Re: Writing for on-line

[David Knopf <david -at- KNOPF -dot- COM>]

Jill,

Jill Burgchardt wrote:

> Thanks for a good post that made me think about how I use browse
> sequences.  Just a few things I wondered about when I read it, though.
>
> You wrote:
>
> (a)  implementing browse sequences well (i.e., a separate path for each
> group of related topics) is very time-consuming both in development and
> maintenance;
>
> I can't imagine implementing browse sequences for each group of related
> topics.  The primary help system I work on consists of over 2500 topics.
> (That'd be an enormous undertaking!) But, I do use three browse sequences
> that link 100 or so topics.  One set links application overview topics,
> one set links help overview topics, and one set links procedures that
> must be completed in sequence.  The first two are basic things a new user
> might want to go through in sequence to become familiar with the system.
> The third is in a browse sequence simply because sequence matters on
> those procedures.
>
> I'm curious, why do you think browse sequences are implemented well when
> one exists for each group of related topics? I find judicious use of
> browse sequences--i.e., where sequence matters--to work best for our
> users. If I were crazy enough to sequence 2500 topics, what would I gain?

Nothing, in my opinion. I didn't mean that the ONLY good way to implement
browse sequences was to create for each group of topics, but rather is this
is ONE good way to do it. I vastly prefer the judicious use in limited
subsets of the help system. What I loathe is the implementation of a single
browse sequence from topic 1 through topic n.

> I find the usability tests interesting, but wonder how much the results
> are a consequence of poorly designed systems. Many users don't like help
> systems at all, because they've encountered too many which explain the
> obvious. Browse sequences require extremely rigorous audience analysis
> which I suspect few people do, but when they include the right topics
> where the targeted user will find them, I think they're very helpful.

The tests I refer to involved good help systems, including some from
Microsoft and Lotus. If you could track the numbers of times your users
actually click your browse buttons, I think you'd be surprised to see how low
the numbers are. I'd bet most don't even see them since you've used them in
only 4% of your topics. (On the other hand, I could well be wrong about this;
every user group has its own characteristics and I know nothing about
yours!)

> I'm concerned if, as you say, "many users may not, in fact, know what the
> browse buttons are for" that they'll miss valuable direction in the
> system I've developed. Where sequence matters, do the usability tests
> suggest an alternative? How are you solving that dilemna?

Several questions here. First, usability tests generally don't suggest
alternatives since they can only test what is actually there. That said, I
think most users would more rapidly recognize and understand authorable
buttons in the non-scrolling or client regions labeled Next and Previous. I
have not tested this hypothesis, though, and it would obviously be still more
work to implement this. As to how we solve the dilemma here, I don't see it
as much of dilemma. We scarcely ever use browse sequences, and when we do we
use them much as you do.

-- David Knopf

------------------------------------------------------------
David Knopf
Knopf Online
Tel: 415-820-2356
E-mail: david -at- knopf -dot- com
Web: http://www.knopf.com/

------------------------------------------------------------
Writing * WinHelp * Web Sites * Training * Consulting
RoboHELP Certified Trainer * RoboHELP Certified Consultant
------------------------------------------------------------