SEO and Usability

I'm running into an interesting conversation with more regularity every week. That conversation surrounds the usability compromise that sometimes occurs when optimizing pages, functionality, content organization and linking strategies for search engines. There is theory being preached within my company that if you optimize for search engines, then you are optimizing for the user as well. I disagree. I think they are two separate sets of logic, that may in fact overlap, but are absolutely not in harmony. Until search engines accurately emulate human thought (I have my doubts this will happen very soon) then there will be compromises to the user while optimizing for search. Any thoughts?

Mark

It depends on what flavor of SEO you're talking about. There's the notion that building HTML pages in a very standards compliant way increases search engine visibility & accessibility "for free" so to speak. Better accessibility could be considered better usability.

But I'm guessing this isn't the type of SEO you're thinking about. If you could be more specific, I'm sure the list could come up with more ideas on this one.

Jeff

On Jan 3, 2008 3:24 PM, Mark Schraad mschraad at mac.com wrote: I'm running into an interesting conversation with more regularity every week. That conversation surrounds the usability compromise that sometimes occurs when optimizing pages, functionality, content organization and linking strategies for search engines. There is theory being preached within my company that if you optimize for search engines, then you are optimizing for the user as well. I disagree. I think they are two separate sets of logic, that may in fact overlap, but are absolutely not in harmony. Until search engines accurately emulate human thought (I have my doubts this will happen very soon) then there will be compromises [trim]

On 1/3/08, Mark Schraad mschraad at mac.com wrote: There is theory being preached within my company that if you optimize for search engines, then you are optimizing for the user as well. I disagree.

I agree with your disagreement, but I will add this. If you optimize for humans, you will be optimizing for the search engines as well.

Because we're talking SEO here I feel I can safely make the assumption that we're dealing with textual content. And we all know that the way humans deal with textual content on the Web is by scanning for what's relevant to them. So if we put the topic of a page or section (i.e., keyword) in the HTML title tag or within H1, H2, etc. tags, that will call it out for the human as well as the machine. It tells both the machine and the human that this particular page or section is relevant to that keyword or idea. I would even go so far as to say that SEO of this nature is an important part of user experience design as a whole... because the user experience begins with Google (or Yahoo or whatever) more often than not. If we're not optimizing for humans, we're not doing our jobs.

Optimizing for the machine, on the other hand, leads to crappy, awkward text, and if you ask me is toeing the line between blackhat and whitehat techniques.

/soapbox

F.

Machines and humans will NEVER have the same complete set of usability needs (if you will).
SEO's sole goal is to index. Human's have many goals when it comes to information and thus their mental models for how to deal with that information will shift dramatically away from the type of goals that are programmed into search engines.

-- dave

Posted from the new ixda.org
http://www.ixda.org/discuss?post=24134

On Thu, 3 Jan 2008 13:12:11, dave malouf dave at ixda.org wrote: Machines and humans will NEVER have the same complete set of usability needs (if you will).

No, not the same complete set... of course! I'm simply talking about the task of finding relevant information.

SEO's sole goal is to index. Human's have many goals when it comes to information and thus their mental models for how to deal with that information will shift dramatically away from the type of goals that are programmed into search engines.

Yes, humans have many goals about why they want a piece of information and will consume it in different ways and take different actions based upon it... but in order to do any of that they have to find it first.

I would modify your statement to say that SEO's main goal is to get stuff found. Humans, while we may have larger macro-goals, will inevitably have at least a micro-goal of "finding relevant information, " ESPECIALLY in the context of the Web. Google's macro-goal is "finding relevant information, " and to do that it does its best to mimic how humans determine relevance: a) do lots of other humans think it's relevant? b) does it mention the topic (keyword) a lot? c) does the topic (keyword) appear in places that are supposed to communicate relevance to humans? And so on...

F.

To pose a tangential question, if you have a design optimized for the user but nobody comes across it, is the design successful? Assuming that SEO and user optimization won't lead to identical designs, is there a tradeoff that can be made between a design's findability (SEO) and its user optimization? The overall effects would be measurable using whatever success measures the business is interested in - conversions, stickiness, etc.

I disagree with your assumptions.
Mental models are a core level of HOW I want to find something.

Let's take this off the web for a moment.
One of the things that DOESN'T work about Fresh Direct is that it is searched based. However when people shop for groceries they are browser based (hunting & foraging). Google can't help you there. So if I want to design a browsing system SEO is actually irrelevant, b/c "engines" are irrelevant from user mental models.

-- dave

On Jan 3, 2008 4:25 PM, Fred Beecher fbeecher at gmail.com wrote: On Thu, 3 Jan 2008 13:12:11, dave malouf dave at ixda.org wrote: Machines and humans will NEVER have the same complete set of usability needs (if you will). No, not the same complete set... of course! I'm simply talking about the task of finding relevant information. SEO's sole goal is to index. Human's have many goals when it comes to information and thus their mental models for how to deal with that information will shift dramatically away from the type of goals that are programmed into search engines. Yes, humans have many goals about why they [trim]

-- David Malouf
http://synapticburn.com/
http://ixda.org/
http://motorola.com/

Search engine optimization is one area in which I expect our progress toward standards-compliance and good structure to pay major dividends and time savings.

Increasingly, search engines are biased away from deliberate manipulations (like metatags) in favor of the actual content of the page. Semantically structured pages benefit by being very accessible to searchbots, but only if the semantics are well designed. Pages that rely too exclusively on CSS layout (you'll know them by their abundant and tags), without adequate semantic structure in the page markup, also suffer. Pages with abundant Flash and javascript will be indexed well if their essential content is described as required by the HTML/XHTML spec in effect on that page.

We used to spend inordinate amounts of time and attention on keywords in the metatags, and other SEO gimmicks; I suppose some people still do. But in the past year or two it's become clear to me that I have more success with search engines if I give each page a unique and meaningful title and assure that all headings are contextually meaningful. It helps too if the text on all pages follows the "inverted pyramid" model familiar to journalists. If you do all that, which is pretty simple, you can be reasonably sure you'll be appropriately indexed and also fairly accessible to assistive technologies (of which searchbots are only one example). All this serves the user well, too.

So I guess my bottom line position at this moment in time is that SEO now relies on semantic structure and other good design principles, and is not the separate consideration it once was. I also think that's the way it should be.

Posted from the new ixda.org
http://www.ixda.org/discuss?post=24134

Todd,

You are of course right. To design a site with no regard for how people might find it... in other words to not pay attention to optimization is just irresponsible. The problem for me is more about the culture that I work (AOL) than a realistic argument regarding SEO. Until 18 months ago we were not a true open web environment and very little attention was paid to search engines. Now, it is like finding free money. At the moment many design decisions are trumped by search potential for expanding our user base (which remarkably, is rapidly expanding). So I am always looking to fortify my defense of the user experience.

Mark

On Jan 3, 2008, at 4:28 PM, Todd Roberts wrote:

To pose a tangential question, if you have a design optimized for the user but nobody comes across it, is the design successful? Assuming that SEO and user optimization won't lead to identical designs, is there a tradeoff that can be made between a design's findability (SEO) and its user optimization? The overall effects would be measurable using whatever success measures the business is interested in - conversions, stickiness, etc.

On 1/3/08, David Malouf dave at ixda.org wrote: Let's take this off the web for a moment.

Erm. You can't take SEO off the Web. : )

One of the things that DOESN'T work about Fresh Direct is that it is searched based. However when people shop for groceries they are browser based (hunting & foraging). Google can't help you there. So if I want to design a browsing system SEO is actually irrelevant, b/c "engines" are irrelevant from user mental models.

Browsing is a completely different situation. For example, I hate trying to find movies at Netflix because I usually don't know exactly what I'm looking for. I much prefer going to a movie rental store and just wandering through the aisles until something strikes me. In that situation, recommendations and other browsing hacks just don't quite do it for me... they don't match the experience of being able to scan through a floor-to-ceiling stack of movies.

It sounds like you're looking at human tasks from a much higher perspective... yes, information finding does not necessarily imply use of the Web, but when it does, optimizing Web content for human consumption will also make it more palatable to the machines. Jeff said this much more eloquently in his comment about SEO being based on "semantic structure and other good design principles." That's really all we're talking about here.

F.

I mentioned one SEO technique as a usability problem a few weeks back (building "filler" content pages for the sole purpose of driving traffic to a web site, regardless of the value of the content itself), and recently raised another issue from our SEO: tweaking the page title to elicit more keyword matches. Essentially, the advice goes that you add your tagline to your page title, on the premise that search engines are more likely to consider that part of your page body (it's a variance of the "pad your keyword metadata tag principle" that no longer works). The problem arises when your tagline, placed at the front of the title, prevents anyone from determining what page they're on, either in Favorites or their Windows toolbar. I raised the issue as a usability concern, but we're rebuilding our site, so I plan to correct the "bug" in the new release coming soon.

So yes, I agree that SEO strategies, followed blindly, can work at cross-purposes with usability. Careful consideration and integration of both should work out well, though, as long as everyone is willing to listen to each other (heaven forbid =]).

Bryan http://www.bryanminihan.com

Original Message
From: discuss-bounces at lists.interactiondesigners.com [mailto:discuss-bounces at lists.interactiondesigners.com] On Behalf Of Mark Schraad
Sent: Thursday, January 03, 2008 3:25 PM
To: ixda Discuss
Subject: [IxDA Discuss] SEO and Usability

I'm running into an interesting conversation with more regularity every week. That conversation surrounds the usability compromise that sometimes occurs when optimizing pages, functionality, content organization and linking strategies for search engines. There is theory being preached within my company that if you optimize for search engines, then you are optimizing for the user as well. I disagree. I think they are two separate sets of logic, that may in fact overlap, but are absolutely not in harmony. Until search engines accurately emulate human thought (I have my doubts this will happen very soon) then there will be compromises to the user while optimizing for search. Any thoughts?

Mark

*Come to IxDA Interaction08 | Savannah*
February 8-10, 2008 in Savannah, GA, USA
Register today: http://interaction08.ixda.org/

Welcome to the Interaction Design Association (IxDA)! To post to this list . discuss at ixda.org
Unsubscribe .... http://www.ixda.org/unsubscribe
List Guidelines http://www.ixda.org/guidelines
List Help http://www.ixda.org/help

If you have a choice... one being a headline and the other being a page descriptor (similar words to what a person might enter in a search engine), then it makes complete sense to tag the description as H1 and not the headline. Where we often have to have the seo vs usability conversations is in regard to multiple H2 tags where there is a constant word, paired with a second word variable... as in 'used trek bikes', 'used schwinn bikes', 'used cannondale bikes', etc. Though there are ways to adjust the visual presentation so that it is not horrible to the user without going black hat.

I noticed that Patagonia hides its search keywords at the bottom of the page with a mouse over reveal. Anyone know how this does not get them into trouble?

Mark

On Jan 3, 2008, at 6:36 PM, Bryan Minihan wrote:

I mentioned one SEO technique as a usability problem a few weeks back (building "filler" content pages for the sole purpose of driving traffic to a web site, regardless of the value of the content itself), and recently raised another issue from our SEO: tweaking the page title to elicit more keyword matches. Essentially, the advice goes that you add your tagline to your page title, on the premise that search engines are more likely to consider that part of your page body (it's a variance of the "pad your keyword metadata tag principle" that [trim]

Sorry, my previous comment should have read:

... Pages that rely too exclusively on CSS layout (you'll know them by their abundant "div" and "span" tags), without adequate semantic structure in the page markup, also suffer....

Jeff

Posted from the new ixda.org
http://www.ixda.org/discuss?post=24134

If you optimize for humans, you will be optimizing for the search engines as well...

Optimizing for the machine, on the other hand, leads to crappy, awkward text, and if you ask me is toeing the line between blackhat and whitehat techniques.

What Fred says above should be the rule of thumb, accessibility for humans first lead to better SEO.

Regards,

Luis

-- This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

Hi Jeff,

In fact, abundant use of div and span tags wouldn't be the problem but substituting semantic markup by them: a span classed as "header2", instead of an H2 tag, for example.

The use of divs and spans can be a necessity depending on the graphical design.

Luis

Sorry, my previous comment should have read:

... Pages that rely too exclusively on CSS layout (you'll know them by their abundant "div" and "span" tags), without adequate semantic structure in the page markup, also suffer....

Jeff

-- This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

I'm not an SEO expert by any means, but reading all this talk about how CSS will effect SEO results really SCREAMS to me my point that designing for SEO is a flawed methodology for total success. It seems the Search Engines actually need to optimize for the real world more than we need to optimize for them. Yes, they control the critical mass, but they are also a responsive system in and of themselves. If people are unhappy with their results over time they will change their systems to make people happier.

The fact that a technology like a search engine (and a search engine that isn't even in our design domain) is something that we even consider is troubling to me to no end.

This is NOT the same thing as considering accessibility which is about protecting the rights of human beings who are not of the critical mass (majority), who need protecting by that very fact. This is about favoring technology over human beings in total which to me is beyond unacceptable.

I am now officially throwing my cap in the wing of anti-SEO.

Fortunately designing hardware and embedded software means this doesn't really effect me to much ... ; )

-- dave

Posted from the new ixda.org
http://www.ixda.org/discuss?post=24134

On Fri, 4 Jan 2008 07:29:53, dave malouf dave at ixda.org wrote: I'm not an SEO expert by any means, but reading all this talk about how CSS will effect SEO results really SCREAMS to me my point that designing for SEO is a flawed methodology for total success. It seems the Search Engines actually need to optimize for the real world more than we need to optimize for them.

I don't think anyone here is saying that SEO-focused design is good design. I certainly agree with you that it is not. What we're saying is that designing and structuring content for human relevance will have the side effect of also being highly relevant to search engines.

While the search engines aren't perfect, Google at least does a very good job of mimicking how humans determine relevance, which leads to good, relevant search results for users. If you have two Web pages that talk about the same topic but are structured differently, both humans and search engines will judge them differently. The page that has the topic in the title, subtopics in headers, and uses consistent terminology will be easier for humans to read and comprehend compared to the page that is unstructured and inconsistent. Similarly, a search engine will rank the structured & consistent page higher than the unstructured & inconsistent page. So it's happy times for everyone, human and machine alike. : )

Where the imperfections of search engines come through is where the blackhat SEO folks get their bread and butter... keyword stuffing, duplicate content, etc... Google's getting smarter about that sort of thing, though, and people are getting penalized in the rankings for pulling that kind of crap. I am very against that stuff. I wouldn't even call that SEO... hell, I'd call it cheating.

I am now officially throwing my cap in the wing of anti-SEO. Fortunately designing hardware and embedded software means this doesn't really effect me to much ... ; )

Heh... This is more a Web IA problem than a pure IxD problem, so that's definitely true. : )

- Fred

Search engines are just one way for retrieval and indexing information, this the problem with SEO: it has a very limited technological scope based on a single model of information retrieval. A bit like a cancer for the cure, SEO makes pages easier to rank in search engine results but then nobody said that the page was relevant for me in first page. There's an aspect of SEO which is very interesting indeed: keyword searching, taxonomy and linguistic trends, but few are the SEO's who even understand what a taxonomy is all about.

Many times the purpose of SEO is to put the page in the first positions so that more people come to a page where they can click on the paid ads for other pages: a bit like the salmon lifecycle, which goes all the way against the stream so that they can lay eggs, die, then when everything is washed downstream, eggs hatch, baby fish eats mum and dad's flesh and when they are adults they go all the way against the stream again for no purpose in life but populate the rivers.

The success rate of most SEO campaigns out there are measured by the position of the web page in the results page, which definitely produces more traffic (another old model: traffic). It is a surge tactics, which might work for the owner of the site, but from the user perspective, if you are lucky to get a page with good content, good for you.

Personally, I don't use Google anymore, I go straight to Wikipedia and then go to the links they have there. I also use Digg, which seems to me the emerging model for search, something SEO hasn't thought of yet, after all SEO is aimed for short-term results.

Nevertheless, there are serious SEO consultancies out there, it is hard to spot them because there are too many failed web designers waving the SEO banner.

Cheers,

Luis

"I'm not an SEO expert by any means, but reading all this talk about how CSS will effect SEO results really SCREAMS to me my point that designing for SEO is a flawed methodology for total success..."

-- dave

-- This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

The users are the same. From my point of
view searching and reading information
is part of a human internet experience

On 1/4/08, Fred Beecher fbeecher at gmail.com wrote: On Fri, 4 Jan 2008 07:29:53, dave malouf dave at ixda.org wrote: I'm not an SEO expert by any means, but reading all this talk about how CSS will effect SEO results really SCREAMS to me my point that designing for SEO is a flawed methodology for total success. It seems the Search Engines actually need to optimize for the real world more than we need to optimize for them. I don't think anyone here is saying that SEO-focused design is good design. I certainly agree with you that it is not. What we're saying [trim]

-- Rafa López Callejón
http://www.nativos-digitales.net

Luis said:
'In fact, abundant use of div and span tags wouldn't be the problem but substituting semantic markup by them: a span classed as "header2", instead of an H2 tag, for example.'

Yes, Luis, I think you've said it better than I did. That's exactly the sort of thing I meant.

I wholeheartedly agree with Fred about structure and consistency. Well-structured semantic markup yields huge benefits in searching as well as presentation, and all of this relates to faster page loads, better accessibility and a more positive user experience.

Posted from the new ixda.org
http://www.ixda.org/discuss?post=24134

I have to weigh in as a usability specialist who is sympathetic towards SEO (white-hat, at least). I've been digging deep in to the SEO world over the past year and have begun to see more and more convergence with usability as search spiders become more sophisticated.

While I agree that building solely for SEO can compromise usability, that's true of any narrow-minded focus. Building solely for cutting-edge design or your marketing department also frequently compromises usability. On the other hand, thoughtful SEO, and the architecture it requires, can often be good for human users as well.

In addition, I think it's very important to usability that we start to consider the entire life-cycle of the consumer, from the moment they sit down at the computer. If a person types a query into Google and gets a link that represents my site badly and then takes them to a completely disconnected page, I've done a bad job of creating a user experience. Understanding SEO is an important part of that pre-site user experience.

Posted from the new ixda.org
http://www.ixda.org/discuss?post=24134

On 5 Jan 2008, at 07:10, Peter Meyers wrote: In addition, I think it's very important to usability that we start to consider the entire life-cycle of the consumer, from the moment they sit down at the computer.

Over the last couple of years I've noticed an upward trend among usability test and ethnographical study participants to use Google rather than a particular site's navigation, e.g. instead of looking for Flight Times in the left hand nav of the Virgin Atlantic site, they prefer to Google "virgin atlantic flight times". This is similar to the increase in users using Google for navigating to popular sites rather than typing in the URL or using favourites/bookmarks as evidenced in the increase of search terms such as Facebook, MySpace etc.

This behaviour relies on sites having accurate SEO and may indicate an increasing importance of SEO to provide an acceptable level of usability.

Diarmad McNally
Interaction Design Studio

UK: +44 (0) 7808 297289
Irl: + 353 (0) 85 7888 085

It also may belie user dissatisfaction with the navigation systems of sites they really want to use, in spite of getting lost or not finding what they may even know is there.

I know for years, on many newspaper and major media source sites like CNN.com or MSNBC, I'd try 3-4 times to find an article I'd read before or saw in print and wanted to forward to someone, only to become so vexed and thwarted by the godawful CMS's and lack of permalinks or ability to search all terms, including author, or article text, that I'd just bug out and go to Google, and inevitably, find the article I couldn't locate on the site from the site itself, to save my soul.

What if the real reason people do it is interface frustration?

Chris

On Jan 28, 2008 3:42 AM, Diarmad McNally diarmad at ixdstudio.com wrote:

On 5 Jan 2008, at 07:10, Peter Meyers wrote: In addition, I think it's very important to usability that we start to consider the entire life-cycle of the consumer, from the moment they sit down at the computer. Over the last couple of years I've noticed an upward trend among usability test and ethnographical study participants to use Google rather than a particular site's navigation, e.g. instead of looking for Flight Times in the left hand nav of the Virgin Atlantic site, they prefer to Google "virgin atlantic flight times". This is similar to the increase in [trim]

On 1 Feb 2008, at 03:20, Christine Boese wrote:

It also may belie user dissatisfaction with the navigation systems of sites they really want to use, in spite of getting lost or not finding what they may even know is there. [snip] What if the real reason people do it is interface frustration?

True. I had exactly this sort of problem recently where Google found a page I couldn't find with the standard site navigation.

However, what I've been seeing is that with some users this behaviour becomes generalised. Since there are lots of bad sites out there, and searching for things with Google works, it becomes the weapon of choice, rather than the last resort.

Like Diarmad I've seen more folk use Google to navigate sites. I find it quite interesting since they usually end up completing the task - just not always using the site being tested :-)

Just one more nail in the coffin of "web site" as a useful concept.

Cheers,

Adrian

Take a look at:

http://www.google.com/intl/en/press/zeitgeist2007/

where Google gives an indication of popular queries for 2007. Among the top ones you will see queries such as "myspace", "facebook", "youtube", "tmz", "dailymotion", "badoo" etc. (I have even heard of a sizable amount of queries for the word "google" ).

All the above are names of website which would have been accessed by typing for example www.tmz.com.
So there is strong indication that people are using google a "shortcut" engine as well as a search engine. Not sure if all these queries come from people typing in the browser search box or if they have google as their home page and type it in the standard google text box.

But I definitely think there's scope for a better shortcut/history mechanism in browsers.

Cheers,
Alex

On Feb 1, 2008 3:43 AM, Dr. Peter J. Meyers peter at usereffect.com wrote: I think you're probably right about that. I've also noticed a tendency for users to get stuck in a pattern. They visit a site the first time by going through a certain chain of actions on Google, and from then on out they continue to visit the site via that chain. Sometimes, this borders on the bizarre, including browsing through multiple pages of search results and/or clicking on PPC ads (very annoying if you're the one paying for those ad clicks). People are strange, but I guess that's what keeps us employed : ) - Pete On Jan [trim]

It is really interesting to think about this in terms of global interaction structure. Search engines are really a layman's command line interface for the web. Very direct access to information, if you know the right syntax. This is as compared to the hierarchical menu structure (GUI-driven) thathas been shown to be the easiest to learn (assuming that the you have the categorization of information correct).

The improvment in the search engine's tolerance for "faulty" syntax, coupled with the high accuracy of the search results really eliminates some of the faults that traditionally are associated with command-line interfaces. And what you are left with is the speed and efficiency of direct access vs. the intuitiveness of a hierarchical menu structure.

Are the bulk of web users becoming "expert" users in which they are looking for the power and efficiency of a command line interface? Is the baseline level of web-user sophistication growing to the point that power, speed, and flexibility (at least in navigation) are as important as rock solid intuitiveness (not to be confused with ease of use)?

Posted from the new ixda.org
http://www.ixda.org/discuss?post=24134

I am one of these users, but with a twist: I use a bookmark keyword (Firefox) to "I'm Feeling Lucky" of Google:

http://www.google.com/search?q=%s&btnI=I'm+Feeling+Lucky

This allows me to type "l ixda", and voilà, I'm on http://www.ixda.org/!

-- Claude

On Feb 1, 2008 1:36 PM, Alexander Baxevanis alex.baxevanis at gmail.com wrote: Take a look at: http://www.google.com/intl/en/press/zeitgeist2007/ where Google gives an indication of popular queries for 2007. Among the top ones you will see queries such as "myspace", "facebook", "youtube", "tmz", "dailymotion", "badoo" etc. (I have even heard of a sizable amount of queries for the word "google" ). All the above are names of website which would have been accessed by typing for example www.tmz.com. So there is strong indication that people are using google a "shortcut" engine as well as a search engine. Not sure if all these queries come from people typing in the browser search [trim]

I think one of the reasons people use Google as a shortcut to find things on other sites is that it drastically lowers the learning and memory costs. It's one technique that works the same across all sites, whether it's a shopping site like buy.com or a news site like cnn.com. Each site has a different (often radically different) way of sorting and presenting its information. From the user's point of view the choice is either to learn and try to remember each site's way of doing things - the varied IAs - or just learn one way and use it everywhere.

I think it's a completely rational and understandable decision for users to pick option # 2.

--Alan

Sorry for joining in on this late but I'm wondering what you folks think of eliminating browsable navigation on Web sites all together and just forcing users to use a search interface to locate what they are looking for. Songza ( http://www.songza.com ) is an example of this that does not allow users to browse, for example, a category such as Rock music.

I've always been of the mindset that we should provide for different user habits but if the majority of users are moving towards search only, then perhaps my assumption should be re-evaluated. It makes me a bit sad to think that serendipity may be eventually lost.

Any thoughts?

Stephanie Walker
Information Architect
Austin, TX, USA

There are fundamental reasons that search of publicly-available specific information works better than pre-built structure.

Setting up site navigation involves choices:
- Which items to make visible and when
- What to call the items

Search cuts across both of these:
- If the searcher gives priority to a lower-level category, the search will match when step-by-step navigation would hit a hiccup - If the searcher chooses a different name from the architect, the content may match anyway

The times when search does not win are:
- When privileges are required to make items visible, and the search engine isn't granted the same privileges as the user
- When multiple distinct items are called by the same name

This first factor explains why search within an e-mail archive is a killer app. The search engine in your e-mail has your privileges, so anything you can get e-mailed to yourself is searchable. If you are able to distinguish among the items in your e-mail, then they become findable too.

Regarding serendipity, there are three phases to search: - Specifying criteria (and later broadening them based on actual or anticipated search results)
- Narrowing the criteria (based on actual search results) - Selecting an item from among the search results

Perhaps what you are looking for is there, but in a different way than you expect. Is there not serendipity even in filtering? That combined with idiosyncratic links within content can give us the appropriate surprises that we crave.

Best wishes,

Bruce Esrig

On Wed, Feb 20, 2008 at 5:58 PM, stephanie . ennea999 at gmail.com wrote:

Sorry for joining in on this late but I'm wondering what you folks think of eliminating browsable navigation on Web sites all together and just forcing users to use a search interface to locate what they are looking for. Songza ( http://www.songza.com ) is an example of this that does not allow users to browse, for example, a category such as Rock music. I've always been of the mindset that we should provide for different user habits but if the majority of users are moving towards search only, then perhaps my assumption should be re-evaluated. It makes [trim]

YAY! I was hoping that I would be able to catch up to this conversation. For me, it is about control. In navigating to desired information, the IA or whomever constructed the site, controls the experience with the customer making step-by-step choices based on what the site chooses to make available. With search, the user "control" the experience, plugging their oftentimes ambiguous terms into a box and getting a list of seemingly appropriate results right back. Google further simplified this process by taking away a lot of the programmatic arcana (Boolean bugaboo that few were able to use effectively) and now the "Google experience" drive search UX.

All of that typed, Jared Spool will tell us (fingers crossed at my end) that customers are more successful if they navigate through the information space than if they trust the information seeking equivalent of crack cocaine usually found in the upper right corner of a site's masthead. I believe that this is because customers don't know what they don't know at the outset of their search. As the customer navigates through an information space, their information need become contextualized within that space and clearer enabling them to make more effective choices and ultimately resolve their need. That's the Disney ending at least.

Web search used to work that way when Northern Lights and Alta Vista presented their cornucopias and folks would click around and make discoveries and figure out that maybe they were looking for the wrong thing or they gave up on expecting a machine to understand their need and asked a fellow thought processing biped (hopefully a reference librarian because they totally ROCK!). For me, serendipity is getting lucky with the I Feel Lucky button, using search engines to help me find something that I know is there (i.e. the IxDA website) and will include the search engine that takes my flailing around and makes sense of it by presenting results that it "thinks" I will be interested in based on what I've told it so far. I believe that this is closer than we might think.

As far as I'm concerned, the only time search doesn't work is when we, the thought processing bipeds, do not avail ourselves of every opportunity presented to describe our content to the machine. Search is far from perfect. It is however extremely complex and robust and a terrific tool, not thoughtful and terrific nonetheless. As for its flaws? I believe that they lie..."Not in our stars but in ourselves."

marianne
msweeny at speakeasy.net

Original Message From: discuss-bounces at lists.interactiondesigners.com [mailto:discuss-bounces at lists.interactiondesigners.com] On Behalf Of Bruce Esrig Sent: Thursday, February 21, 2008 2:04 AM To: stephanie .; discuss at ixda.org Subject: Re: [IxDA Discuss] SEO and Usability There are fundamental reasons that search of publicly-available specific information works better than pre-built structure. Setting up site navigation involves choices: - Which items to make visible and when - What to call the items Search cuts across both of these: - If the searcher gives priority to a lower-level category, the search will match when [trim]

I just visited Songza and was completely stymied. After clicking around bit, I had that sinking sensation of "Oh. I get it, that's what they're trying to do here." And the only reason I even got that far was because I was in Investigation Mode. Had I stumbled across this site on my own, I never would have guessed that all the functionality offered by Songza actually existed.

They're borrowing a bunch from Google, which is fine, but they may have actually gone overboard with the whole simplification thing. Also, several of the interactive controls are so cutesy/clever that I'm way too conscious of the hand of the designer, something that was rampant with the Flashturbation sites of the 90's. As a result, I'm left wondering who the site is for. Is this one of those "if they're not smart enough to get it, they shouldn't be here" sites?

Overall, the sense I get from Songza is that a refreshingly forward-thinking businessperson got together with a really talented designer and they went for it. Great! Unfortunately it also feels like a seasoned info architect was missing from the mix...someone who would have pointed out that however "uncool" and irritating it may be, there are still some very real principals about users that need to be accounted for...such as non-directed search.