determining reasons for user dropout

11 Dec 2008 - 6:12pm
6 years ago
6 replies
632 reads
Michael Stiso
2006

Hello. I'm curious about methods for finding out why subscribers to a
site stop visiting that site. For example, imagine that I have some
social networking site, and I notice that some subscribers (some of
them new members, some old) simply stop visiting it. They don't cancel
their accounts, they just stop showing up.

How might I discover their reasons for dropping out? I imagine that
surveys can be automatically sent to people who don't log in for a
specified period of time, but I'm wondering if others may have
developed user research methods that would be less potentially
irritating to the users.

Thanks for any thoughts on the matter.

Mike

--------------------
Michael Stiso
HCI Research Scientist
SINTEF ICT
Oslo, Norway
http://mikestiso.com/

Want to help with a survey on design patterns? Please?
Check out this link, if you can spare 5-10 minutes. Thanks!

http://preview.tinyurl.com/63f9gq

Comments

14 Dec 2008 - 2:20pm
Michael Stiso
2006

Thanks, Diana, and I agree with what you say. I'm more curious,
though, about how to determine the reasons why people drop out, other
than via surveys/emails.

I'm not working for a place where this is an issue, really, it was
more just a research interest on my part -- how to get information
from users of a subscription-type site when those users disappear.

Thanks!
Mike

> There are lots of good reasons users stop coming, so context matters:
>
> Were they ever regular users, or were they just checking you out (Second
> Life, which has a huge fall off after registration)?
> Have they forgotten their password and would log on if you reset it for
> them?
> Do they just need a gentle reminder/invite and coupon or tour of what's new
> on your site? Or do they have a reason for not wanting to log on--for
> example, despite weekly e-mails from my brokerage firm, I simply do not want
> to face how much money my retirement account has lost since my last log on.

14 Dec 2008 - 6:04pm
Jared M. Spool
2003

On Dec 11, 2008, at 7:12 PM, Michael Stiso wrote:

> How might I discover their reasons for dropping out?

I know this might sound radical, but have you considered just talking
to them?

Picking a handful who have not logged in for a while and sending them
a personal email, asking if you can have a phone conversation.

I bet some will take you up on it and it'll prove to be a very helpful
way to learn about your users.

Hope that's helpful.

Jared

Jared M. Spool
User Interface Engineering
510 Turnpike St., Suite 102, North Andover, MA 01845
e: jspool at uie.com p: +1 978 327 5561
http://uie.com Blog: http://uie.com/brainsparks Twitter: jmspool

16 Dec 2008 - 12:45am
Atul Thanvi
2008

Ya i agree with Jared. Telephonic conversation will be the best option
to go ahead with this research.

Before few weeks four of my friends on a community site stopped
visiting the site and i was surprised why they did so, as they were
earlier addicted to this.

The only reason they told me that more of business ads and networking
on the social community site.

It also depends on user to user as some user like advanced features
on the community site for better networking environment.

Surely this subject needs a research......

Thanks,
Atul

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=36413

14 Dec 2008 - 2:02pm
Diana Wynne
2008

There are lots of good reasons users stop coming, so context matters:

- Were they ever regular users, or were they just checking you out
(Second Life, which has a huge fall off after registration)?
- Have they forgotten their password and would log on if you reset it for
them?
- Do they just need a gentle reminder/invite and coupon or tour of what's
new on your site? Or do they have a reason for not wanting to log on--for
example, despite weekly e-mails from my brokerage firm, I simply do not want
to face how much money my retirement account has lost since my last log on.

The other reason of course is they don't need what's offered by your
company. Reminders and incentives are unlikely to affect those people, so
good to separate them out if possible, and perhaps follow up with them
separately. If you were charging a monthly fee, they would cancel their
accounts rather than drop out.

I'd recommend testing responses to various approaches rather than sending a
survey, which is unlikely to engage anyone except the small demographic of
people who like to fill out surveys (most of whom are members of this list
:)

A friendly "we haven't heard from you in a while" e-mail accompanied by an
incentive can revitalize lost visitors/customers, at least the ones who saw
value in your service in the first place.

Finally if you do follow up (and you should), please offer an easy
unsubscribe. There's nothing I hate more than being spammed by some site I
logged onto once three years ago that wants me to remember a user name and
password in order to stop getting mail from them.

Diana

On Thu, Dec 11, 2008 at 4:12 PM, Michael Stiso <mikestiso at gmail.com> wrote:

> Hello. I'm curious about methods for finding out why subscribers to a
> site stop visiting that site. For example, imagine that I have some
> social networking site, and I notice that some subscribers (some of
> them new members, some old) simply stop visiting it. They don't cancel
> their accounts, they just stop showing up.
>
> How might I discover their reasons for dropping out? I imagine that
> surveys can be automatically sent to people who don't log in for a
> specified period of time, but I'm wondering if others may have
> developed user research methods that would be less potentially
> irritating to the users.
>
> Thanks for any thoughts on the matter.
>
> Mike
>
>
> --------------------
> Michael Stiso
> HCI Research Scientist
> SINTEF ICT
> Oslo, Norway
> http://mikestiso.com/
>
>
> Want to help with a survey on design patterns? Please?
> Check out this link, if you can spare 5-10 minutes. Thanks!
>
> http://preview.tinyurl.com/63f9gq
> ________________________________________________________________
> Welcome to the Interaction Design Association (IxDA)!
> To post to this list ....... discuss at ixda.org
> Unsubscribe ................ http://www.ixda.org/unsubscribe
> List Guidelines ............ http://www.ixda.org/guidelines
> List Help .................. http://www.ixda.org/help
>

18 Dec 2008 - 12:46pm
jaketrimble
2008

If your site is configured for it, look at their activity before they
"dropped off". You can also compare their actions to other active
users.

You should be able to determine quite a bit about why a user might
have stopped visiting your site from their action history.

To understand the why, you need to already know the what, when, and
where. This means configuring your application to track clicks, and
other actions that will help determine a whole plethora of why's.

I also agree with Jared and Diana. You should use this method
especially if you have no action history configured in your
application.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=36413

18 Dec 2008 - 1:48pm
Torey Maerz
2008

I agree that the best information will come from an interview of some
users. To help you identify trends and potentially specific groups
of people to interview you should consider creating a timeline of
specific events a user might encounter. For example registration,
first post, number of friends, number of replies etc. Trend this
timeline along with drop off data. You should be able to see around
what events a user is more or less engaged. For example, are they
more or less likley to drop off depending on the number of friends or
length of time from registration to a specific amount of contributions
etc. There will be a lot you will learn from your users with this
information and you could also use to to determine different groups
of people to talk to.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Posted from the new ixda.org
http://www.ixda.org/discuss?post=36413

Syndicate content Get the feed