Re: Step n of m

Hi Hal,

I used to do a lot of testing of online insurance quote forms.  Users
would get quite annoyed when the numbers of steps on the progress
indicator were not perceived to be an accurate reflection of actual
screens, and often thought this was a deliberate and sneaky ploy on
the part of the insurance company to dupe them into getting a quote by
pretending the process was quicker than it actually was.

Trying to make the process appear shorter and therefore more appealing
was counterproductive: if the system had added in 'extra' screens
once, users felt it might do it again, leading many to fear that the
process must in fact be prohibitively long and therefore increasing
the dropout rate.

In this context it was often impossible to predict accurately just how
many steps a user would have to go through, as some users would have
to enter more information (and go through more steps) than others,
depending on certain conditions.  For example, users with endorsements
on their driving licences would have to provide codes and dates of
each offence before being given a car insurance quote.

I always recommended being as accurate as possible in matching the
apparent steps to the actual number of screens, making sure it was
always clear why all steps were necessary and exactly what was going
to happen next, and explaining the need for any deviations from the
'normal' process which might add in extra screens.

NB: This was a few years ago when adding extra questions always meant
adding extra screens.  If you're able to use a more interactive (e.g.
Ajax) interface, you could now dodge the issue by adding extra
questions inline into the form as and when required.

This was a different context so YMMV (insurance customers can bail out
and phone, or  get the same product elsewhere).

However, from my experience I would recommend always showing actual
screens completed/remaining.  If it helps to clarify the process then
show how these map to logical steps too, but make sure you show what
this means in number of actual screens.

HTH,
Claire

On Wed, Mar 11, 2009 at 1:29 PM, Hal Shubin <[email protected]> wrote:
> You're designing a wizard in a Web app. Choices:
>
> 1. Each step is one page and each page is one step. Benefit: Clarity --
> click NEXT and get to the next step.
>
> 2. Make the steps be at a higher level, and each step could be spread out
> over two pages. For example, the "Setup" step includes logging in to your
> account, downloading software and installing it -- maybe it's just neater
> (conceptually and/or visually) to break that up. Benefits: Simpler pages,
> and it looks like fewer steps when you see "Step n of m".
>
> I just ran a study where one person was vocal about wanting one step/one
> page. I don't think it mattered much to the others. Other observations?
>
> thanks                          -- hs
>
> . . . . . . . . . . . . . . . . . . . . . .
> Hal Shubin
> Interaction Design, Inc.
> 617 489 6595
> www.user.com
>
>   --------------------------------------------------------------
>          Tip of the Day: Postings must be in plain text
>    CHI-WEB: www.sigchi.org/web POSTINGS: mailto:[email protected]
>             MODERATORS: mailto:[email protected]
>      SUBSCRIPTION CHANGES & FAQ:  www.sigchi.org/web/faq.html
>   --------------------------------------------------------------
>
>

    --------------------------------------------------------------
           Tip of the Day: Postings must be in plain text
     CHI-WEB: www.sigchi.org/web POSTINGS: mailto:[email protected]
              MODERATORS: mailto:[email protected]
       SUBSCRIPTION CHANGES & FAQ:  www.sigchi.org/web/faq.html
    --------------------------------------------------------------