Execution is the key #7: the magic funnel

Internet is worse than tv. You can zap faster and for many more reasons. Once you get a user into your new website for the first time, you have to give a pretty good reason to get him use your service and eventually come back. One of the biggest challenge is to manage to convince him/her to try the service with no friction (one of them being the registration process). This is in my view one of the hardest part of the execution and there is no real recipe on how this can be solved since it merely depends on the nature of the service. But the end result should be that the user will get close to immediate understanding and reward by trying your new site.

There are many tactics to remove the friction between the first visit and the actual usage. Here are a few

• Site Tour: a screencast to explain what you'll get
• Demo account: so you do not need to register
• Freemium approach: don't pay now, pay later (animoto, Flickr)
• taste me first: give an inch of your service without registration and provide more services by registering (Youtube)
• No registration at all: all is based on a open cookie based service (like stat service whos.among.us i use on this blog at the bottom right)
• Simulation approach: if you were a user this is what you would get. You enter some information and there you go (Pandora)

All of those tactics could be resumed by what i call the magic funnel, which is a path the user is following and that drives him to not only try the service but also get instant reward and understanding by doing so. I think the best option is to give you an exemple. Lijit

 

Lijit provides a customized search service around your universe: it allows you to aggregate your blog and other online services (bookmarks, photos,..) and will provide a search box that you will be able to embed anywhere you want. In addition you'll get TONES of useful services like search stats, popular search tag (you can see that on my blog on the right)

Lijit has a KILLER magic funnel. The way they drive you in the service is seamless and simple. With very little effort (meaning in that case providing the URL of my blog) i got the service set up with already very useful info and tools. Even better, somehow Lijit managed to identify my blog open from my browser i did not even have to write the URL. Here are examples of instant benefits: Lijit identified from my URL the blogs presents in my Blog Roll on the right and asked if i wanted to include them in my search. Lijit provided me a widget that i could install in one click on my blog. Lijit provided me with the option to add the search to my Firefox search tool.

The user interface in addition is clear, simple, fast and provides a kind of WOW effect, which helps of course.  In a word i have the feeling that Lijit was obsessed with the question "How can i make the first time experience as simple of possible" and tried to resolve every single step from first visit to registration. They use a mix of TasteMeFirst and simulation approach in a progressive way

Finding the magic funnel is a very hard task and is the result of an iterative process based on analytics and optimizing the user experience. you won't get it first. This is normal. but you should focus on it real hard.

PS: another good example of magic funnel can be found at Jiwamusic (althought i hate the design) but there is a zero friction user experience between trying the service and registering (thank to ruby on rails)

Technorati tags: lijit, funnel, user experience, internet, execution, jiwamusic

Comments

[David Kadouch]

The posts I like most are the ones that provide a well-thought analysis of real problems like this one (on this blog I also enjoy the refreshing posts on Jazz music).

[Stan James]

How did you know that we are so obsessed here at Lijit?! You are right that we think hard about every part of the signup process. Our mantra has been that there is no reason to ask the user to do something or to enter information if there is anyway that we could do it for them.

We're proud, but want to go further. Users should be able to test out their search engine before creating the account, for example.

And Ouriel, I think maybe you met my friend Fabian at the 2006 NextWeb conference in Amsterdam? Back then Lijit was known as Outfoxed and we had a *very* complicated process--not a funnel at all! Since then we've been working hard to make it easy. Glad that you appreciate it.

-Stan
Founder & CTO, Lijit Networks