In order to gain a quantitative understanding of the extent to which the number of words in a keyword phrase (that is queried in a search engine) impacts the number of search referrals a website receives from the natural (or “organic”) search results produced by that query, iProspect analyzed three years of keyword and referral data from its SEMCARD (Search Engine Marketing Campaign Aggregated Results Data) data warehouse to create the iProspect Natural SEO Keyword Length Study.
Built by iProspect, SEMCARD is a data warehouse that contains five years of search engine rankings and traffic data for most of iProspect’s clients, across all the keywords which produced traffic for those clients, across all the major organic search properties.
Methodology
The iProspect Natural SEO Keyword Length Study is based upon natural (or “organic”) search referral data collected from a sampling of iProspect's clients between 2002 and 2004 (amount of data prior to 2002 was not statistically relevant). The number of websites whose data was aggregated for this study ranged from 39 in 2002 to 44 in 2004. The client companies represented a variety of industries, including B2B, retail, pharmaceutical, automotive, travel, and financial services.
Most of these clients were Fortune 1000-level companies, with websites that can be characterized as “high traffic.” Hence, the conclusions drawn in this study should only be applied only to high-traffic sites at Fortune 1000-level companies.
In order to protect the confidentiality of our clients, iProspect can neither disclose their identities, nor the true numbers. Therefore, all the metrics reflect combined, aggregated data, and have been expressed as percentages.
Disclaimer
Percentages expressed in this report are rounded to the nearest one percent, so where a percentage is expressed as zero, its real value is in fact less than one-half of one percent.
Findings from this study apply only to natural (or “organic”) search referrals and the keywords that drive them – NOT to words targeted by pay per click advertising or the traffic produced by that channel.
Key Findings
Search referrals produced by querying one-word keywords still account for a healthy percentage of overall search referrals, and are not going away any time soon.
Though experiencing a decline over the last 2 years, single-word keyword searches still account for 39% of natural search referrals, and so should continue to be targeted by your natural search engine optimization efforts.
One- and two-word keyword queries combined account for the vast majority of search referrals.
A full 88% of search referrals are generated by searches for one- and two-word keywords. Because shorter keywords account for such a high percentage of traffic, they must continue to be targeted by your search engine optimization campaign.
There is a definite growth trend in two-, three-, and four-word keyword phrase search referrals combined.
Over the past three years, referrals generated by multi-word search referrals have increased by 12% of total referrals, and as the percentage of referrals which come from longer keyword phrases continues to grow, it will be more and more important to target this higher converting category of keywords with natural search engine optimization.
Detailed Findings
Search referrals produced by querying one-word keywords still account for a healthy percentage of overall search referrals, and are not going away any time soon.
As you can see from the chart below, over the past two years, the number of search referrals produced by 1-word keyword searches has declined (12%). But despite that decline, single-word queries currently generate only 10% fewer referrals than 2-word keyword phrases (which produce the highest percentage of referrals).
Contrary to a number of industry reports that describe the steep erosion of one-word keyword searches (which haven’t always addressed the number of actual search referrals (clicks) generated by those searches) and the growing sophistication of search engine users toward multi-word queries, data from SEMCARD indicates that though the number of referrals from single-word queries is certainly decreasing, they still account for nearly four out of ten natural search result referrals. In addition, the rate of decline between 2002 and 2004 has been just 6% of total referrals per year.
Potential reasons for this include: an increasing number of new Internet users who are not yet search-engine-savvy – who still search on single-word keywords and click on their results; the number of searches on single-word branded keywords – especially in the case of Fortune 1000-level brands; and the number of referrals garnered by big brands on unbranded single-word keywords because the user recognized the big brand’s name in the search results – so select it over lesser-known brands.
Regardless of the reason, the implication from a natural search engine optimization perspective is clear: big brands still need to target branded and unbranded one-word keywords that apply to their products or services. Though website visitors who arrive as a result of searching on multi-word keyword phrases often convert at higher rates than those who searched on one-word keywords, 39% of all referrals still come from one-word keywords. So failing to target those keywords means ceding the conversions that come from that substantial 39% of search referral traffic.
And big brands, by their very definition, have an advantage when it comes to generating referrals and conversions from these single-word keywords due to the searching public’s comfort, and familiarity with their brand.
One- and two-word keyword queries combined account for the vast majority of search referrals.
As you can see from the chart below, almost half of all search referrals are currently being produced by 2-word queries (49%), with one- and two-word queries accounting for almost nine out of ten search referrals (88%). This percentage has remained pretty flat over the past three years (88% in 2002 – 91% in 2003 – 88% in 2004).
Despite much of the buzz in the industry about the rise in the number of three- and four-word keyword phrases, data from SEMCARD shows that these numbers have fluctuated a bit, but have not, in fact, shown any considerable growth over the last three years.
The continued strength of two-word keyword phrase search referrals – especially for well-known Fortune 1000-level brands – may be partially explained by the volume of searches that combine a brand name and a product name (e.g., IMB laptop, IBM ThinkPad, etc.) that ultimately result in a search referral (click) to a website. And many of these searches and resulting clicks may, in fact, be taking place after an initial one-word keyword search that returns too general a results set – from which the user does not click on a result.
Industry data that demonstrates a trend toward higher conversion rates on longer keyword phrases – perhaps due to a higher degree of keyword specificity the closer the Internet user gets to the purchase (conversion) decision – is not being questioned here, but the data from SEMCARD shows that failing to focus optimization efforts on shorter keyword phrases ignores the categories that are currently generating almost 90% of search referrals across “natural referral” data.
Even at a lower conversion rate, this quantity traffic cannot be ignored – as its overwhelming volume may very well generate more conversions in real numbers than the much higher converting three-, four- and five-word referrals. And it is important to recognize that once attained, a natural search engine ranking on a single-word keyword target has no incremental cost associated with it over time – which is unfortunately not the case with pay per click advertising campaigns.
There is a definite growth trend in two-, three-, and four-word keyword phrase search referrals combined.
Referencing the same statistics we’ve been using thus far, you’ll see that over the last three years search referrals generated by multi-word queries have increased by 12% of total referrals, while single-word search referrals have, by definition, decreased by the same percentage. The percentage of 4-word search referrals has actually tripled just over the past year.
As discussed earlier, despite the relatively small percentage of search referrals generated by keyword phrase searches of 3 words or longer, these longer keyword phrases typically convert at a higher rate than their shorter counterparts. We also intuitively know that there are far more possible combinations of words that can produce three- and four-word keyword phrases than there are of one- and two-word phrases. So targeting as many three- and four-word keyword phrases as possible, at that higher conversion rate, is an important strategy in generating as many referrals as possible from this keyword category.
As noted previously, the effectiveness of this strategy comes precisely from how natural search engine optimization works, where – unlike pay per click advertising – there is no additional incremental cost associated with targeting additional keywords. As long as site content supports a large number of multi-word keyword targets and site optimization elements, and supporting in-bound links can be provided, you can test and measure as many multi-word terms as time allows – at no incremental cost – until you find those that generate the highest conversions for the site.
These multi-word keywords, used in combination with the single-word keywords that produce the highest volume of conversions in real numbers, appear to be growing in importance each year, and if the trend continues, important decisions will need to be made regarding the appropriate mix of keyword lengths within your natural search engine optimization campaign.
Summary Conclusions and Actionable Implications
Many search engine marketing campaigns currently omit single-word queries (often due to perceived difficulty in achieving high rankings on those keywords), and therefore target longer two-, three-, and four-word phrases instead. The data that serves as the basis for this study clearly establishes that avoiding single-word keyword targets (remembering that we are addressing marketers responsible for large Fortune 1000-level company websites) means ceding a substantial proportion of traffic to competitors who ARE targeting those keywords.
Similarly, targeting a fixed number of multi-word phrases does not consider the characteristics of natural search engine marketing strategies – namely, that most search engine marketing firms assess no incremental charge for increasing the size of keyword target universe – and that there’s certainly no incremental cost (aside from time) associated with targeting a larger set of keywords if you manage your SEM campaign in-house.
Because there is no economic basis to pursue a fixed, smaller universe of keyword targets there is little reason not to target a larger keyword universe – especially because doing so often serves the campaign objectives and the physical characteristics of larger company websites (dynamic or static sites comprised of thousands of pages typically contain many more “targetable” keywords than smaller sites).
If, however, you limit your keyword target universe to a contract-specific target set that is defined by your SEM firm – or which you arbitrarily create in order to maintain a more “manageable” quantity of keywords, this strategy does not reflect successful search referral trends experienced by the websites of large companies.
About iProspect
Founded in 1996, iProspect is the Original®Search Engine Marketing Firm, helping many of the world’s most savvy marketers at many of the world’s most successful brands maximize the return on their online marketing investment. iProspect creates custom search engine marketing campaigns for each of its clients based on their specific needs and marketing goals – selecting the appropriate mix of natural (organic) search engine optimization, paid inclusion management, pay-per-click advertising management, web analytics, and website conversion enhancement services.
iProspect was named to the Inc. Magazine Fast 500 and Deloitte & Touche Technology Fast 500 lists in both 2003 and 2004, was awarded the honor of Best Search Engine Marketing Vendor by ClickZ in 2004, and is the only search engine marketing firm to have ever received an A+ rating in the MarketingSherpa Buyers’ Guide to Search Engine Optimization Firms.
Attribution and Further Information
Any references or quotes pertaining to this document should be attributed to the “iProspect Natural SEO Keyword Length Study.” Any questions or inquiries from the press, should be addressed to iProspect’s Media Relations Manager, Colleen Reed at colleen.reed@iprospect.com.
