• Register
• Login

SEO Theory - SEO Theory and Analysis Blog

Algorithm analysis, Web community relationship analysis, SEO practices and techniques, industry news, etc.

• Home
• About SEO Theory
• SEO Glossary

How competitive is a query space?

Posted by Michael Martinez on April 30, 2008 in Intermediate SEO

How competitive is a query? When SEOs want to piss on each others’ parades, the first words out of their mouths are often along the lines of: “That’s not a competitive query”. That usually means, “I don’t compete in that query so it doesn’t mean anything to me so I’m going to be a complete ass, arbitrarily dismiss it from my sight, and assume you’re easily fooled by bluffoonery”.

To me, when someone tells me “that’s not a competitive query”, I assume they mean, “I really don’t know how to determine if a query is competitive so I’m just going to disagree with you because that makes me look like I know how to do it.”

A query is competitive if at least two people are jockeying for control over the top position in the search results. Period. End of definition.

However, some queries are more competitive than others. It’s one thing to bet your best friend you can beat him in a race to the end of the street; it’s quite another thing to enter the Boston Marathon and beat thousands of people who have been training for the race, sometimes for several years.

The most competitive queries tend to attract more resources. For example, if you search on moissanite jewelry you’ll see that JC Penney ranks very well in the query (twice). Last time I checked, JC Penney was not recognized as a leader in the Moissanite industry, so why should they have such favorable treatment in the search results for “moissanite jewelry”? Not that I would necessarily attribute the obviously spammy links pointing to their site as being directed or sanctioned by JC Penney, but they are clearly benefiiting from an allocation of search optimization resources (at least 2 Web sites and a fair number of links).

Disclaimer: I am not presently contracted to consult or deliver services for anyone who competes in “Moissanite Jewelry” but I have performed consulting services for that query in the past (not for JC Penney).

We can also look at a variation on one of Jason Calacanis’ favorite queries, hotels in paris (which makes more sense to me than “paris hotels”). You get a lot of crap results in the organic listings, so the consumer should be grateful for the local search listings that Google promotes at the top of the page. Still, this is obviously a query where some people have invested time and resources in obtaining prime search results rankings, but they don’t actually have any hotels in paris.

You get better results with a more logical query like find paris hotels because, in this case, the query has not been hyperoptimized by spammy competitors. I say it’s more logical because this is the perfect query for a paris hotel search resource but the spammers have left the query to Yahoo! Travel and TripAdvisor (not that there aren’t spammy sites in this query — it’s just that the spammers aren’t overwhelming it).

Think of Web spam as falling into two categories: useless junk that is intended to send people clicking on advertising links and semi-useful affiliate link directories that would be better positioned in queries where people want to find searchable resources. Spammers don’t care how relevant their content is to the query. They just want to make a profit. JC Penney has been making profits (on and off) for a lot of years so I don’t blame them for wanting to profit from moissanite jewelry, but if I were to ask six strangers on the street about moissanite jewelry, how many of them do you think would say, “Oh, yes, I buy that at JC Penney all the time!”?

JC Penney is as closely associated with moissanite jewelry as Xenite.Org is with the Parisian hotel industry (don’t bother looking — there is NO content on Xenite about Parisian hotels). Sure, JC Penney sells moissanite jewelry but who cares? Lots of other people sell it too. Why should JC Penney be listed on the front page twice? That’s not doing the consumer any favors.

Nor are all the spammy Paris hotel search sources doing the consumer any favors unless they really do provide deep discounts, but I suspect that Expedia, Hotels.com, ORBITZ, and Priceline are probably the leading hotel discount resources on the Web. It’s a pity the consumer has to know to search for them rather than being offered their sites in listings.

This is a failing of the search technology that indexes Web content, of course, and the human limitations that prevent that technology from advancing to a point where it should “know” how to help consumers find what they are looking for. Pay-per-click advertising helps to alleviate the situation. Nonetheless, these types of queries are pretty competitive.

Contrast the Parisian hotel queries, however, with a search for news. You will agree, I am sure, that “news” is a very competitive query but that its top results are nonetheless very relevant. I’m not familiar enough with the Paris news industry to know how good the search results for “paris news” are, but searches for “atlanta news”, “houston news”, etc. return very relevant results. There’s not much incentive (if any) for spammers to optimize for local news searches, so the search engines are able to do their thing and serve pretty relevant content.

One lesson we can take away from this quick analysis is that spam is not a clear indicator of how competitive a query is. Quite often when people conclude that queries are competitive they are mislabeling highly incentivized queries — incentivized in the sense of advertising budgets being allocated across resources that spammers can leverage. Anyone who has followed local news sources in major American cities knows just how intensely competitive those markets are. The television news stations advertise on cable, in print, and over radio (and radio stations advertise on television and cable). They definitely want to be found first for “city name news”.

Another lesson we can take away from this quick analysis is that advertising budgets are not always allocated across competitive queries. We often look at search results pages and conclude that a query is competitive simply because a lot of people are buying ads in the margins. But there are competitive queries where no one pays for PPC advertising. Those queries are either less monetizable than more familiar queries or else they are undeveloped monetizable queries.

I think we can develop some rules of thumb for determining how competitive queries are, but there are no universal rules. For example, if you see a lot of pay-per-click advertising in a query, it’s clearly being monetized but it may not be profitable. If, however, you see relevant organic listings with the PPC listings then there is probably an achievable ROI for at least part of the query space.

On the other hand, relevant organic listings can still be spammy. Searchable resources or resource directories that are positioned highly in search results are not necessarily either the most relevant or the most useful content for queries. They tend to be the most heavily optimized, most strongly linked sites. Remember the Wikipedia Principle tells us that “search results tend to favor minimally acceptable content because of the prohibitive cost of identifying highly acceptable content”.

It’s necessarily the fault of the search engine that minimally acceptable content ranks first. Search optimizers work very hard to insert minimally acceptable content into the top ten for their clients. SEO clients often only want to be the resource that ranks first rather than the resource that provides the most value. Search engines are simply not equipped to ferret out the resource that provides the most value for every query.

The search technology works better for queries that don’t monetize well. There is a strong correlation between incentivization and the presence of spam or minimally acceptable content in search results. There is a weak correlation between incentivization and competitiveness in search results.

A query is competitive because two or more Web sites are vying to be ranked first. The competition becomes more intense as more competitors join the query. Competitors usually rely upon value, optimization, or brand. Value competitors are recognized and acknowledged as providing a superior resource (not necessarily a superior product or service). That is, value competitors have exceptional Web sites. Optimization competitors offer little to no exceptional value (their Web sites could easily be replaced by other competitive Web sites). Brand competitors can have crappy Web sites with little optimization but because their brands are so closely tied to the queries (either through link anchor text or natural relevance) they compete on “name”.

Our first rule of thumb can be expressed like this: “A query’s competitiveness corresponds to the number of obvious competitors appearing in the top search results”. That is, if a query is incentivized but many of the listings are not dominated by commercial content, the query is not very competitive (there is too little incentive to inspire full commercial development of the query).

Our second rule of thumb can be expressed as: “A query’s competitiveness corresponds to the number of different types of competitors appearing on the search results page.” If you have a search results page that is covered with advertising but the organic listings provide few if any advertising or affiliate sites, the competitive roles are being waged in the search results advertising.

Our third rule of thumb can be expressed as: “A query’s competitiveness corresponds to the number of brand chasers who appear in the search results.” If a query is closely tied to a brand and there are obviously incentivized relevant sites dominating the search results, the query is competitive. Or, if a query is closely tied to a brand and there are obviously relevant (but unincentivized) sites in the results, the query is competitive.

A good example for that last type of query would be aspirin, which is a former trademark that has passed into general use. Although there are obvious commercial interests associated with “aspirin” you won’t find the search results have been incentivized. Who pays for aspirin advertising on the Web? (That situation could always change, of course.) There is a “brand” quality to “aspirin” despite the fact that it’s no longer a trademark. That is, we associate “aspirin” with a very specific type of medicine. Hence, all the top search results for “aspirin” are very relevant because there are brand chasers for “aspirin” — Web sites that, through their content and/or links, deliberately associate themselves with the brand.

Other examples of a brand-dominated search include “google”, “yahoo”, and “microsoft”. Although “microsoft” produces some incentivized PPC results the organic listings are highly relevant to Microsoft. These are brand query spaces where’s relatively easy for a search engine to pick out the most relevant search results (even if the search engines are hand-editing the results, the human editors don’t have to look hard to find the relevant content).

By disassociating query spaces from the SEO community’s obvious interests in monetizing queries we can better understand how competitive queries become. If we look at why a search results page is highly relevant to the user’s intended query, or how the page is not as useful to the user as a highly relevant page would be, we can better estimate the true competitiveness of a query. We no longer have to rely on smoke and mirrors to prove that one query is more competitive than another.

This is a topic worth discussing again in the future.