Archives: Information Architecture

UX v. SEO: Content v. Form

//4 Comments //Link

Imagine a hypothetical Web site that is ready to launch. It looks great, test users love it, the client loves it, and everyone is happy. Everyone, that is, except the SEO expert.

The SEO expert is, after all, responsible for guiding and delivering search engine optimization (SEO). SEO experts provide methods to drive a site up the Google results page, and because what they do often can be quantified quite easily as success or failure, SEO experts can be very specific about what changes are necessary to ensure that the site shows up when people want to find it.

Or, rather, they help ensure that the site shows up when people search for certain things. We’ll get to that why that difference matters in a moment.

An SEO expert might want the page titles changed. No biggie, right? They might want keywords integrated into site copy and image alt tags. Again: that’s cool. So far the great-looking, user-approved, client-adored site is intact, and the “under the hood” SEO adjustments don’t mean any concessions.

But what if the results aren’t good enough? What is the SEO expert to do?

Answer: they want more keywords. They want more and more keywords. They want links and other adjustments to the site that often not only spoil the aesthetic nor the way the client views the site, but the way that users experience it.

The stage has been set, then, for the classic SEO v. UX battle.

User experience (UX) encompasses how the users navigate a site, how they feel about a site, and how key features and messages are delivered. Good UX adds value and, when coupled with content users want, helps define the value of a site.

An argument for compromising UX in the name of SEO is relatively easy to make: if users cannot find the site, they cannot enjoy the experience. By giving a little on the experience front, the thinking goes, traffic will be generated and the site will be better off for being … worse.

In order to challenge this position, I think it’s necessary, first and foremost, to examine what SEO is at its most basic level. It is the manipulation of the difference between the value of a site and what search engines perceive to be the value.

Let’s deconstruct that quickly before we move on. A perfect search engine would know what the user wanted to find and give them the best possible value given that request. It would not be fooled by fake Web sites and it would be able to account for minor user error to point them in the right direction. No search engine is perfect, though, and because of the algorithms and logic used by the software of Google and Microsoft and others, inefficiencies exist: gaps between what the user wants and what the search site delivers. SEO manipulates that inefficiency by knowing (or guessing) at what search sites look for and “gaming the system” for positive results. If that perspective is too cynical for you, gentle reader/conscientious SEO expert, another way of looking at it might be that SEO strives to accurately convey the true value of a site given the search engine’s inability to correctly judge what the user wants when she searches.

SEO, then, is a cosmetic cover-up for the acne-riddled face of search sites.

The early days of search sites saw directories dominate. Microsoft and Yahoo employed teams of worker bees to review sites and classify and rate them, allowing users to have a virtual yellow pages. This didn’t scale, of course, as the number of Web pages exploded. Early attempt at user-updated directories were prone to severe levels of cheating and it was nearly impossible to trust ratings and/or believe that the directory was anything near complete.

The next step was the use of software to supplant directories with search engines. Bots would read pages and see what terms showed up and then spit out results for users on the search pages. Proto-SEO actors would game the system, search sites would adjust, and search engines evolved into what we have today: a reasonably fleshed-out coverage of many parts of the World Wide Web.

The good news is that as technology has improved the gap between search site perception of value and actual value has decreased. The bad news is that the gap is still relatively wide in many cases.

It is in this environment that the hypothetical Web site we’ve created finds itself. Search site placement can be very important, but sacrificing value to users in order to gather users can result in an inferior Web site.

Fortunately, the future might allow the best possible sites to exist with the most possible users. The future might have a superior search approach: one that combines technology with communal eyeballs and collective-approved expert opinion.

Earlier efforts at this sort of thing (by sites like Wikia) have fallen short, but hope is not lost. Zakta seeks to augment its search results by letting users edit, save and share the results for their search terms. A standard Google search for “best HDTVs” might have two pages of questionable or even unusable results—either because of SEO or because of the passage of time and weakness of the search algorithm. On Zakta, the same search may point to a user who has taken the time to build a top 10 list of HDTV review sites and five of his favorite online retailers. Eventually, the same search might result in a half-dozen similar lists, with the searcher able to differentiate between the quality of the lists and/or quality of the list-maker as rated by other users of the search site.

Zakta might not be the next step, but the next step is coming. It’s Wikipedia meets Digg meets Google. It encourages value over gaming and substance over style. And it might signal the end of SEO as we know it.

Drop Down Menus: An Information Architecture Conundrum

//0 Comments //Link

We’ve been working on the web redesign of a local community college for the past several months and as with any large, complex organization, the information architecture has been a bit more challenging than your average display site. Any time you try to formulate a simple, useable navigation schema while respecting the hierarchical structure of an academic organization and the sundry needs of students and faculty alike, you’re going to end up with a very deep site. Deep sites are problematic for various reasons: important pages can get lost in the bowels of the structure, navigation elements can become unwieldy as we try to design for six or seven levels of hierarchy, and the overall usability suffers and people get frustrated.

There are solutions to the deep site problem. A couple we’ve tried in the past include highlighting popular or important links in special “Quick Link” areas. Another was to create special persona-based “portals” that work alongside the deep hierarchy to promote important links to the top based on user profiles. The one solution we have always tried to avoid is the dreaded drop-down menu. In the 6 years I’ve worked here at PBDH I can’t remember a single project where we have recommended using drop-down menus. There’s been a couple projects where we were coerced into using them, but never without a good deal of “informed pushback” from us. Drop-down menus present all sorts of problems, from allowing users to skip over important pages, to removing any frame of reference,  and mostly frustration at having to keep your mouse in the elusive “sweet spot” as you scan the link offering quickly before the menu disappears again.

So a few weeks ago our community college client came by the office with the results of a pretty extensive card-sorting exercise. We all headed down to the ideation lab and began pouring over the data. As expected, we were faced with the potential for more than 6 or 7 levels of hierarchy and all sorts of varied user needs. Eventually, as we sat there pondering the white board scribbles, the D-word came up. What about drop-downs? Could they actually work here? This time around they didn’t seem all that bad. Especially when we started calling them “flyouts”. Flyout menus sound sleek and purposeful. And there is a real-world distinction: we were envisioning large, multi-column menus that required little in the way of mouse dexterity to operate.

Flyouts make sense in this situation, especially since all of our top-level section index pages were essentially devoid of any real content—they would be reduced to, well, indexes. But, if we could create a section flyout for the first 3 levels and skip the index, we could save the user a lot of time clicking through to the meaty links and offer a nice section overview at a glance. (This is not a site concerned with a sequential narrative, so “stepping” the user through the content is not a requirement.) The file structure would remain hierarchical, with a browse-based interface built on top if it, and anything beyond the third level would become side navigation.

This sounded good. But what about all the anti-drop down usability studies out there? Usability experts are always coming out against the use of drop-down menus—and heck, we were too. Was this really a good idea? So we did some research and it turns out the king of all usability gurus, Jakob Nielsen, has recently published a pretty extensive study called Mega Drop-Down Navigation Menus Work Well. (This is sort of like the Pope recommending that married couples start “seeing other people” to spice things up.) It’s a great read and there’s some good information on what to avoid when traveling down the path of the now renamed: Mega Drop-Down Menu.

From the article:

Given that regular drop-down menus are rife with usability problems, it takes a lot for me to recommend a new form of drop-down. But, as our testing videos show, mega drop-downs overcome the downsides of regular drop-downs. Thus, I can recommend one while warning against the other.

I’ll keep everyone posted on the progress and we’ll see if our assumptions about Mega Drop-Down Menus are correct.