photoshop but it still had a very amateur and unpolished feel to it.
I decided I needed a new logo!
Not only that but I wanted to be able to measure exactly what impact a new logo would have on key metrics such as bounce rate, time on site and pages per visit.
So to find out I got a logo designed and set up a split test!
Finding A Logo Designer
So how do you go about finding a decent logo designer?
Forums
The first thing I did was jump over to some popular internet marketing forums to see who was offering logo design services.
I shortlisted a few I liked and got in touch with them and got a set of concepts created by one of the designers.
The only problem was I didn’t like any of the concepts and after trying to work with a few designers I still couldn’t find one I liked better than the original one I made myself.
Logo Contests
During this time I came across a post on TrafficPlanet that recommended logo design contests.
Essentially you create a contest with a prize fund; people submit entry’s that you can rate and feedback until you select a winner.
At first I looked at 99Designs but the prices were shocking and then I came across 48HoursLogo.com which was much more budget friendly.
So I setup my logo contest – you can check out the actual project page here to see exactly what I asked for.
Within a few days I had loads of logos to choose from!
Not too shabby huh? I actually had a hard time picking the one I liked most!
Choosing The Winner
To choose the winner I created a vote and asked my readers to vote for their favourites!
After all, it’s you guys that will have to look at it all the time 
In the end I settled on this-
I also asked the designer if he would mind doing another version with the character facing the other way.
Which looked like this-
So now I have the original logo, the left facing logo and the right facing logo to split test!
Will it actually make a difference?
Setting Up The Split Test
To setup the split test I used the SES Theme Split Test plugin which is free to use.
This allows you to split test themes and then track custom variables in Google Analytics.
Duplicating The Theme
I took a copy of the original theme, changed the folder name to ‘meleft’ and swapped out the logo for the new one.
Then open up the stylesheet and change the Theme Name to meleft.
Then do exactly the same but this time for the right facing logo called ‘meright‘
So at this point I had 3 themes installed and the only difference between each of them was the logo.
Configuring SES Theme Split Test
This has got to be the easiest plugin in the world to use.
Just choose which themes you want to split test and hit Save Settings.
Setting Up Google Analytics
Last but not least is setting up Google Analytics to track the different themes. Don’t worry this is really easy!
From your main visitor dashboard click on Advanced Segments and then New Custom Segment
Have a look at this screenshot to see how I track people using the meleft theme-
You need to create a new custom segment for each theme you are split testing.
It will take about 24 hours for the new data to start appearing in your analytics account once you have set everything up.
Note: You will see my segment only includes new visitors. Repeat visitors are already engaged with my brand, the trust is already built.
I wanted to measure the ‘first impression impact’ of the logo with people that have never seen the site before.
Split Test Results
I ran the test from February 8th-26th through a total of 5,459 unique users.
I’m just going to let Analytics do the talking here-
click the image above for a full version
To make it a bit easier to digest let’s focus on the 3 important metrics – pages per visit, average visit duration and bounce rate.
| Logo | Pages/Visit | Visit Duration | Bounce Rate |
|---|---|---|---|
| Original | 2.08 | 00:03:43 | 57.82% |
| Left Facing | 2.52 | 00:04:01 | 55.70% |
| Right Facing | 2.67 | 00:04:12 | 54.64% |
Digesting The Results
As you can see both the professionally designed left facing and right facing logos offered better engagement all round.
However what is interesting is how the left facing and right facing logos performed differently.
They are essentially the same logo with the character facing the other way, yet one performed significantly better than the other.
This is why I love split testing 
You can settle most arguments with data!
So Was It Worth It?
For a one time investment in a logo contest I was able to-
- Increase pages per visit by 0.59
- Increase time on site by 29 seconds
- Decrease bounce rate by 3.18%
So this means that every other visitor will now browse 1 extra page than they would have before giving me extra opportunities to convert them.
This also leads us into visitors spending an extra 29 seconds on the site than they would previously!
And that decrease in bounce rate means more people are moving around my blog, sharing my content, clicking on my affiliate links and subscribing!
You can start a logo contest for just $99 but I decided to offer $130 for my contest.
So was the $130 worth it? Hell yeah it was!
All day every day while I’m working, sleeping or eating more new visitors are getting reeled in with my blog than ever before! WINNER!
Wasn’t even hard work either =D
No not Matt Cutts, but I am better at SEO than him :P I make a great living online with SEO and I will teach you 
total
twitter
facebook
google +1
linkedin
High quality tutorials like you've never seen before!
24 Responses
Very interesting! I’ve been looking into different graphic designers for my own branding, but so far there’s been tons of schedule conflicts so no actual roughdrafting has been done. I like the idea of turning it into a contest because it gives a lot more to choose from (assuming a lot apply.) The split test results were also interesting but I wonder if the numbers were too close to really distinguish whether it’s coincidence or actual results.
Reply
Matthew Woodward Reply:
March 13th, 2013 at 4:44 pm
Hi,
Yes the contest was great because you got lots of design minds involved instead of just 1!
The split test ran enough data to be conclusive
Reply
Peter Reply:
March 13th, 2013 at 7:59 pm
You would have to know the standard deviation or variance of the data in order to determine if the results are different. Minitab will do it all for you in seconds. Just need to enter the data.
Reply
Matthew Woodward Reply:
March 14th, 2013 at 9:33 am
Feel free to drop me a mail with that data and I’ll update the post
Reply
Now who would have thought that a logo could have that much of an impact?
Great stuff Matthew. I am definitely going to put that on my things to do list.
Reply
Matthew Woodward Reply:
March 13th, 2013 at 4:45 pm
Well first impressions count and you have a split second to make a great first impression!
The logo is one of the most prominent design features and the first thing that hits the eye!
Reply
Ryan @ Trusted Site Seller Reply:
March 13th, 2013 at 10:15 pm
Nice case study on how a log change can make a big difference. Also, I had been looking at 99designs and always thought their pricing was pretty crazy too. Thanks for the link to an alternative.
Another interesting split test would be to change out your picture in the “Welcome to my Blog” box on the upper right of the page. My first thoughts when seeing that picture is you look like a punk and your drinking a beer. After reading your content for a few months I’m pretty sure the first isn’t true and not sure about the second, but in my opinion that picture had more influence on my first impressions of your brand than your logo did.
Just food for thought, but it might make for an interesting followup to this post.
Reply
Matthew Woodward Reply:
March 14th, 2013 at 9:35 am
Hi Ryan,
Funny you should mention that – I have had similar feedback from others but hadn’t thought about split testing it. I was going to get a better image done on a red background to improve CTR from SERPS as well.
Next I’m going to test sidebar position but once thats locked in, this is next – thanks for the tip!
Reply
great stuff as usual Matt specially the case studies and tutorials. graphics indeed play a important role it isn’t just limited to logo and this post has given me an idea that i’ll be trying using facebook ads =)
Reply
Matthew Woodward Reply:
March 14th, 2013 at 9:29 am
Oooo Facebook ads – do tell
Reply
You are truly the master of the split test – absolutely love it and wish I had quite the scientific dedication you have Matthew! I will definitely be blogging about this and split testing WordPress in general – great tip on the plugin.
Reply
Matthew Woodward Reply:
March 14th, 2013 at 9:30 am
Hahaha I would say I’m the master, this is low level testing really. I much prefer multivariate stuff but its a reeeeet pain with WordPress.
The next split test is left aligned or right aligned sidebar
Reply
Jeremy Chow Reply:
March 20th, 2013 at 4:26 am
The F-shaped eye tracking monkeys might argue that they have that already figured out but I love a debate and hope you come out with a completely different answer. Controversy always smells good in the morning.
Reply
Matthew Woodward Reply:
March 20th, 2013 at 9:05 am
Hi,
Well I’ve actually got a bit of experience in that field and I can already tell you what the results are going to be
You know as well 
Reply
well done and thank you for the walk through of how to do all this.
What is surprising is that designers so often create images of people looking away from the midline of the page and not towards it.
Alex
Reply
Matthew Woodward Reply:
March 14th, 2013 at 9:31 am
Thats an interesting observation – I hadn’t thought that you could use the direction of the eye to lead people
Reply
Absolutely stellar post, I love it… the idea of a logo contest is great. Your results just goes to show how much difference design can make to engagement.
I used to work for a company called Interwoven and they had a Multivariate Testing product called Optimost, which simply tested a rearrangement of multiple design elements on registration/conversion pages (such as text boxes, buttons, images etc) and found out what design converted best. Basically the next step up from A/B testing. The customers were mostly big ecommerce brands and some were able to get a decent percentage increase in conversions or registrations using MVT (which meant big increases in revenue for them – I think Love Film experienced a multimillion £ increase in revenue using MVT back in 2008). Interestingly Google had a free MVT product called Website Optimiser which did something similar but it was discontinued in 2012 and replaced by a function in Google Analytics called Experiments, which can now be found under the Content menu. …so Matt maybe a nice idea for a new post would be doing some Registration/Conversion page optimisation testing using Experiments in GA?
Reply
Matthew Woodward Reply:
March 14th, 2013 at 9:32 am
Hi,
Experiments suck with GA – the older website optimizer was much more dynamic and allowed for multi variate testing where as now it is just Page A and Page B, which actually makes split testing your logo across the entire site impossible.
Still not found a decent solution that plays nice with WordPress for the multivariate voodoo!
Reply
Great post, thanks. The info on split testing was great – easier than I thought too.
Incidentally, I used to work (in telesales, briefly) for a logo design company. For those that happen to be looking, I’d have no hesitation in recommending them (if you don’t mind Matthew). TheLogoCompany.net*. Cheers.
*I am not an affiliate and have no incentive, financial or otherwise.
Reply
Matthew Woodward Reply:
March 14th, 2013 at 9:33 am
Hi,
Yeah its easy when you know how
Reply
Love your reporting style, easy to follow steps that always make the task you’re describing seem eminently doable.
On the to-do list!
Reply
Matthew Woodward Reply:
March 14th, 2013 at 9:37 am
Well its easy when you know how – took me a bit of trial and error to come up with a decent way to do the actual testing =D
Reply
Very interesting, indeed!
And those logos are all so very sharp looking. They were designed by some real professionals.
I’m developing this comedy blog ( http://www.ssstlg.com ) where I’m going to offer my services as a writer (as I keep getting contacted out of the blue from this particular site because my craziest writing lives there), and I’m actually proud of the terrible job I did with that logo – it’s the best I’ve done before anyway. I think I’ll let the professionals handle instead…
Reply
Matthew Woodward Reply:
March 14th, 2013 at 9:38 am
Hi Robert,
Yeah I was impressed with not only the quality of the entrys but the huge variety in style.
I’ve got to say, your logo does hurt
Reply