Posted in: Archive
So you’ve managed to actual generate traffic to your website. You’ve got a steady flow of visitors, merrily munching on your content, and life is good. Not only are the consuming what you write, but they are interacting with it as well, sharing it on Twitter, Liking it on Facebook, Bookmarking on Delicious. etc. etc.
Where do you go from here?
It’s all well and good stating that your website has traffic. You can boast about the numbers until the cows come home, but what really matters are conversions. Give me a website with one hundred visitors a day that converts 70% of visitors – over a website with one thousand visits a day that converts merely 7%. There’s certainly less work to do with the first scenario, to drive more business through the door – as the hard part is more or less done.
Thankfully we are spoiled online with tools that can help us increase the conversion rate of our website, and indeed understanding the impact that design has on conversions. The most common way to garner this information is to perform split testing (sometimes known as A/B testing), or multivariate testing. E-commerce sites in particular, with established end goals can benefit greatly from this methodology, rather than just throwing extra traffic at the site (which in some cases actually costs money in additional server hosting fees). The following review post takes a look of some of the online offerings we have at our fingertips in 2010 for increasing conversions, and improving the overall usability of your website.
Five Second Test
Five second test is a great usability tool, created by the two webby geniuses at Angrymonkey’s. 2010 will be the year of web based conversion testing, and five second test allows designers to get a feel for opinions quickly and easily from other designers and web users. You can perform either click tests (which require users to click on the areas which stand out), or memory tests (where users are given 5 seconds and asked what they remembered about the design).
The beauty of five second test lies in the fact that all you have to do is upload your design, and away you go. There’s no scripts to mess with, and you can begin to get feedback straight away once its set up.
Both of these tests can focus designers on the immediate goals of the site – which can’t be a bad thing. If you aren’t interested in actually performing a task, you can choose to help other designers out by taking a random test.
Feedback Army is a paid for service that again connects designers with an audience to receive feedback on certain areas of a site design. Think of it as a paid review. Under the bonnet Feedback Army uses Amazon’s Mechanical Turk (God I can’t wait until this service hits the U.K. at present its a US only offering). Mechanical Turk essentially is a programmatic interface to humans performing tasks. Requests are made and received via an API, and fed back to your web application. Genius. Anyway, that’s what you are paying for within Feedback Army – make of that what you will as far as the actual feedback you receive. The demographic of the audience is going to be very mixed. For the price however, it is better than no testing at all.
Usertesting is a higher end offering than its competitors. It allows an exact demographic to be chosen by the tester, and the cost is just $29 per participant. Every participant in your test records their actions and voice for later playback. Screen recording software is used to let you see exactly what is happening as they browse your site.
One of the main benefits of this is that it allows you to take a step back. Often software developers and web designers are too familiar or “close” to a project to really get a feel for problems.
Visual Website Optimiser
Visual Website Optimiser is a killer app. I was introduced via an invite from Dave – and is exactly what the web needs. Whilst Google’s A/B optimiser tool takes alot of thinking, reading and general too’ing and fro’ing – Visual website optimiser is the exact opposite. It’s main strength lies in the fact that very little changes are needed to your actual website code, and its very simple to use. One snippet in the header, one snippet in the footer. Done.
There are two main options for testing. The easy to use “visual” point and click tool, and the slightly more complex split test option. With the visual optimiser everything after this point can be setup and changed via the web application itself.
Call to actions in many web applications are in textual format, and visual web optimiser allows you to play with that text. Want to find out whether the words “signup” or “subscribe” result in more RSS subscribers? No problem. Simply setup your variations, click a few buttons and away you go. If on the other hand you use images and want to test two versions, the split tester is what you need.
Results are presented in a clear and easy to read fashion letting you see your conversion rate for each variation over time. Form subscribes as well as landing page visits are supported (after a suggestion from yours truly – and the vwo team turned it around the next day!) – and I believe that they are fighting the good fight in creating one of *THE* best web based testing tool on the web. One to watch in 2010 for sure. I’ve a few invite codes for this one – so if you want to leave your email address in the comments I’ll sort you out.
Kissmetrics are one of the few analytics tools online that concentrate on people over numbers. They use the notion of events that people perform to construct funnel reports, allowing you to see where people drop out – all in real time.
Some of the main benefits of the product lie directly in the API – which you tie each visitor to an multiple interactions on your site. Even visitors who block cookies can be identified, assuming you can provide a unique identifier such as an email address or a GUID. This facilitates tracking over the duration of a session, and only needs to be set once at the start of a request.
Kissmetrics have a good track record – and have launched awesome web based tools before in the past – tools such as Survey.IO , ProductPlanner and ShareFeed – all these apps are great examples of the innovation they obviously thrive on.
MixPanel describes itself as a funnel analysis tool. Its complete free, provided you are only tracking small numbers of events, and would be a perfect tool to sit alongside Google Analytics to get a feel for what is really going on on your site. If your traffic is more substantial the pricing reflects that, but at $25 for 100,000 events it will still remain affordable for many. Some of the larger clients they have on their books are Posterous – the microblogging platform , and Uservoice – and we all know how easy their signup process is.
Mixpanel track real-time user engagement in a similar way to KissMetrics – although their API isn’t quite as simple to implement.
I could browse this site for hours. Ab tests allows testers to share their results with the world, and explain exactly what they have done to change the conversions. Whilst it isn’t a “tool” in the traditional sense of the word, learning from other people’s mistakes is always a win win situation. If you have performed tests on web sites that you own or manage, ab tests also lets you link back to the site in question.
Abtests is the brainchild of the same team behind KissMetrics (see earlier) – and it makes sense for conversion tool companies to be thinking in these veins, and really proving their worth.