A recent study suggested Wikipedia appears on page one of the search results for 99 percent of search results. Conductor analyzed 2,000 informational and transactional queries and found Wikipedia on page one for 6 out of 10 informational and 34 percent of transactional queries.
Last month Danny Goodwin at Search Engine Watch and Matt McGee over at Search Engine Land each wrote about a study from Intelligent Positioning that found Wikipedia ranks on Google UK for 99 percent of searches. The study looked at Wikipedia’s appearances in the search results for 1,000 keywords.
From Search Engine Land on the methodology:
As Intelligent Position explains, the company used a couple random noun generators to come up with a list of 1,000 nouns — words like “ashtray” and “volcano,” “snowflake” and “melody.” It then did 1,000 unique searches on Google UK and charted if and where Wikipedia showed up in the first page of results.
Many in the comments on the Search Engine Watch article raised questions about the methodology used, and over on Search Engine Land, McGee expressed his own reservations while suggesting an alternative approach:
Most search queries are longer than one word nouns. Chitika recently pegged queries at between 4.07 and 4.81 words on average, depending on the search engine. A couple months ago, Hitwise reported that 27 percent of searches that produced clicks were one word – leaving 73 percent of searches not represented in this study.
What I’d love to see someone do is this: Do a thousand searches (or more) that represent actual search engine behavior. Make 27 percent of those random searches be a single word (like “tortoise” or “liquid”); make 24 percent be two words (like “buy laptop” or “ankle pain”); make 19 percent be three words (like “u2 song lyrics” or “funny Valentine’s cards”), and so forth up to seven or eight words.
And then, using a variety of search terms that mimics actual search behavior, show how often Wikipedia appears in the first page of results. I’m pretty sure it’ll still be very high, but it won’t be 99 percent of the results.
With no disrespect to the authors of the original study (in fact, a thank you for sparking the discussion on the subject is in order) we followed McGee’s suggestions for expanding the study methodology, to incorporate a broader range of queries and reflect searcher behavior.
Based on Search Engine Land columnist Shari Thurow’s comment on the writeup –
“Searcher goals are currently classified as informational, navigational, and transactional…”
– we gathered a total of 2,000 keywords: 1,000 keywords each for informational and transactional queries (an analysis of a subset of navigational keywords showed Wikipedia was largely not visible for these kinds of queries, so we did not complete a full analysis of these terms.) In breaking out results by informational vs. transactional queries, we felt that Shari made a good point– a more complete view of Wikipedia in the SERPs should include an analysis of how its visibility differs based on query type.
Using the Hitwise breakout of queries by keyword length cited by McGee, our distribution of keywords looked like this:
Keywords were gathered using a combination of Google Suggest, keyword research website Soovle (which queries multiple search engine for search suggestions simultaneously), including a sampling from their ‘top internet keywords’ list, and the Google Adwords tool. Care was taken to mix high volume searches with lower-volume, self-generated queries to produce a balanced keyword list that is reflective of what actual searcher activity might look like.
Examples of each keyword type:
In the beginning of February, we entered the keywords in Searchlight, Conductor’s enterprise SEO platform, which then gathered ranks on Google.
Our analysis of the keyword rankings showed Wikipedia appears on page one of the search results for six out of ten (60%) informational keywords. It was far less visible for transactional keywords, appearing on page one 34 percent of the time. These findings line up with Wikipedia as an online encyclopedia, highly visible for searches for information, less so for transaction-based queries.
Interestingly, and perhaps contrary to how many might think about Wikipedia in the SERPs, not appearing on page one of the search results did not mean it did not appear at all: for one-quarter of the combined set of keywords Wikipedia appears on page two or higher.
When Wikipedia did appear on page one of the search results, it appeared in top-visibility positions. For 66% of informational keywords Wikipedia was in position 1-3 of the search results, where studies show up to 58 percent of all clicks occur.
Analysis of page one appearances by query length shows Google treats Wikipedia as a particularly relevant result for one word queries: Wikipedia appeared on page one of the search results for 85 percent of informational queries. 73 percent of one word transactional queries were on page one, but that may have more to do with the fact that one-word queries are difficult to convey searcher intent. For example, searching for the term ‘headphones’ might have transactional intent, but absent elaborative keywords, Google likely treats it, and queries like it, as an informational query and therefore presents Wikipedia as a search result. A drop to Wikipedia appearing for 19 percent of two word informational queries supports this assertion, with longer string searches expressing user intent more clearly.
Wikipedia has long been a thorn in the SEO professional’s side. Our analysis at Conductor shows when it does appear in the SERPs, it is high up on the search results and is extremely visible for broad, one-word queries. But in looking at a holistic keyword set that more closely represents actual searcher behavior it is clear that Wikipedia is not appearing quite as frequently as earlier research has suggested.