Trip report: 47th STC Annual Conference (long)
SteveFJong -at- AOL -dot- COM
Mon, 5 Jun 2000 16:08:09 -0700
In response to a request, I am posting my trip report. This report covers
"Using ISO Standards to Evaluate Quality and Usability of Ongoing Product
"Measuring the Quality of Technical Publications"
"Information Process Maturity: Going from Oblivious to Organized"
"You Be the Judge"
"STC's Art, Online, and Pubs Competitions"
"A World View of Good Technical Publications"
"Design Surveys to Extract Usability Information from Tech Support Calls"
"How to Ask for What You Need So That People Want to Say Yes"
"Creative Ways to Reward Employees"
These reflect my interests and my company's needs. Hopefully, if others
post their trip reports, a more representative picture of the conference
will emerge. I apologize for its length. (It was too long for AOL mail; I
sure hope I haven't posted it three times!)
SESSION: "Using ISO Standards to Evaluate Quality and Usability of Ongoing
Product Development Processes": David Dick, editor of the Usability SIG
David, the most dynamic speaker I've ever seen at an STC conference,
described how groups in a typical company talk to each other, but enjoy
limited access to customers, during product development (this sounds
familiar...!). Without customer feedback, the development cycle is liable
to spiral off in the wrong direction.
Every group in a company needs coordinated customer contact?in short,
feedback on usability. But justifying the perceived expense of usability
studies can be difficult (even though the cost of providing an unusable
product is 100%).
For companies that are (or plan to be) ISO 9000 compliant, there is now a
direct justification. (Quoting the ISO Website, "The ISO 9000 standards are
a set of international quality management standards and guidelines. Since
their initial publication in 1987, they have earned a global reputation as
the basis for establishing Quality Management Systems (QMS).") The Year
2000 revisions to the standards add the requirement that products or
services measurably conform to customer requirements. Determining
conformance to requirements falls squarely into the arena of usability (and
quality metrics!). He advocated task analysis, which involves understanding
both goals and measurements.
SESSION: "Measuring the Quality of Technical Publications": Dr. JoAnn T.
Hackos, Comtech; Julie Bradbury, Cadence; Connie Lamansky, Cadence
I eagerly anticipated this presentation. For Dr. Hackos, the place banged
out?over 200 people. Her newest book, Quality and Technical Communication,
will be published by the STC Press next year.
In her opening remarks she said she knew of two or three cases where
documentation actually saved companies. (In one case, installation support
calls ate up the entire profit margin on a high-volume consumer software
product. After writing professional documentation, calls for installation
Whenever Dr. Hackos speaks, she drops quality metrics like pearls, for
which the audience scrambles. Process milestones are quality gates: don't
miss the opportunity to measure quality. What fraction of the development
cycle do documents spend in review? How many errors are found, and how many
corrected? What percentage of reviewers actually return comments, and what
errors did they miss?
She also suggested a survey methodology. Rather than hand-pick users as is
often done, the statistically valid way to do it is to select, based on
user role, a random subset of the user community, and then aggressively
pursue them. If you can't get a response from someone, randomly select
someone else, but be sure to get the requisite number of responses! She
said that faxed questions and answers works well, if there are fewer than
10 questions that take less than 20 minutes to answer. She suggests asking
if users even have the current documents, and how long it takes them to
She plans to put her slides for this presentation up on the Web at
SESSION: "Information Process Maturity: Going from Oblivious to Organized":
Dr. JoAnn Hackos, Comtech; Michel Lanque, Alcatel; Lois Renaud, Alcatel
I got so much from Dr. Hackos's session that I changed my plans and
attended this one as well. Dr. Hackos adapted Watts Humphreys's Software
Process Maturity Model and created the Information Process Maturity Model,
which ranges from Level 0 ("Oblivious") to Level 5 ("Optimizing"). She
listed these characteristics of immature and mature organizations:
Chaotic, reactionary, crisis-oriented Significant management presence
Work done by individual "heros" Roles and responsibilities well defined
Work stops when times runs out Budgets and schedules are realistic
Quality compromised by unrealistic budgets/schedules Organizational
commitment to process
Significant differences in quality Continuous monitoring of/improvement in
Organization is unpredictable Organization is predictable
What moves a writing organization up the scale?
Level 0 to Level 1: Hiring a professional technical writer
Level 1 to Level 2: Customer complaints about documentation, translation
needs, grass-roots efforts by writers, or hiring/creating a doc manager
Level 2 to Level 3: Written policies and procedures, and the commitment to
follow them all the time
There's also a road back down, which at Level 3 is paved with creeping
bureaucracy and stifled innovation.
The Alcatel speakers gave a hair-raising case study of bootstrapping their
division from Level 0 to Level 3 in less than a year. They kept extensive
stats on elapsed project time. Why? On one project, the doc costs were way
out of whack. The fix was to replace the development manager (!).
Dr. Hackos plans to put her slides for this presentation up on the Web at
SESSION: "You Be the Judge"
I got in on the tail end of this session, led by Ken Ries, the Pubs
manager of this year's ITPC. For this year's competition, there were 68
online entries from 38 chapters. He said that the form, which we used as a
basis for our overhaul of the Boston form, is being reworked a little more
for next year.
Chip Jones, the eDoc manager, gave a verbal review of the level of
quality of ITPC winners, which bears repeating:
Distinguished: No flaws
Excellence: No major flaws; can have minor flaws
Merit: Can have one major flaw
Entries are not judged against each other, as we stress; but Chip
continued by saying, "so therefore there can be multiple winners." I would
interpret his remark as meaning it's OK to judge one book against another,
so long as you allow the possibility of more than one winner in each
category--a different spin. Commenting on judges, Chip said, "If you can't
be constructive, we don't want you." In his view, judges must have
Chip showed an eDoc entry and led the audience through an evaluation.
I noticed the same tendency to trash the example we've seen here. Before
people reached the point of declaring the entry shouldn't win anything,
Chip said it had actually won an award of Distinction (but no one gasped).
SESSION: "STC's Art, Online, and Pubs Competitions"
This session was presented by Ken Ries (ITPC Pubs), Chip Jones (eDoc),
Ann Blankinship (assistant to the STC president for competitions), Tom
Barnett (Art), and Hary Janos-Bottka. It was frustrating at times, because
presenters flashed slides containing statistics, which they thought no one
cared about, but I did.
For 1999-2000, the ITPC reorganized the categories: three software
categories into one, hardware into computer and non-computer hardware. They
clarified policies about credit, putting entries in multiple chapters
(prohibited), and entering illustrations from a document in the Art
competition (allowed). Next year, they will allow photocopies of articles
in addition to tear sheets.
Chip, the ITPC eDoc manager, spoke about ITPC judge training. He
recruits from the pool of previous winners, from the ranks of nationally
known speakers, from local competitions managers, and from others
recommended by regional directors. ITPC judges must pass three or more of
the seven listed criteria. Chip wants qualified judges, not just "warm
bodies." Judges must be interested in constructive criticism, and he's
willing to turn over a third of them each year. In his home chapter,
Atlanta, the competition is the #2 money-maker, but to maintain the
standards of the competition he limits entries rather than overburden
judges! He also mentioned a disqualifying flaw in an eDoc entry: if it
contains a virus.
Ken gave a statistical breakdown, which I didn't fully capture: 173
entries (up from 132 last year), 24 Distinguished (14%), 35 Excellence
(20%), 60 Merit (35%), 54 No Award (31%). He said over the past 12 years,
between 30 and 35 percent of ITPC entries have won no awards. (Remember,
these are all chapter-level Distinguished award winners.) The trend is
toward more PDF entries (are they books or eDocs? pick one, probably
books); more HTML Help but still mainly WinHelp; more Director and Flash
demos; and Web sites (16% of the total). Web sites are problematic because
they can change between entry and judging, and because submitting them on
CD-ROM can break links. There were 36 Pubs judges (same as last year), 27
who gathered in Atlanta and 9 teleconference judges. (I was a
teleconference judge.) There were 16 entries per judge (historically it's
been 11-12). For next year, he wants to provide a sample statement of
purpose (which is crucial at the International level) and samples of good
and not-so-good judges' comments. He also would like to see the evaluation
forms and all administrative paperwork on the Web (even entry-fee payment
by secure shopping cart), and he thinks there should be a PERT chart of
competition milestones. However, he thinks consensus judging still needs to
be face-to-face. Ken was asked how to prevent disputes by a member whose
chapter competition saw judges walk out after arguments; he said the
Atlanta chapter uses a matrix of major and minor flaws. Does he blacklist
judges? In one chapter, yes they do; he advises recruiting as many judges
as possible and holding some in reserve. Must team leaders attend training?
Yes, said Bonnie Graham of SoCal; she disqualifies judges who can't or
won't attend training (although I think she referred to lead judges, who
were invited to stay, but only as regular judges.) What if we don't like
our best entry enough to give it Best of Show? Don't give a Best-of-Show
Tom discussed the Art competition. There were 46 entries (down one
from 1998), of which 9 won Distinguished, 6 Excellence, 12 Merit, and 19 No
Award. There were nine judges (up from six last year), and 16 entries per
judge (down from 24). The Silicon Valley, Boston, and Twin Cities chapters
(among others) submitted no entries. Tom said that three judges provided
"outstanding" comments, three were average, and three were "pretty poor."
For next year he plans to provide sample commentary for training purposes.
Tom was asked several interesting questions. Can I enter proprietary art?
No. How do we get more entries? Advertise for non-members, from hospitals
or other businesses. Where's Canada? No chapters outside the US held art
competitions last year.
SESSION: "A World View of Good Technical Publications"
Presenters: Carol Luttrell, US (chair of the STC International SIG);
Cathalynn C. Labonte-Smith, Canada; Luc Bouquet, Belgium; Terri Morgan,
China; Norio Kobayashi, Japan; Chris Rose, Canada
Consistency of terminology is crucial, because synonyms are very hard to
translate. Some language-specific tidbits:
US legal warnings are by far the longest.
Translation into French is tricky: If it gets too formal, Canadians
perceive it as "Parisian French" and object.
Why are Japanese consumer guides so poor? Translation is almost always
from English to Japanese; going the other way, it's easier to write in
In any language, spelling affects trust. (Moral: Spell-check your work.)
The biggest problem with graphics is embedded English text labels, which
often sneak in to clip art. Chris said that to save money, he once removed
a number of illustrations he judged "unnecessary." Customers complained, so
he had to put them back in. (Moral: There are no unnecessary
Translation Issues: Our decision to use translations agencies that rely on
the tool TRADOS was a good one: Luc commented that in Belgium, where
translation is commonplace, there's little reliance on machine translation,
but heavy reliance on TRADOS. He added, "Tools make a good translator
better, and a bad translator worse."
Layout Issues: Chris hired a professional to create a page layout for him,
and was very happy with the results. He also follows the tenets of Edmond
Weiss. Normally we worry about leaving extra white space for translation,
but that's not always necessary; for example, when translating from English
to simplified Chinese, there's a dramatic shrinkage in text area.
Production Issues: Print production is always PostScript or PDF when
SESSION: "Design Surveys to Extract Usability Information from Tech Support
Calls": Lori Kipnes-Bailen, Better On-line Solutions, Ltd., Galilee,
A tenet of usability design (and documentation!) is to know the user. Why?
Because knowing users lets us:
Challenge or verify our assumptions
Attend to users' goals and tasks in users' words
Plan easy transitions to new releases
Build in flexibility without complexity
Improve the user interface and documentation according to actual needs
Provide fast and effective technical support and reduce the number of
1) Identify available sources, their roles, and the users. Lori identified
Technical Support, which already asked questions when users called for
2) Design the survey. To the questions already asked (experience, nature
of the problem, and what was tried first) Lori added two questions: what
was the user trying to do, and what did the user do to get to the point of
calling for help?
3) Conduct the survey. Lori's company averages 100 calls a week.
4) Organize the results. (Lori used a spreadsheet.)
5) Analyze the results. Lori noted that only one time in three did the
subject in the trouble ticket match the actual problem.
6) Report the results. In Lori's case, the information she discovered led
to a formal agreement to send feedback from Tech Support to Documentation.
Based on problems users encountered, Lori's company wrote wizards to cover
some trouble spots.
SESSION: "How to Ask for What You Need So That People Want to Say Yes":
Kathleen Holm, Polly Joy, Sakson & Taylor, Inc.
This was really a session about negotiation and communication. (Polly was
a substitute presenter.) E-mail is a notably poor medium for persuasion (as
shown by a hideous example), because it's so limited: body language, tone
of voice, facial expressions, etc. aren't transmitted. Persuasion is best
accomplished face to face.
Understanding the other person's position is a key to negotiation. One
effective technique is to repeat what the other person says until they
agree that you're summarizing accurately.
I enjoyed this presentation, but it was more of a personal-development
session, so there's not much I can pass along. A question from the audience
touched on gaining respect. This is a deeply painful issue for technical
writers, and Polly answered very carefully. Her simple answer is that you
can't demand respect?you can only give it, and hope to earn it. She
acknowledged, though, our profession's long struggle for respect. I was
reminded of another STC conference some years ago, and a presentation by a
Digital writer on assertiveness that went spectacularly awry. The speaker
played videotaped scenarios of passivity, aggression, and assertion. His
point was that while passivity and aggression are bad, assertiveness is
good. But the vignette of a writer aggressively dressing down an engineer
triggered among the audience of writers a riotous standing ovation! Every
one of us could identify: been there, longed to do that.
SESSION: "Creative Ways to Reward Employees": Gloria Reissman, Diane
Beadle, Ellen Goldhaar, Michele Gordon, Doug McIntyre
Even though this was the "getaway" session (the last session timeslot of
the conference), it was overfilled; I stood in the hall and, as people
bailed, gradually worked my way into a seat. The audience was mainly
managers, and the discussion was lively and candid. I'll keep specific
suggestions to myself, but I can share these thoughts with you:
Rewards should be meaningful; they need to be fair, timely, and in
proportion to the achievement.
Interesting assignments are perhaps the best reward, along with
opportunities for team leadership; but not everyone can get them.
Make sure your stars know how much they're valued.
Training is considered "a huge reward, one of the most appreciated."
Training helps people acquire, maintain, or advance key skills, which keeps
them feeling competitive.
Career paths, both management and technical, provide choices, which
increases satisfaction and retention and reduces stress.
Service awards (5, 10, 15 years, etc.) are common.
Other major rewards are telecommuting, flex time, comp time, and the
universal favorite, free food. (But don't overlook dietary or religious
Rewards that mean additional time away from home and family might not be
Sometimes we don't appreciate what we have. One manager worked in a
Japanese subsidiary, where the corporate culture ignored rewards: if you
showed up and worked as hard as you could, you'd get a paycheck. Asked
about comp time, another commented, "Fifty hours a week isn't overtime."
(She was hissed down.) Finally, someone asked how to reward employees at
her company, which had low salaries and no medical benefits. (We suggested
she should resign.)
One manager said her responsibilities were project management, building
and retaining a good staff, and evangelizing for her group. I thought those
were excellent marching orders. Another manager boasted that her company
reimburses for trips to seminars and conferences, including a Saturday
stayover which lets employees have a little pleasant free time?and reduces
the airfare and hotel rates. I note that Lightbridge did the same for me,
and I appreciate it!
Visit TechWhirl's Other Sites