Helping Jane

Something that isn't covered too often; search engine interface guidelines. You know, it really shouldn't be too difficult old chap, one unified interface meeting the needs of millions of people...

Imagine if you will, the way people conduct their searches. Jane wants to travel. She's new to the web so she visits her favorite search engine (Gobble) and types in: travel. Hmm, not bad, only 52,200,000 results (another fifty Websites were uploaded while she performed her search). Jane quickly learns to be a bit more specific: safari travel. This time she gets 432,000 results but importantly, in the continent she intends visiting. It took time to reach this point but now she types: travel visa (not bad considering 2 keywords worked before) although this time she gets 1,740,000 results and the first 30 (3 pages) have nothing to do with visas for Africa. Once more she learns within a few searches to be more specific: safari visa nets 92,400 results and the first link clicked through within the top 10 results page, yields the required information.

All right Jono, where are you going with this? Well we need to understand that a chasm exists between the potential client base (Jane), the search engine developers and thirdly the search engine optimizers (SEO's) or Website marketers. On one hand, the public want to be able to find exactly what they want when they type in: travel, some search engine developers want them to use Boolean logic: +travel +safari +bookings whilst all the search engine optimizers (SEO's) want their client's Website to pop up within the top 10 results. Do you see how difficult it is going to be to satisfy all three?

Jane doesn't want to have to learn Boolean code or be told that syntax specific semantics will serve successful searches. She just wants her information, thank you. In a way Jane is putting a lot of trust in that search engine - she requests an answer to a complex situation within one word: travel. In reality she still needs to decide on 53 countries within Africa, time of year preference (high/low season), visa & health requirements, preferred accommodation & activities, tours, budget, transport etc. She scans the top ten results hoping that those links will provide answers to her questions.

What about those top 10 results? Lets consider how few Websites have benefited from professional search engine optimization. The result is some rather disappointing top 10 or even top 30 results. Lacklustre titles: Travel Central, Walkabout Online, none of which really tell you if you are in the right territory or continent. As a SEO consultant I would prefer to see: Travel Central: specializing in travel to Africa, Australia & New Zealand.. Or Walkabout Online: down under travel to Oz.. As soon as the searcher has to click through to see if they're in the right place, that search engine's credibility and effectiveness is compromised.

How about the good old description tag? Well let's consider Google making use of it's on the page snippets or ransom notes. Here they are looking for a page word match for the term searched for. Worst-case scenario? The snippet/ransom note reads: you need the Flash 6.0 plug-in to view this page or some other mealy mouthed text dredged the page. For those search engines, which still use the description tag found in the site's code, it's crucial to write descriptions, which work in association with the title and help our searcher (Jane) choose to click through.

I would like to thank Marc L. Resnick and Rebeca Lergier of Florida International University's Industrial and Systems Engineering department for the following research data. * Denotes terms they have created. They showed how users approach the search interface with a problem and look for an answer within the results links.

Preferred input >> "problem"

Preferred output >> "1. This link has the answer you want."

Do you see the importance of the title/description that needs to be seen in relation to the "problem" or query typed? Imagine if you will, how many technical skills Jane would need to master if each search engine had its way? Think about the differences between them, how they have quirks, foibles and a preferred route. Add the vast lack of SEO input into most Websites and it's no wonder then, that WebTop's 2000 study found that 71% of search engine users are frustrated. Currently several research programs at universities and companies around the world are devoted to creating smarter (more effective) search interfaces. We know Jane might not change - but the search engines and SEO's certainly will adopt new methodology.

I was interested to read that these research programs include cognitive engineers who study various users/searchers and develop a conceptual model of their tasks. There were instances where their tasks require very specific information (Jane really should have typed more than the word travel when she wanted to go on a safari in Africa). Then there's the scenario where the user or searcher is looking for less specific information or of less quality, in this case: Jane needs to find a suitcase. No problems; within the first 10 results there are 8 which can meet her needs. In this instance the link quality can be poor (Untitled, Suitcases & Backpacks, Leather Suitcases etc. will suffice for her needs). Her suitcase criterion is far less demanding than that of the tour or travel operator.

In some instances the quality not only of the resulting links but also the site itself is a must. Credibility, reliability, accuracy, content date or age can all play a part. The fact that Jane found a Website within the top 10 results could cause her to overlook certain credibility issues within the sites themselves. In a way the top 10 position implies extra value or credibility than what some sites deserve.

In Resnick and Lergier's studies, two patterns emerge; some searchers will scan down the list of resulting links and select the one, which is the closest match in their minds. During testing, they determined a *Pre-Click Confidence (PCC) measurement - this relates to how confident the user is that the selected link will yield the required information. Jane scans the links, as soon as one exceeds her match quality threshold; whamo or clicko! They refer to this as a *satisficing constraint due to the user being satisfied with the *PCC, even though the link may not be the absolute best one. Imagine now if Jane had a lot more time to browse, how this could result in a higher *PCC threshold. Having more time could also provide her the opportunity to start fine-tuning down to specific keyword phrases we SEO's love (unique combinations and more than the usual two word options).

Of particular interest to me, is research that shows how few people actually request a further set of 10 results if the first 10 (or page) were unsatisfactory. I'm a firm believer that the high yields are all within the top 10, or at a push the top 20. Those who searched according to the full power available per search engine, are indeed in the minority (power users). To me, this highlights the importance of a solid search engine optimization strategy. Make it easy for Jane to find your site. Do that and you certainly can't miss the rare power user or advanced searcher.

Personally I would like to see more interaction between user/searcher groups, search engine developers, cognitive engineers and search engine optimizers. How about search engine wizards for those newbies? Perhaps even examples that could subtly prompt Jane to be more specific (i.e. when she typed in travel, an example box displayed a more specific approach like: travel australia, travel to europe, 21 day safari etc.). Imagine if a box displayed recent searches similar to Jane's travel africa safari - perhaps: south african tours, okavango delta safari etc.) - in other words it matched or associated similar recent click through & top 10 searches to assist her. I believe search engine developers do have a responsibility to train and educate their users. A link to their Advanced Search interface just doesn't cut it.

Last week I was explaining to a colleague the importance of ranking well, not only for the obvious or specific search terms (i.e. 21 day safari) but also for the slightly more obscure ones; (safari visa) since travel visa would produce unacceptable results and force the searcher to redefine their query. If I could help Jane find the visa information she required after two attempts (the first: travel visa, was too broad) but the second produced a first & second place hit, then certainly Jane & I were happy. For my client it meant Jane got to their site and would hopefully respond with an inquiry. I imagined her asking if her new suitcase would be safe upon the back of that elephant she wanted to ride across the Okavango Delta. Jane, don't forget your sunscreen and hold on tight.
August 16, 2002