The Relationship Between Search Behaviour and SERP Landscape (Part 2)

This is the second and final part of a 2 part series on the relationship between search behavior and SERP landscape. Part 1 is here.


The seeds of this article arose when I was studying differences in searcher behavior between countries.

I noticed that in many countries, especially those who are relatively new to the internet being widely available, searchers seem to show, in general, lower stages of search behavior, often progressing no further than Stage 2 - Exploration.

At first, I assumed that this less sophisticated behavior was due to less experience with the internet, an assumption I now know to be false and, to a degree, ethnocentric.

The assumption that searchers in these countries are less sophisticated due to a lack of experience can be supported by the massive growth rate (500% or higher is common) of internet availability, but does not explain why even people who have had the internet for years in these countries still tend to show less sophisticated search behavior as a whole than searchers in developed countries.

There is no "national searcher" - everyone progresses by themselves as an individual due to their own experiences. Additionally, in many of these countries (typically developing countries) people who have internet tend to be well educated and spend a lot of time online - they should progress fairly quickly, yet often do not. So the question is - what's going on?

I believe the answer lies in the SERP landscape, and it's been hidden there due to the halo effect of the online industry as a whole.

You see, my research appears to show that searchers everywhere will naturally try to progress through the stages of search sophistication as they use search more and more.

However, if the SERPs do not support that progression, the progression stops. In short, increases in the sophistication of tool use is limited by the sophistication of the tool itself. Some tools simply lend themselves to more sophisticated actions than others.

How could this happen? Many ways.

Sometimes, it's due to limitations of the search engines themselves. Back in the days before Google, searcher behavior was also fairly unsophisticated, since attempts at sophisticated searches were generally foiled by small indexes filled with spam. It simply wasn't worth the effort to try to improve your search techniques when the search engine still gave you poor results.

Another reason this could happen is due to the available results. If nothing but spam is available, then you will get nothing but spam. If local websites are designed in a manner that is not search friendly, then no matter how good you are at searching, you are still unlikely to find what you need, because the search engine simply doesn't have it to give you. It may exist, but it's not available through a search engine.

GIGO - Garbage In, Garbage Out.

It is this second reason that appears to be why searcher behavior tends to "stall" in some countries and languages.

Although part of the blame sometimes lies on the search engines themselves for not adapting to local restrictions, culture and resources, the majority of the blame for this is, very simply, poor websites from an SEO (and often usability) perspective.

As websites for the region or language become more sophisticated and accessible to search engines (and start carrying useful information rather than being just web brochures), the search engines will begin to be able to provide a better sampling of results, thus allowing more advanced searcher behavior.

Analogy - Buying some Beans

As an analogy, think of going shopping to go get a can of beans at two different stores. The first store is a well run national chain with a huge selection and a logical, clear layout. Since you have been there many times before, and you already know exactly what you want, you will probably know exactly what to do, even if you have never bought beans before at that store.

You will go to the aisle that is most likely to contain the beans, glance quickly down it for the area most likely to contain the beans and head right there. On the way, you have probably already checked for coupons and sales, and have a very good idea of exactly what brand and size you want, along with what brands you may not want. You may also know whether you want to get the beans from the "All Natural" aisle, the "Ethnic Foods" aisle, or the general "Soups and Canned Goods" aisle.

Within seconds, you have exactly what you want, after following what is, if you think about it, a very sophisticated and effective search pattern, that was helped by a large, well-organized selection to choose from. This is Stage 4 - Control mode. You take control of your search and make it work for you.

Next, contrast this with going to get a can of beans from a nearby store that is very messy, disorganized and has little stock. First, forget coupons, the "All Natural" aisle and all of that. If you are lucky, there might actually be a "groceries" aisle. Your can will be there, in with the other cans, if you are lucky. If there is an organization to it, it's not readily apparent.

At this point, you have to throw planning and sophistication out and basically just start hunting through the shelves until you find a can of beans. The chances of it being exactly what you were looking for are remote, so you may then end up sifting through yet more cans in a vain hope that there might be another choice (hopefully one that has not expired). You may even decide to give up and either skip the beans altogether, or go check a different store. You are in Stage 2 - the Exploration mode.

In this scenario, it's not YOU who has changed and become less sophisticated, it's the shopping environment that has. You basically had to degrade your planning and shopping behavior to deal with the fact that sophisticated actions can only take place at the top end of the available actions, and the available actions are reliant upon the choices, support and quality present at the time.


In short, searchers in China, Mexico and elsewhere are only searching in unsophisticated manners because the SERPs themselves are unsophisticated, not because of some sort of cultural norm, which is often currently the assumption.

I hear this all the time: "The Chinese search like this" or "Mexicans tend to do searches this way", but this is ethnocentric and misleading.

It would be more accurate to say that "searchers in China do this" or "searchers in Mexico have to search this way". The sophistication of the searches are based on the sophistication of the search landscape, not the searchers themselves.

There are 2 major conclusions of interest to the search community that can be taken away from this, IMO.

  1. If a market shows unsophisticated searcher behavior (as evidenced by the types of searches performed), then there is almost certainly an excellent potential market for SEO (along with an attendant lack of awareness of SEO in the first place). Additionally, due to this lack of sophistication within the market, SEO and PPC are likely to very effective. There is simply less high-quality competition.

  2. Since searchers will increase in sophistication as the available search landscape evolves, it is important to prepare websites for more sophisticated searches (ie long tail terms, searches for specific on-site information rather than just contact information, etc) rather than simply rely on current KW research. You will need to evolve within your market as your searchers do. And they will, as quickly as the market does.
Hopefully these 2 articles have been helpful to you. I know the insights I've taken from coming up with them are serving me well with my current international SEO efforts.


The Relationship Between Search Behaviour and SERP Landscape (Part 1).


The following is original research developed from several hundred interviews across 9 countries (Canada, USA, Mexico, Brazil, China, Japan, Korea, England and France), combined with insights gleaned from both formal training in cultural anthropology and HCI (Human-Computer Interaction - aka usability). That said, I could be totally wrong. But I don't think so.

I was explaining this theory to a businesswoman in Brazil a couple of days ago and it occurred to me that I hadn't written about it yet, nor had I actually shared any of this with any of my SEO colleagues - forgive me, I'm writing a book on international SEO and as a result my blogging has been slow lately.

Here is the insight: People search differently at different stages in their comfort levels and experience with search engines, but progression through these stages are in turn affected by the SERP landscape they are provided.

This is most prominently (though not exclusively) visible within intranational searches, but this behavior can also be seen in any area where the SERP landscape is different from "standard", such as mobile search, image search, local search, and so on.

Searcher Behavior Stages

The speed of progression through these stages can vary based on how often the searcher searches (dozens of times per day vs a few times per month), the topics for which they search (each major topic will have it's own progression), and whether they get help or training (or have previous experience).

But this is the general progression of search stages:

  1. Trust. A searcher at this stage is a novice, and has no idea what to expect as far as results for their search. In general, they will click on the first link that is not clearly spam or inappropriate (some novices to computers in general as well as search will not even attempt to look for spam or inappropriate results - they will simply trust the search engine to answer their question). If they don't find what they are looking for, they tend to blame themselves for making a bad search or assume the information is not available.

    Searches tend to be simple and general ("taxi" or "how to babysit"). Repeat searches may get slightly more complex (ie the addition of a location - usually at the end of the search as an addition to the original) but not too much more complex - they tend to assume the search engine is much smarter than they are and will figure it out. Users at this level often do not realize the difference between PPC ads and organic results (or don't care).

  2. Exploration. Fairly quickly, searchers tend to progress to the next stage (since the first tends to be unfulfilling), which is exploration. At this stage, there are several ways it will play out, but the most common is the "serial-clicker" - someone who goes though every (or almost every) result in the desperate hope that one of the results will be an answer. Another common scenario is the searcher, rather than going back to the original search results, will begin to click on links within the site they landed on, surfing from page to page and site to site. MFA sites make a lot of money based on this behavior. This is a very important stage to remember, and I'll tell you why in Part 2. These searchers may go several levels deep in the results (page 3 and beyond).

    At this point the user is still trusting of the search engine, so search queries tend to be very similar to the trust stage. Instead, the searcher is modifying their own behavior (still assuming the problem is with themselves or with the available data) by clicking differently and exploring the results to try to find their answers.

  3. Analysis. At this stage, the searcher starts getting smarter and more experienced. They have come to realize that clicking on more links isn't really the answer. At this point, depending on their personality, the results they have seen so far, and other criteria, they will begin to change their tactics. Some of the tactics they may try include one or more of the following:

    • looking at the results page for likely candidate sites before clicking on any (aka "sniper" mode)
    • trying other search engines
    • beginning to use more sophisticated searches and planning ahead (ie putting the location first, then the query)
    • Figuring out the difference between PPC and organic listings (and tending to avoid PPC)
    • Finding and re-using "tried and true" search patterns (like "X reviews" "X FAQ" or "X wiki"

  4. Control. At this stage, the searcher becomes sophisticated and takes control over their searches. They realize that the results a search engine provide are in part controlled by the search query itself, as well as the results available. This is usually the level most people eventually find themselves at.

    At this stage, advanced search tactics are used, such as:

    • tiered searches (searching in a general manner, then using information gleaned from those results to perform the "real search" using the information and keywords from the previous search - like looking up the wikipedia entry for a topic, then using keywords and ideas from that to perform a second, "real" search)
    • searches based on likely content or title of a desired result, rather than the user's question
    • long tail searches become more prevalent
    • Simple parameters such as quotes or "results from this country" are more likely to be used
    • Actively trying to prevent bad or off-topic results by using negative parameters or less ambiguous terms.

  5. Expert. Most searchers do not reach this stage, as it requires study and is more difficult to do than the results are generally worth for most searches. Experts will use advanced search parameters, tiered searches, and other advanced techniques that require a good knowledge of search engine behavior. This category includes information and search professionals (SEO's, researchers, topic experts, advanced students).

    Searches are planned out, often "long tail" or tiered, and can include advanced and multiple parameters. PPC ads, often avoided at the control and analysis stages, will begin to be clicked on if they appear to answer the query. Experts are after the best result, and don't usually care how they get it. They will intelligently make exceptions to general rules of good searching if they believe the result will be good.

Well, that's my list of searcher behavior stages.

The next thing to realize is that these stages may repeat themselves for different queries or topics.

For example, someone may search at a Control or even Expert stage for a topic related to their work or hobby, but when confronted with something totally new (like planning a wedding) is likely to go through the stages again, though usually at a much faster rate than a new searcher would (sometimes in a few hours or less).

They know nothing about the topic, so they need to start by trusting the search engine again.

While the above information is useful in and of itself, in Part 2, I'll go over how people (and entire cultures/nations) can get "stuck" at certain stages, and the effect this has on international SEO and SEM.

Stay tuned.


Keyword Research Schedule

I was doing some keyword research this morning and was reminded of a very important issue that I'd like to share.

It's February, and I just did some KW research related to Calgary (my home town,and a place I'm very familiar with, as far as KW research goes).

Services like KeywordDiscovery and Wordtracker typically default to showing what's in their database for the last few months (usually around 3 months or so).

That means that right now, I'm seeing KW's like "calgary valentine sale" and almost no searches for what I *know* is one of the biggest KW's for the city - the Calgary Stampede.

If I didn't know better, and used this KW research, I might come to the conclusion that the Stampede wasn't that big of a deal. But it is, and KW research done in August will clearly show it.

This seasonality especially affects seasonal things like tourism, tax preparers and so on, but keep in mind that it will have an effect on almost all industries, either directly or indirectly. For example, the end of summer is when a lot of people move to go to school. This results in more searches for anything from pots and pans to computers to local nightclubs as the new residents start to settle in.

As a result, I recommend doing KW research at least once per quarter (ie every 3 months) and to aggregate the data at least once per year.

Aggregation should be the following reports (keep them in a binder or in a folder somewhere):

  1. Quarterly KW Reports, saved by year.
  2. Yearly aggregate reports of the quarterly reports for the year, also saved by year.
  3. Quarterly Reports, aggregated and *trended by KW* from the same quarter for each successive year.
  4. Yearly trended reports of KW popularity. This is so the temporary huge popularity of a topic isn't directly affecting your decision making. "Pam Anderson" was an extremely popular KW back in the early 2000's, but not so much today, for example.

This will typically give you enough good information to act on, yet not be terribly onerous to create, analyze and report on.



The new thing all the "cool" social media spammers are doing now is impersonating famous people on twitter. They sign up for a name that is very similar to a well known person (ie "mattcutts" becomes "mattcutts_" and then sets their screen name to the exact name of the person they are impersonating: "Matt Cutts". I use Matt as an example here because it happened to him recently by a particularly clueless twitspammer.

They then "follow" a few thousand people, most of which are honored to have such a famous person following them, and thus go visit the person's profile to follow back. At this point you'll typically see only one or two posts, always to some sort of spam. Ironically, the one I've seen most is a link to a "whitepaper" on how to market your crap in Twitter.

So they spam you on Twitter in order to promote a paper on how to spam people on Twitter. Ugh.

To their credit, Twitter is usually pretty good at finding and removing these, but they are usually up for several hours and get a few thousand visits before they go down - enough, apparently, for some twitspammers to feel it's worth it.

The really scary thing (to me) is how many people fall for this, and start following back. Yikes! Hey, I've got $78Million I stole from the Nigerian government that I'd like you to help me get out - how about it? You can by that swampland in Florida, the Brooklyn Bridge, and still have some left over to invest in Bre-X!


ADDED: Andrew Girdwood posted about this earlier, as well.