This came up recently for me because of two things:

First, I received a regularly scheduled newsletter that apparently decided to hit send even though they didn’t have content to fill out their mailing. The screenshot of the cse in point even admits they didn’t find enough good content to send, but send they still did. Now, in their defense, the email is typically full of good links, and this particular edition did too, just not in this particular section. So I wonder, why not wait on this mailing until there’s enough good content to justify hitting send?

Second, I’m in the process of thnking about e-newsletters on behalf of my university’s alumni relations group. They currently send an email on a bi-weekly basis. That’d be fine if the content in those emails supported such a constant stream. However, and it’s one man’s opinion of course, I don’t think it does. It’s chock full of stuff I don’t care about (I happen to be in their target audience so I can make that claim), is way too long, has no focus/theme/glue to what is included, and it doesn’t look appealing. So why not abandon the bi-weekly schedule and move to an as-needed basis? It’d be less work internally to create and become more valuable to recipients because it’s only sent when there’s something valuable worth sending.

User tests show that students have consistent informational needs when deciding which university to attend:

1. Do you have a program of study that interests me?
2. Can I afford that program?
3. Does the school’s culture/vibe feel right (will I fit in)?
4. Grad level students’s needs will lean more towards faculty, their interests and research opportunities away from cultural fit on a social level (grads don’t usually live on campus so the social component isn’t as important)

This basic set of questions all revolve around degree programs, not the broader departments within which they exist. Because of this level of specificity, departments should take a secondary role in how a higher ed site is structured. Degree programs should instead be the central organizing framework.

I see too many university sites where I can find a program of study through the top level pages only to be taken to a departmental site’s homepage where I have to find the same degree information I thought I was originally linking to all over again.

With a shift in how higher ed sites are organized, other pieces begin to fall into place: building communities around logical points of interest, presenting appropriate content (research, faculty, pricing, culture, etc.) within context and, importantly, filtering out a lot of stuff that’s not relevant because it has nothing to do with a student’s preferred degree program.

I love Boston.com’s Big Picture and want to bring it to the University of Denver. We already get a stream of student supplied photos, but we hide them behind our password protected internal system. A true shame, but hey, new year’s resolutions…

My bias is toward a balance between marketing and usability, but the bias gives more initial credence to usability than to marketing. Why? Because you can market the hell out of a site, a subsection or page, but if people don’t know it exists, then it doesn’t matter. That’s my main criticism of the DU homepage- It looks great (which is valuable in and of itself), but it comes at the expense of visitors (prospective students) getting their tasks completed slower and more inefficiently. The most powerful way to manage this tension is to 1) ensure your basic usability requirements are met and 2) make that basic usable framework as beautiful as possible.

If we look at the Apple and Google examples on this page (you can click them to access full size versions), you see this 1-2 approach put into practice. Google is a usable site. It gives visitors a short overview of the conference, easy to find/use links to additional information and a big button to join. But is it beautiful? Has it leveraged good marketing principles? Well, it does provide user generated content to breathe some life into the page, but I would argue that it’s not an engaging presence. It has leaned too far toward usability by disregarding the power good marketing would provide.

Apple’s page also provides basic information on the conference, links to additional information and a big button to join. The difference is that Apple takes that basic usable form and packages it with more marketing muscle. The page is beautiful. It’s engaging, active and makes me want to go in order to be a part of the energy depicted. This is a page that has found a good balance between marketing an usability.

Both conferences, being as high profile as they are, will undoubtedly sell out (Google’s already has as of this writing) so you could make a counter argument that Google’s lack of marketing didn’t hinder them. In fact, they likely saved money and time by avoiding the marketing point of view. True enough. But which conference do you want to be a part of?

Before we dive into the details, we need to define marketing and usability. By “marketing” I mean a perspective that exults the intangible- largely subjective areas like branding and visual aesthetics. By “usability” I mean a perspective that exults the tangible- things that are perceived to be objective through testing and measurement like navigation and functionality controls, categorization and flow of information. I realize you may disagree on my definitions, but for the sake of argument, I’m not here to say one is better than the other so feel free to change the definitions in your own mind. I support both as I’ve defined them. You should ensure that your visitors are represented through testing and measurement, but you also need to be a leader sometimes and do what you feel is necessary even if it’s contrary to user’s wishes. The two can work in tandem, but so often fail to do so. However, that is not a recipe for disaster in and of itself.

A higher ed site could go in either direction. DU’s homepage is squarely in the marketing corner while the task of finding degree information is squarely in the usability corner, yet neither truly works as intended for me, not to disrespect Nick’s conclusions. The homepage does indeed have impact and bucks the usual higher ed trend, but does it work? It does if you want to grab attention and differentiate yourself from the pack (I’ll assume a prospective undergrad student audience). But do students want different or do they want ease of use or a sense of what life at DU is like or something else? Is the leadership inherent in publishing such a bold homepage good or bad? I’d argue that the homepage misses the mark.

I don’t see any reason to be bold here when so much of our own research and those of DU’s consultants over the years points to the fact that people researching what college to attend are more interested in getting the four big questions answered as opposed to being “marketed” to:

1. Does the school offer the degree I’m interested in?
2. Am I qualified to attend this school?
3. Will I fit in socially/do I see myself being happy at this school?
4. Can I (and/or my folks) afford it?

How many of these does the DU homepage answer (or how many of these questions can you easily get to if you make your way to an internal page within the DU site)? To varying degrees, there are links and clues for each of them, but they’re overwhelmed by the gigantic photo and audience links. This page is more about DU than it is about DU’s customers.

UPDATE 4/29/2010: I decided to expand on this idea using a non-higher ed example.

• Fewer in-person client reviews: Wayne is hired by various departments for photo shoots. After a gig, he used to schedule an in-person meeting with his client to review and choose the final photos that the client would ultimately take away with them. With the introduction of Flickr, he now uploads all the photos from the shoot into a Flickr set and gives the client access to it. The client then goes through and chooses the photos they wish to keep and deletes everything else. Our photographer saves himself an average of half a work day per week which frees him time to shoot other jobs, time to post process a client’s final selections and time to take care of non-billable, house cleaning tasks.
• Fewer photo searches: Wayne, as the sole photographer for the university, continually receives photo requests for use in various materials (marketing collateral, website, banners, etc.). Each request required him to go through his archives and ferret out an appropriate sampling of photos. Clients would either come to him in person to review or he would burn a CD with images and send it to them. With Flickr, he is now able to send people to an online archive of photos (in this case, he sends them to either the internal account or the public one). Once there, clients can download high res versions of anything they find and know that whatever they come across is approved for usage.
• Fewer variable costs: While not a huge area of servings, Wayne is able to cut down his use of CDs, jump drives, etc. because he now uses Flickr as a delivery method instead of physical media

Current Workflow

To make this work for Wayne, we plugged Flickr into his existing workflow so that his routine wouldn’t be overly disrupted. That process goes something like this:

1. He downloads photos from his cameras into Lightroom
2. He does a first pass through the raw files and throws out any obviously problematic photos
3. He uploads the photos into a Flickr set using Jeffrey Friedl’s “Export to Flickr” Lightroom plugin
4. If the set is uploaded to the public account, the set is marked as private and an automatic Twitter message is sent to one of our editors for title, description, and other metadata inclusion before being marked to public
5. If the set is uploaded to the internal account, he gives his client access to the page and waits for them to choose the final shots they want

Truth be told, the above ideas are still being tweaked as the dust settles. Even so, Wayne has saved himself a good deal of work while our department has better served our internal clients as well as expanded our content offering to our various audiences (through direct hits to Flickr as well as embedding content into our core du.edu website- our annual report site is a good example of that).

Opportunities and Problems

One other workflow idea we’re working to incorporate now is to include the university’s archive team. The holy grail here is to have Wayne send his photos to archives for inclusion into their storage system and then pull the images we want to show in our public Flickr account from their database. The benefit gained is that Wayne has to send his work to archives anyway (per university policy) and, since the archive team appends a consistent set of metadata fields to each image, we can skip the step of using up an editor’s time to do the metadata work on the Flickr site.

Another idea we may try is based on an idea from Brad Ward. The jist is to use some fun gadgets to auto upload images in real-time from Wayne’s camera while he covers an event live.

One issue we’ve encountered is whether or not to make the internal account private or not. We didn’t, for example, want to post hundreds and hundreds of photos of any single event for public consumption, but Wayne has found that managing client credentials needed in order to access the private account was becoming more work than it was worth. So at the moment, all the photos are public, but not promoted in any way.

Other little issues have cropped up, but nothing that can’t be solved. We’ve reaped a lot of benefits from this move and are happy we did it.

Since title tags are an exercise in telling people what they’ll find behind a link before they actually go there, the act of writing it requires you to justify the relevance of the link in the first place. If you’re at Apple’s website on the Macbook page, you might see a link to their Macbook Pro page. Makes logical sense, right? If you’re interested in a Macbook, you might be interested in stepping up to a Pro model. A title tag might say “Step up to a Macbook Pro for added performance, storage, memory and more.” The sentence establishes relevance and a reason why you should click or not click. Job done, move on.

Let’s take another example, however. Let’s say you’re on a university’s annual report site, on any page. There’s a global link to the chancellor’s site. You write a link title that says… what? “Go to the website for Chancellor so and so.” No, that’s too obvious. “Get information about Chancellor so and so.” No, that’s not relevant to the annual report as a whole. “Get Chancellor so and so’s impressions on the year’s events.” No, if that information existed, it would be part of the annual report site itself.

The above reasoning hints at the utility of link titles. Writing them forces you to double check your architecture. Why does a link exist on this particular page or in the global nav? Is it relevant to include here versus over there? How does the inclusion of this link in this area on this page help the visitor accomplish their goals or further their aims?

All of these questions should have been asked early in the process, but things slip through or circumstances change. Writing link titles help verify that your user experience goals are kept intact and on track. Try it, it works.

So where does decentralization make sense? The obvious answer is content. Higher ed sites are large, if not huge, relative to many websites and the thought of centralizing that amount of content into a few hands doesn’t seem practical. The sheer workload would jeopardize the distribution of time sensitive information. Plus, no content person wants to work in a sweat shop environment were quality takes a backseat to simply getting the work out. And beyond even those practical concerns, will a content person be as passionate about every subject that comes across their desk as the people who live and breathe it?For those reasons, content ought to be unleashed.

This isn’t anything new Mike

OK, so a decentralized content scheme isn’t a new idea, but my point is that simply because it’s smart to decentralize content creation doesn’t mean ALL aspects of the web effort should go along with it. This includes strategy, information architecture, visual design and functionality. There’s no credible reason that I’m aware of that would cause me to believe that every dean, vice chancellor, and director at the university should control how their part of the website works or functions. That’s not to say they shouldn’t be at the table to discuss matters that affect them, but neither should they be allowed to dictate needs on a per site section basis either. This is how workloads get out of hand.

Give me an example

If group A wants to promote news, group B wants to promote events and group C wants to promote both, what happens in a decentralized world? The web team gives each group exactly what they want. But if information architecture is centralized, then you would look at all groups in totality. You would take all wants and needs into consideration en masse. Only then can you create a solution that can be sustained over the long haul in an efficient manner. In our example, you might come to the conclusion that lots of groups across campus will want to promote news and/or events. So instead of building standalone news and event solutions plus one that handles both, you could merge news and events into a single steam called what’s happening. Any group on campus could then tap into this single solution. Group A doesn’t have events? No problem, no events populate their content, but the “what’s happening” label still makes sense. Same goes for group B in reverse.

Technology folks will of course try to build solutions with modularity in mind, but this only goes so far. News and events is an easy example since we all know those ideas will port across many departments. But what about requests that have limited universal appeal or practicality? Do you build it? The answer lies in your overall strategy. 

You may not have a global strategy precisely because of the decentralized environment. But once centralization occurs, that all changes. You should have a basic understanding of what your site is supposed to achieve and for whom. In that light, a request for a specific feature that is only useful to a certain group or for a particular, somewhat rare occasion should be looked at suspiciously. If your site is supposed to communicate with prospective students and entice them to apply, then there may not be a good case, for example, in building a Blackboard login function. I’m not saying Blackboard doesn’t have utility, but not for your prospective student audience nor to achieve your goal to drive applications . The correct answer in this case would be a polite “no.”

Instead, I see many sites, our own included, take on projects to fulfill any and all requests without much critical thought as to their strategic value. So more projects enter the queue, timelines stretch out, both clients and customers get frustrated over the slow progress, staff morale suffers and the website becomes ever more confused. I would imagine that if you looked at your institution’s workload, you’d find a significant amount to be, at best, tangential to your main goals and, at worst, unrelated at all. Kill those projects and focus. Centralize the high level, strategic functions your group provides while you decentralize the ground level content that gives visitors localized flavor.

• A lot of the “early” tweets I see, ironically, link to or are part of a link chain that trace back to a traditional news media outlet which squashes my main point. That said though, I don’t follow every news company out there so in effect, comparing the news outlets I do follow vs. all my other non-media follows, the peeps win the race to inform me often.
• This only happens with national or international news. For local coverage, @denverpost is much faster. In fact, I don’t think I’ve ever gotten a local news tweet from anyone I follow. @denverpost is the only game in town in this regard. 
• I’m not the only one who has noticed this trend: @isaacson said not long ago… wow twitter is ridiculously ahead of CNN and the mainstream news in reporting the #GAZA ceasefire. Also, I took a screenshot (but it’s at home, I’ll update when I can UPDATE: Screenshot below) of two tweets on the same subject next to one another in my Twitterific stream. CNN was the later tweet.
• I wonder if anyone DMs or replies to their local newspaper or TV news station about niche news events in hopes that the news organizations could then broadcast (or, more likely, re-broadcast) to its followers as a micro-local news stream. They could add those mentions in their main Twitter account or, if the barrage became too spammy for a general public audience (which by definition it would be), they could set up separate Twitter accounts. But I suppose at that point, the Twitter accounts become so niche focused, the effort involved would outweigh the number of followers. Then again, maybe they could have a local feed and a micro-local feed and let people choose for themselves. Anyway, something to consider- and it probably has, I just haven’t come across it.

Here’s the Twitterific screenshot I mentioned above. Cameron Moll beat CNN to the punch.

Let’s face it, a CMS isn’t much good if its more painful than beneficial. The promises sound great, but the reality may not be realized unless you inject yourself into the process early. But don’t take my word for it, take Jeffery Veen’s advice: get your editorial process worked out before you do anything.

You need to know exactly how work gets done in your organization before you begin any CMS selection. Every organization has a unique mix of people, deadlines, dependencies, hierarchy, new requests, ongoing maintenance, visitor testing, stats analysis, strategy refinement, and on and on that affect editorial flow. If you don’t know how all of these pieces fit together to form your editorial process, then you have some homework before any CMS talk is begun. Editorial processes are not static, they allow for some wiggle room and that’s OK. But knowing where wiggle room needs to occur helps you determine what CMS will work well for you. The better you understand your work process, the clearer your needs become and the easier it is to see which CMS is right.

If you leave this decision to the tech department, you could be in for a long long trip. Now, please note that this isn’t a rant against developers. It’s a rant against a flawed process where the CMS is seen as a technology issue to be worked out by technology workers. Of course the developers should be central to the decision, but the CMS choice should revolve around the people who use it, how they use it and why. Anything less will result in just that- less. Less enthusiasm, lower morale, perhaps even less productivity.