This post has moved. An updated version is on my Medium here: Every Step Costs You 20% of Users
Tuesday, October 23, 2012
Subscribe to:
Post Comments (Atom)
About Me
Gabor Cselle
San Francisco, CA
San Francisco, CA
-
I work at Google, where I'm a Partner at Area 120. Previously, I worked at Twitter on trends, the logged-out homepage, and on MoPub. Twitter acquired our startup Namo Media in June 2014. Before Namo, I was a Product Manager at Google working on Google Now, Android and Gmail. I started reMail, a mobile email startup which was acquired by Google in February 2010.
Before reMail, I was the VP Engineering at Xobni, where we invented a popular plugin for Outlook, and a Software Engineer at Google. I have an MS degree in Computer Science from ETH Zurich.
Gabor Cselle Bio
13 comments:
stockpile for the appocalypse ;-)
That's why I like signup/onboarding flows that use gradual engagement. Show them the utility of the app/guide them through their first action before you actually have them sign up. We do this with Picplum -- drag photos in first before signing up.
Wil Wright of the Spore game says he strives to make users have their first "win" within 5 seconds. I think that definitely applies to any new user experience as well.
I was about to make the app-ocalypse joke too :-)
A typical consumer web service signup is 2-4 steps. I've done a few of these, but have never once seen a 20% drop-off rate per step. I can't imagine that participants in this discussion have experienced a cumulative 80% drop-off in the funnel?!
So, while I like this discussion, perhaps the headline is misleading..
It's implicitly true that giving more up-front and getting users up and running as fast as possible is beneficial... however there is tension between fulfilling customer acquisition goals and managing your numbers for very good reason.
One tip I can share is to get their email address (or FB/Twitter account) first. The value of recovering from a failed funnel signup goes unstated. You should have a list of abandons on Mailchimp (or wherever) that you blast periodically, to try to recover from aborted signups.
-Ian.
Ian - thanks for your comment.
As Paul said, the challenge becomes locking people in as quickly as possible - even before signup.
I guess my main point, aside from the specific numbers, is that mobile app users are incredibly fickle. You can't expect them to step through layers of signup and registration. Apps don't have very much time and space to tell their story before they have to deliver satisfaction to the user. That wil remain a big advantage for the web for many years to come.
There are just so many apps out there that users download apps to be able to find them later, though they aren't ready to actually play with it yet.
(That might be about 15% of the downloaders, but a third of them actually get back to opening it.)
The more complex but potentially useful the app is, the more likely it will be downloaded but not run.
Apps that are designed to help organize the clutter ought to have a higher rate of unopened downloads than yours does.
We have a registration process on one of our mobile coupon sites. We try to collect the user's phone number on a splash page before entering the site in order to do SMS messaging. We've struggled with this...is it better to collect the phone number (lower opt-in rates) to do SMS messaging (higher open rates) or have them opt-in with Facebook/Linked-in (higher opt-in rates) and do email blast (lower open rates)? What a dilemma!
How about this: skip signup altogether. Distribute a GUID with the app, so you can take metrics on any given installation. You don't NEED anyone's name or email address.
When apps ask me to give them personally identifiable info prior to providing me any benefit, I immediately quit and never look back.
Post a Comment