Session notes from HighEdWeb09 ============================== > See also [my conference recap], > my photos ( > [set 1], > [set 2]), > the [conference Flickr set], > and > [the conference Twitter stream]. Here, in mostly chronological order, is a big dump in one giant page of all my session notes from HighEdWeb 2009. Most of the presentations themselves will be available for download [on the conference site]. These are just the points, links, and quotes I found interesting enough to type into [Evernote] while attending the sessions. APS1: ADA and Section 508: Best Practices for Web Accessibility for People with Disabilities —————————————————————————————————————————————— This was my first session of the conference, and it set the bar pretty high. Lots of good stuff about accessibility. The main point seemed to be that while standards compliance is good, just fixing the errors in your page to get a clean run through a validator tool doesn’t create usable, accessible web pages. It is much better to learn to code in an accessible manner, and to take the time to understand *why* the accessibility guidelines are there. Coding HTML in a usable, accessible manner produces better, more usable code. Loved the quote from a designer, “I learned this stuff because of Web accessibility. I use this stuff because its better Web design.” - Instead of coding around the errors from accessibility evaluation tool, use design pattern that support accessibility. - People w/ disabilities want an accessible / useful site, not just a site that passes validation tool. - Try using opera — turn off images, turn on high-contrast style sheet. Can users w/ low vision tell what site they are on? - Illinois Information Technology Accessibility Guidelines — technical best practices document with what/why/how. - UIUC — iCITA Best Practices & FAE & FireFox extensions - Check out ARIA — Accessible Rich Internet Applications Suite —
- UIUC FireFox extension allows you to send rendered DOM from Web app
to FAE for validation.
- Great quote: “I learned this stuff because of Web accessibility. I
use this stuff because it’s better Web design.”
- cf. [www.admissions.illionois.edu] for visually compelling
- Accessibility users group:
APS2: Website Editing Using Ruby on Rails
I’d never seen RubyCMS before. Never used rails, never coded in Ruby.
About the time Ruby became a *thing*, I was making the transition from
mostly-programming to mostly-managing. This was an interesting session.
Not sure RubyCMS is the right tool for the tasks we use a CMS for, but
man, some killer features. Talked to the presenter after the session to
ask about the scalability issues (which is the biggest criticism I’ve
heard of Ruby), and he acknowledged it is an issue, but they are making
great strides in that area. In any case, good to know it’s out there,
and it might be a good option for some side projects, or for doing rapid
prototyping of interfaces to be fully fleshed out in some other system.
- Goals — make editing site easy for user; simplify
uploading/sizing/captioning of images
- Looking at RubyCMS
- Ruby - dynamic scripting languages written in C programming lang -
elegant, easy to read/write, requires less code; won’t make you a
better programmer, and definite learning curve, and some syntax not
as intuitive as claimed; once you get hang will become your favorite
- Rails - Web framework written in Ruby - Favors convention over
configuration; is “agile”: minimal planning, lots of prototyping,
frequent revision — working software more important than
comprehensive documentation… “Show me something that works.”
- Rails: [twitter], list apart, basecamp all built using rails
- Put /edit in front of URL to log in — LDAP connected. /staging for
staging version, shows most-recent draft, versions
- RubyCMS — inline editing, appearance schedule, killer image &
document manager. Drag & drop images. Navigation on-page drag & drop
editing. Forms not done yet. Object embedding not done yet. Awesome
way of using hashtags sort of like Views in Drupal
- Mongrel - http server written in rails
- Passenger - module for Rails on Apache
- Question: Are images & docs tied to a page? Any sort of asset
library so you can use same image on multiple pages
- Question: Is there a publishing workflow so a copyeditor can review
/ release the page?
- Question: This is very cool, lot of interesting features. Issue with
scalability and enterprise use? That’s the one criticism I’ve heard
of Rails specifically and Ruby in general?
APS3: Implementing Reason CMS with Small Teams and Small Budgets
This was a session about Reason, a home-grown CMS developed at Carleton.
For certain purposes it looks great. They have some pretty significant
campus Web presences built out in Reason, and it was impressive how much
content could be *managed* by a small team. If I was ever in the
situation where I had no budget, no manpower, and needed to let lots of
people work on a site, this might be an option.
- Kalamazoo College for International Studies. Using “Reason” PHP5 CMS
developed at Carleton.
- Beloit College runs whole site w/ 2-person Web team
- “Our content management strategy was to give the staff assistants
DreamWeaver and hope for the best.” \#heweb09
- Design idea:
- Luther College » phasing out RubyCMS in favor of Reason, migrating
SOC3: Talking to Your Boss About Twitter…
This was a best-of-track session, and was really good. Lori presented
some great strategies for talking about the value of social media
technologies, and for handling one of the great difficulties of higher
ed: dealing with brilliant faculty and administrators who aren’t used to
being unaware of things. My favorite recommendation was to start talking
about the work-related value of your *personal* twitter or facebook
accounts. Funny example story: one dean just starting with social media
tweeted that he’d give some departmental shwag to the first person who
saw the tweet and *called* his office… using new media to drive people
back to old media.
- We all have a VP or chair, etc., whom it is our duty to keep
informed of our activities and elicit their support.
1. It is called twitter. Tell me what’s going on in myface and
spacebook and tweeter.
2. They know… but they don’t know. They are smart people, they’ve
heard about it, and think they know about it. But they don’t use
it. So they know what it is without understanding what it is.
Twitter is easy to misunderstand. People tend to mock what they
don’t understand as a defense mechanism against the appearance
of ignorance — don’t take it personally.
3. “Justify” often equals “quantify.” It’s tough… how do you
quantify the value of a relationship? How do you put a dollar
value on a conference backchannel that enhances the conference
experience for some attendees.
4. Frequent question: shouldn’t you be working? They think of it as
a social thing that students use. You are working… since this
is your job.
- Strategies & Tips:
- These are smart people who are used to being in control and
knowing what is going on. They have reasons to be skeptical that
twitter is just another pets.com
- Statistics can be useful door-openers. These things are real
- Value not in information being exchanged but in network itself.
- You need to let them know this isn’t Santa Clause, I didn’t make
this up, this stuff is real.
- Facebook is 5.5yrs old; 300million users; 99% incoming freshman
have FB profile; 300K new users added every day; Look at \# fans
of university site.
- If users of FB were a country, it would be the 3rd
most-populated country in the world. Not insignificant number of
people who have taken time to sign up for this thing called
- It is much easier to show than it is to tell; let them “see” the
conversation instead of trying to explain how it works.
- Show difference between “This is me” and “This is me on
Facebook.” Demo — log into FB as yourself, get on site, and let
them see status updates. It’s simple, but they’ve never done it
themselves. These folks are used to knowing things, but this
stuff they don’t know.
- Let them see the conversation. Start searching for name of
institution on all these things. University FB page is only tiny
portion of institution’s presence on Facebook. The conversation
is going to happen with or without them — you might as well
- This is the point at which your boss may get scared. You have to
talk ‘em back from the edge — remember its not about control,
its about authenticity. It’s about having an authentic message
by letting people who are part of your community tell the truth,
be it good or ill.
- Search for university on youtube. See what is going on. If this
stuff was produced by communications office, nobody would be
- Take them for a test drive. Let them (help them) use the tools —
help them create a twitter account or facebook page.
- Suggestion from [@LoriPA] on social media strategy — first just do
it, then develop strategy. Add layer of virtual community to some
event on campus, call it a pilot program. Create sense of ambient
- What about your own social presence, not just the university — make
the case for institutional presence by showing work-related benefit
and benefit to university from your personal social media presence.
Experiment with ways to use twitter in your profession; work in PR?
Follow reporters & local media.
- The line to draw in sand: participate vs. dominate. You can’t make
viral video. Institutions need to find their feet in the virtual
world can learn to contribute, not control. You have every right to
participate in the conversation, but not the right to squash other
participants of the conversation.
- Above all else, **be useful**; give people something they can use,
do, have fun with — something they couldn’t get from anyone else —
then get out of way. It’s like a university-hosted social event:
school provides the room and the snacks, but they don’t create or
control the social interactions that happen at the party.
- Shared governance of institutional facebook page — add admins to fan
page to represent areas of university (e.g. one from sports, one
from marketing, one from admissions) then soft-coordinate updates
page updates; don’t overwhelm fans, try to limit yourself to one
update a day.
TNT3: Increasing Web Site Usability (With or Without a Redesign)
This session was best of conference, and was really good. In terms of
usability and writing for the Web, they didn’t have anything *new* but
they presented it in a succinct and memorable way, with great examples
and the perfect touch of whimsy. My first attempt to attend this session
was thwarted by overcrowding… I arrived about five minutes before the
session was to start, and was about \#10 in line out in the hallway, as
every available chair and place to stand inside the session was
My favorite thing they talked about was the rollout of a new CMS for
their school. They held open training sessions every day for six weeks,
and the rule was that nobody got access to the CMS (or their
department’s content loaded) unless they attended the session. Because
they were willing to hold them every day for a month and a half, they
got the backing of the administration: nobody had a legitimate
scheduling excuse for not finding one day in six weeks. The training
sessions, then, were full days, the first half was all about writing for
the web. The second half of the day was actual training on the CMS. I
think that is just brilliant.
- Obstacles: Information Architecture / wireframes / design / content
- Great quote: “Our site could have been featured on Webpages that
inhale vigorously dot com”
- Information Architecture: content revision may expose problems in
the IA; start small then work your way to the upper-level pages.
- Information architecture, **then** wireframes, **then** design, then
- Only ready for design **after** other work has been done.
- Did one-day content/CMS training sessions every day for six weeks.
Nobody got on site unless they went to the training. Half day of
writing for web, half day on CMS.
- People don’t want to ‘marketed *to*’; they want to be ‘communicated
*with*’. - Flint McGlaughlin
- Write content like a conversation.
- Users don’t read, they scan, hunting for keywords they are already
have in mind. If you can’t give them something they are looking for
in 10-15 seconds, they will go away frustrated.
- Don’t waste first sentence welcoming users to a page they already
know they’re on e.g. “**Welcome to the dept of who gives a crap.**”
- Talking about respecting diversity doesn’t tell users what classes
they need to take, what sort of degrees are available, what kind of
jobs they can get.
- Content — Reference & Research (Jakob Neilsen - [useit.com] / Bob
Johnson - [bobjohnsonconsulting.com])
- Content — Location, location, location: eyetracking studies. You can
assume users believe they are on the page they *think* they are
supposed to be on.
- Users don’t read text thoroughly. First two paragraphs = most
important info. Subheads, paragraphs, & bullets should start with
- Information about the page you are already on does not provide add’l
useful information to users.
- Place self in position of users: assume users don’t know anything
about your org chart, institutional / higher ed lingo, or the
program or department to which you’re referring.
- User’s don’t know what “core curriculum”
- Head it with what they’re calling it + what you call it; e.g.
“Basics / Core Curriculum”. Help them find things in their terms &
simultaneously educate them about your terms.
- Writing for Web: people scan Web pages, not read them. Go for 50% of
text on Web as print counterparts. Don’t give half the info — give
the same info in half the words. Cut out “marketese” and welcome
mat. Don’t be cutesy, make puns, or use clever spelling — nobody
will find the information.
- Find out what terms students are using.
- Users expect underlined or colored test to be a hyperlink. To make
key words stand out, use bold rather than underlining. And highlight
key words, not entire sentences. When scanning, the eye can only
pick up 2-3 words at a time. If you make a whole sentence bold,
you’ve eliminated the benefit of making words bold.
- Hyperlink keywords — they serve double-duty. Linking out, and color
draws attention to keywords.
- Keywords in headings vs. warning against bolding whole sentences —
headings should not be full sentences. Font differences & whitespace
help them stand out. Chunk out content w/ headers that break out
- Consistency — contact info same written format, text style, and page
location on every section of site.
- Good writing for web will end up laying out content in F-shaped
pattern w/ important stuff toward top and toward left… which
matches the way people scan web pages.
- Comments — While talking about using the language of your visitors,
they used example that people search for “masters degree”, not
“graduate degree”. I had to chuckle because we tend to use the term
post-baccalaureate. Lots of people searching for that one.
Jared Spool Keynote: Cooking Up Gourmet Experiences On A Fast Food Budget
First… my moment of geek glory… Jared Spool [retweeted my joke about
Very entertaining and mildly snarky, had a couple good routines about
the TSA as an example of standards turning into dogma. He also had a
bunch of slides illustrating what he called the design trap of higher
education: girls under trees. School homepage after homepage showing
pictures of girls sitting under trees. Too funny.
One of his main points was the results UIE has found of some extensive
on-going research of what makes Web team successful. On the process
spectrum of tricks, techniques, methodology, and dogma, they expected
the best, most successful teams to be in the “methodology” camp, but the
tighter a team adhered to a methodology, the less likely they were to be
successful. Not that methodologies are inherently bad, but institutions
that implement them promote a mindset that stifles creativity. The best
thing to do is to find your vision, then give your developers and
designers the toolbag of tricks and techniques that enable to fulfill
Probably the most controversial thing Spool said was something along the
lines that templates kill success, that the point of Web design is not
filling content into boxes, and that extant research provides no
evidence that templates result in good design. That’s not be the exact
quote: Spool was very careful and nuanced in his language. But his point
is that attempts at excellence which have, as their foundation,
*reliance* on a template or methodology, are doomed for failure.
- Process spectrum: Tricks → Techniques → Process → Methodology →
- The great higher ed design trap: girls under trees.
- No research-based evidence that templates result in good design.
- Research on successful web organizations shows the best don’t have
methodology or dogma, but rather focus on techniques and tricks.
- » Give your Web team lots of tricks & techniques.
- Three core UX attributes: vision, feedback, culture.
- » Can everyone on the team describe the experience of using your
design 5yrs from now.
- » In the last six weeks have you spent more than two hours watching
someone use either your design or a competitors?
- » In the last six weeks have you rewarded a team member for creating
a major design failure?
- Amazon.com is not a university. Look @ data, not at what someone
else in some other industry is doing
- 5-second page test. It’s called a 5-second page test because it
takes 10 minutes to do. Look at page for 5 seconds then write down
everything you can remember about page, On scale of 1 to 5 how well
does page communicate the mission of page (e.g. “Customer Service”
or “You can trust us” or “We’re here to help”)
- Great when pages do specific things — identifies if pages quickly
communicate their purpose. Take a page comp down to the student
union and ask the kids to do 5second page test.
- Book to get [“Paper Prototyping” by Carolyn Snyder].
- Focus on a few quality ingredients in order to produce the best
- Inuksuk: a sign that someone has been there before. “I have made
this journey before you and you can make it to an be happy.”
- News feed on campus home page is usually just stuff that only the
marketing department cares about and usually isn’t news. Doesn’t
reassure students or parents that others have succeeded.
- There is a difference between being creative and being clever.
- Its not about the money you spend It’s about focusing on what you
want to do and building the toolbox of tricks and techniques that
enable you to do that.
- You need to focus on developing great tricks and techniques — don’t
let methodologies and dogma bog you down.
- Look for opportunities for creative approaches.
- First step: get all people who work on public facing site together
- Great quote: Rick-averse organizations produce crap.
For the 4th session I went to a vendor demo of OmniUpdate. They have an
interesting SaaS model. Pros: use push (or baked) model of content
creation, XML & XSLT, very fast, lots of higher ed experience. Cons:
lack of pull (fried) model of content delivery creates difficulty
attempting to deal with user-generated content (e.g. Web forms,
comments, user profiles) on
dynamic or social-network-type sites. Sounds like they are doing
something to address the
user-generated content issue, ala RedDot’s LiveServer. Definitely worth
MMP5: Starting a Web Office from Scratch: Trials and Tales
This was a pretty interesting session co-presented by two relatively new
Web team managers in two very different organizations, talking about
what does and does not work when setting up a Web team. I’m not sure
there was so much that I took away from this other than an appreciation
for how good we have it in Outreach Marketing. But if I was ever able to
split our web technology team out from Creative Services into a separate
team that worked across silos, some of the “how” in the presentation
would be useful.
- Committee policies and guidelines oh my
- HE Web — everybody wants to be able to have their say and touch
everything in case its a success.
- Creating a Web team is about getting the right people in the room at
the right time: this means the people who are paying.
- There is a compliance group, which is never the right group to start
- Committees are good when they server a function, but if they hinder
instead of enable they are a problem. Governance can be effective as
advisory or best practices, not as compliance enforcement body.
- Politics — don’t talk about them
- “Matt, I’m really happy for you, and Imma let you finish, but Wayne
State had the best Web committee of all time.”
- Completely decentralized — every dept has own web team, tasked w/
bringing everyone into unified group and single CMS. Find IT folks
who have been roped into Web as “other duties as assigned” and
position centralized CMS as way to get them back to concentrating on
- When plowing a field, if you hit a rock, just plow around it.
- Structure of Team
- Project Lead oversees
- \* Tech Team: Sys Admin, Programmer, HTML Coders
- \* Design Team: Graphics Designer, Photographer, Multimedia
- \* Content Team: Communications Specialist, Editor, Writer
- \* Quality Team: QA, IA, UX Designer
- Function: Service oriented group (works on projects) vs. global
functions group (oversight of enterprise web). Can they be the same
- \* Service group can build relationships with people who really want
to make a difference at university and can steer around roadblocks.
Notre Dame has good agency model. Can produce things of quality with
focus on user.
- \* Central group cannot always set timelines and have luxury of time
to create quality stuff.
- W-S Location Unknown — IT vs. Marketing vs. Hybrid
- “A CMS is not a magic bullet that will fix every problem in your
organization.” Technology will not fix your people problems.
- Point of CMS is to remove technology barrier that prevents end users
from updating the Web pages for which they own the content.
- WebDev mail list — Having somewhere people can send an e-mail to
w/out getting a negative response or flamed builds trust and
- Starting group — started ning group, hunted out people who do Web,
have meetups. Everyone has some talent, get them to present on it.
Start internal awards. Encourage people to share their talents.
TNT6: Get Your Easy Button: Web and Marketing Working Together
This session was more fascinating as a success story than as anything
with real strong takeaways for our team — because we don’t do any
undergraduate admissions work within our team, there wasn’t a whole lot
to apply in my day-to-day job. But what Xavier was able to pull off with
a small team with tight deadlines and minimal budget is very impressive.
Their average freshman class was around 850. They were shooting for a
very agressive 950… and ended up with a freshman class of over 1100.
Were I ever to work in admissions, I’d go back to this presentation and
steal ideas. I also appreciate the fact that they tracked *everything*
and were able to pull out good statistics, like the fact that on average
20% of accepted students actually attend, but that number jumps to 40%
if they received a peer phone call, and 70% if they attended an
on-campus event. Also an interesting idea with their “Road to Xavier”
site, which is basically a social networking portal for *accepted*
students to get to know each other and learn about the campus and their
program between their acceptance in Spring and move-in in the Fall.
- Charge - drive applications & deposits — focus on recruiting for
larger programs, and marketing struggling programs
- Project: Marketing (University Relations) / Admissions (Enrollment)
/ Web Services (IT)
- Transition from high budget local awareness campaign to
program-specific and outcome focus.
- “[Road to Xavier]” intranet for accepted students. Financial aid,
interact w/ other new students.
- Plan: 3 print pieces (create awareness, promote events, highlight
outcomes) / Landing pages (for each academic dept, use Road to
Xavier design, ability to create profiles, videos, event reg) /
Events (admissions working w/ faculty)
- Add’l elements: profiling of students / alumni, e-mail, alumni
phoning, faculty calling, feeding back data.
- Division of Labor:
- \* Admissions (events, faculty contact, fact sheets, phoning -
conduit to academic side)
- \* Marketing (print, email, profiles, photography/video, alumni
involvement - work on content)
- \* Web Services (landing page framework, incorporate road to xavier,
data back to admissions, event registration system).
- Messaging: I am… (role currently) / I am… (role in service) / I
am the Power of X:
- I am a teacher-in-training / I am a shelter volunteer / I am the
power of X.
- 3 print brochures for each program: outcomes (big deal for parents),
program specifics, corresponding Web site, visit opportunities.
Mailed in clear envelopes. Each one tells a story — Focus on current
students & young alums (no more than 3yrs out), tell a story.
- URLs: [roadto.xavier.edu][Road to Xavier] »
- Faculty member contact for each dept w/ name, photo, quick contact
- Mail pieces followed up 2wks later w/ e-mail. Teaser about
student/alumni profiled, link to video, links to accompanying Web
page and event registration.
- Phoning campaign. Student phonathon call accepted students 2nights a
week for 1month. Connect w/ Alums campaign → 110 alums participated,
each assigned 3-5 students based on location — provided calling
scripts. Faculty → if they want students, they need to put on events
and make a few phone calls. Got support for deans. Took panic of
economy and push from administration to make it happen. Website w/
online calling reports.
- Students who attended events 70% enrolled, 40% students contacted
ended up enrolling, vs overall 20% of admits enrolled. Some combined
departmental events w/ general open house event to give students
more bang for buck on trip to campus.
- Enrollments went from 860 to 1174.
- Question: how did you measure success?
- Question: Recruiting in Cleveland
- Question: Cultural question — Was this a cultural transformation?
did marketing/admissions/IT have good relationship before, or did
this only happen because administration put the pressure on?
TPR9: An Argument for Semantics — Why Developers Should Give a Hoot about OWL
Without question, the session from which I learned the most new
information. The session was presented by Brian Panulla, whom I’ve known
*about* for several years (he used to work at Penn State) and who I’ve
followed and conversed with on Twitter for quite some time… but I’d
never met him, or had any real idea what he actually did.
Anyway, this session was all about the semantic Web, or Web 3.0, or
whatever. Which I’ve heard about, but wasn’t real clear on: I knew it
was something to do with tagging data, and [FOAF] was Web 3.0, but
[microformats] weren’t, and it was all very confusing. By the end of
Brian’s presentation, I was able to say, “oh, I get it now.” On a side
note, I was a philosophy major in college. The intersection of RDF
ontologies and philosophical ontologies was particularly interesting.
Brian made some comment about how ten years ago it was the library
scientists who were able to usher in whole new ways of looking at
information on the Web, and that over the next 10 years it might be the
philosophers who have something to offer. It will be interesting to see
where this goes.
The part that was missing — and this was only a 45-minute presentation,
so there wasn’t really enough time — was some real-life examples of how
this was actually being used in a way that made people’s lives better.
- Semantic Web - Anti-social Web
- Looking for smarter web - ontology: the “o” word
- Semantic Web / Web 3.0 / Linked Data — all same thing
- New W3C langs: RDF / RDF Schema / OWL - each builds on another, but
all fundamentally RDF
- RDF - mark up information in ways that are useful to machines (not
particularly humans) to ease pulling / aggregating / connecting /
displaying information to users.
- If humans can’t get implied meaning right, how are machines supposed
to get it right (context & baggage when dealing w/ language)?
- Google is very very clever, but it’s not very smart.
- \* Why do we maintain local copies of public information
- ? Is there a public API or updated database of record for
- Semantic Web - parallel infrastructure - design patterns for smarter
infrastructure. Web content, pages, and sites do not need to change
to be made ready for the web 3.0.
- More meaningful markup allows software to make more robust
- Roadmap to smart data: entities as resources / specifying
relationships / drawing inferences.
- Differentiating between conceptual entities creates the need for an
identifier: indefinite article (a college of IST) / definite article
(THE college of IST at PENN STATE). — how do we get those proper
nouns into our data?
- Implicit semantics: USPS state code / ISA country codes / 2-letter
symbols for chemical elements
- Two columns called “state” could be USPS state code or FSA state or
- In the absence of a good key, we make one up, but how to we
differentiate between entities even if they have the same name?
URIs. We can create string that works as global identifier.
- “You need to be careful about the difference between a URI and a URL
when you’re dealing with semantic web people. They get real touchy
about those things.”
- There is no guarantee a URI will be accessible from a browser… but
it usually is. “I link therefore I am.”
- Normally entities identified by two different URIs are distinct but
you can override it.
- RDF: fundamental knowledge representation declaring resources
(nouns) and specifying properties (verbs). — specifies is-a,
is-a-member-of, has-a relationships.
- RDF/OWL stored as triples. Subject (Penn State), predicate (is a),
- Creator of ANT has apologized for using XML as the configuration
- RDF can be expressed as XML, or as N3, or other formats.
- Types/properties in RDF can be defined in RDF schemas or Web
ontologies — think of classes in RDF schemas (RDFS) as sets than
than OOP classes. RDFS adds limites set theory properties
subClassOf, subPropertyOf, Domain, Range.
- Ontology - the study of a being or reality. Higher level of a schema
that describes a being. Computer Science profs saw word in
thesaurus, saw it was bigger than schema and said “we’ll use that.”
A Formal Ontology is a representation of a true ontology in some
sort of communicable format.
- Schema describes structure / Ontology can describe structure and
- OWL classes adds expresivity of sets. Can say psu:College and
umich:College are the same thing. Enables us to get over differences
- War brewing between taggers and top-down hierarchical ontology
wonks. Class structure is not a strict hierarchy: it’s all about
sets; classes can belong to more than one parent.
- Inferences are statements that are derived from other facts —
getting at new information based on what is already known; by
declaring relationships as transitive/inverse we can get information
w/out storing it separately (as you might need to do in RDBMS).
- Using these: add a semantic data to your existing app by reading
source, and query ontologies using SPARQL.
- \*Look for an ontology for your domain before you create your own.
- \*Don’t try to model the universe. It gets the philosophers mad,
because its their job.
- \*Don’t use the word “ontology” in mixed company.
- \*Do use an upper/foundational ontology — BFO, Cyc, DOLCE — Easier
integration later, makes philosophers happy.
MMP10: This Is Not A Brand
This was a co-presentation by Doug Gapinski from mStoner (who I’ve met
before) and Patricia VandenBerg from Mount Holyoke College (who I’ve
never met). Doug presented *via negativa* what a brand is not, and
Patricia presented the successful implementation of some branding
strategies on the Mount Holyoke site. The big takeaway for me was that
your Web presence needs to tell prospective students what it’s like to
go to school there, what other people think of the school, and what’s in
it for them.
- Using the Web to better present your institutional brand
- The Treachery of Images: “Leci n’est pas une pipe.” It’s not a
pipe… it’s a painting of a pipe.
- What a brand isn’t: not a marketing platform, not advertising
messages. Blue is not a brand — it is part of a brand.
- What a brand is: what is it that is special about this institution
to our various constituencies, and what is it that will add value to
what they looking for that will make them choose us?
- Web is *the* brand medium:
- place for authentic content
- experience-oriented (rich media as storytelling devices) —
storytelling one of best ways to carry brand
- effective distribution — investment on front end, but no
recurring printing costs
- Brand answers three questions:
- What’s it like to be at *your institution*
- What do other people think of *your institution*
- What can *your institution* do for me?
- Ask yourself: are you answering what’s it like to be there, what’s
in it for me?
- Mount Holyoke — appeal to international students — have current
students in videos speaking in native language, use English
subtitles. Reaching out to parents of non-native English speakers.
- MHC — tested w/ background colors. Found incoming students had no
association w/ blue, found it cold. Testing found yellow worked
best. Alums up in arms. Put blue on alumni pages.
- Color is a symbol, but only if it means something to you if you have
a relationship with the university already.
- This is the first generation that will go to Facebook and Youtube
before going to your main school site, so you need to watch how you
are represented there.
- — great example of true authenticity in truly
- **Books to check out**
- [Strategic marketing for educational institutions] Philip
Kotler / Karen F.A. Fox
- [Zag]: Marty Neumeier
- [Beyond Branding]: Nicholas Ind
- Link to presentation:
MMP11: Maybe the Purpose of Our Redesign is Only to Serve as a Warning to Others
This was Anthony Dunn of [CSU Chico], author of the snarky [Tales From
Redesignland] blog. This was a hoot, but Tony delivered what he
promised in the session description: an open discussion of a painful
redesign process with honest discussion about things that worked
amazingly well and things doomed from the start. This session had me
laughing more than any other, but it also had lots of good info about
how to organize an institutional Web redesign team that works, the value
of testing, and the importance of a content strategy.
- “Our information architecture hasn’t changed since 1999. Next month
we’re throwing a 10th anniversary party.”
- “Our current site design was launched in April 2004. It is entering
kindergarten this year, and we’re all very proud of her.”
- Institue a Web governance structure.
- Cabinet Level ITEC » Web Management Committee & Web Content
Committee » Web Design Team.
- Web Management Committee examines recommendations of Web Design
Team to see if they fit the mission of the University. They do
not direct, they approve. Web content committee is not
decision-making body, but a communication channel (advisory
body) who communicates out to constituents.
- Academic Politics: Their viciousness is only matched by the
insignificance of the stakes.
- Engagement — not as interesting as watching COPS.
- Creating a competent and sufficient team is VITAL. A team of Web
nerds is not necessary but not sufficient.
- skillset.org “Web Team” Kristina Halverson (book on content
- Danger: getting stick in the mode of brainstorming. Bright ideas —
too many of them give you a headache.
- Need outline of steps, milestones, and deadlines. “The journey of a
thousand miles often ends in a mud hole.”
- Budget problems provide constraints that create focus on results.
- **Do the research** — you can’t fix it if you don’t know what is
- Used crazyegg for heat tracking. Heat map of home page was single
most valuable piece of information. Visual representation of how
people using home page. Totally internally audience.
- User survey important from political standpoint — went to every
single faculty & staff member — let everyone feel they had a voice,
and that you are doing something.
- Best Practices — took 20 top sites, and inventory every link and
piece of content on home page and create a matrix who is doing what.
Useful as research to back up decisions… 80% of successful
competitor sites do this, only 5% do what you’re asking form.
- “If people *think* you have done the research they will believe
anything you say.”
- What were jettisoned due to budget constraints: personas. usability
testing of current site.
- Focus groups — sometime their body language is all you need to know.
Two types of prospects (early (jrs) / soon (about to pick campus),
faculty, staff, current students. Invaluable to get real information
about what users want from homepage and top level pages.
- Stakeholders: there’s a lot of them, and they are angry, but hey, at
least they’re wearing campus colors. The top-level pages need to be
a clearinghouse for information; needs to communicate to audiences
in their language, and guide them out to the complicated processes
managed by disparate groups on campus.
- **Plan your content** → \#1 reason redesigns fail is content. If
“under construction” is your content strategy, you’re doing it
- Mindmap for info arch of top-level site.
- Great quote: “Prospective students can smell 1 part per billion of
marketing on a Web page.”
- CMSs will not magically create your content. Someone needs to write
all that copy. Someone needs to load it all into the pages. Have
content strategy and plan for implementing content, or you will
spend rest of your life trying to wrangle content.
- Just because you wear a suit doesn’t mean you are qualified to
critique Web design.
- Research helped w/ design. Big battle ended up being over CSS
frameworks. “I’m not going to recode that to make it work w/ Yahoo
- Get buy in and high-level ownership. If you don’t have that, you
aren’t going anywhere.
- Make sure you have the right people on your team.
- Clearly define the project and it’s scope or you are doomed.
- Do the research. Do what you can. Anything is better than
- Get input, feedback, and buy-in from EVERYONE, or at least
everyone you can.
- Have content strategy and plan for implementing content, or you
will spend rest of your life trying to wrangle content.