Here we will look at parameters that we can add/modify in the URL of the the Google SERP (search engine results page.)
The operators listed in Part 1 should suffice for most searches. Yet, certain types of searches are most efficiently accomplished by directly modifying the Google SERP URL — for instance, changing the number of results displayed per page, adding omitted results back in, going to the last page of search results, and turning off personalization.
URL parameters are name/value pairs. An equals sign sits between the name and value, and an ampersand separates each pair. These name/value pairs are placed in the “query string” portion of the URL — in other words after the question mark. Assume a base URL of eitherhttp://www.google.com/search? or http://google.com/search?. Both will work equally well.
Chart of Google SERP URL parameters
| URL Parameter | Format Example | Description |
|---|---|---|
| q | http://google.com/search?q=swiss+cheese | The search query. This is the one required URL parameter, all others are optional. When the value for q contains multiple words, separate each word with a plus sign. |
| num | http://google.com/search?q=cheese&num=100 | Set the number of results per page to display in the SERPs. It is SO much faster to switch to 100 results per page by adding &num=100 to the end of the Google URL rather than going into the Search Settings screen. |
| start | http://google.com/search?q=cheese&start=990 | Display the SERP that starts with the specified listing number. |
| filter | http://google.com/search?q=cheese&filter=0 | Include back into the results the listings that Google omitted because they looked too similar to the ones already displayed. This is equivalent to clicking the link that says “repeat the search with the omitted results included” that you’ll often see on the last page of SERPs. |
| pws | http://google.com/search?q=cheese&pws=0 | Turn off personalized results. PWS stands for “personalized web search”. Previously you had to be logged into your Google account for search results to be personalized. Not anymore. Personalized results are given to all — whether signed in or not — based on 180 days of search activity linked to a cookie in your browser. Why hassle withturning off personalization, when you can simply append &pws=0 to the Google SERP URL and retrieve non-personalized results from Google for that query. |
| as_qdr | http://google.com/search?q=cheese&as_qdr=d | Search within a date range to present. Use y for year, m for month, w for week, d for day, h for hour, n for minute, and s for second. Follow the letter with a number if you want more than one. For example, “m3″ is for three months, “h5″ is for 5 hours. Want to see only the freshest of results, like within the last 5 minutes? Specify “n5″. If you want to search for a date range that ends prior to the present day, use the daterange operator or use the “Custom range” under the “More search options” link in the left sidebar of the Google SERP. |
| as_rights | http://google.com/search?q=cheese& as_rights=cc_publicdomain | Restrict results to re-usable (Creative Commons licensed) content. This is probably a more straightforward exercise from the Advanced Search screen, but if you’re feeling adventurous, cc_publicdomain is public domain, cc_attribute is re-usable with attribution, cc_sharealike means you’d have to offer the derivative works under a Creative Commons license, cc_nonderived means no derivative works allowed, and cc_noncommercial means non-commercial use only. You can combine these into groups with parentheses and pipe symbols, and minus signs to exclude certain license types such as the commercially restricted license. For example, (cc_publicdomain | cc_attribute | cc_sharealike).-(cc_noncommercial | cc_nonderived) is equivalent to “free to use, share or modify, even commercially.” There should not be spaces in the above (they were merely added for readability). |
| safe | http://google.com?q=cheese&safe=off | Toggles off or on Google’s safe search filtering. Values are active and off. |
| strip | http://google.com/search?q=cache:www.amazon.com&strip=1 | Only works on the Google cached page. Adding &strip=1 to a Google cache URL causes the cached page to display without images, Javascripts etc. This means that you don’t leave any trace of your pageview in the visited website’s analytics. |
| imgtype | http://google.com/images?q=cheese&imgtype=face | Works in Google Images only. Options include face, photo, clipart, lineart, and news. |
Note that the above is not a comprehensive list of parameters. Some parameters are not usually useful to modify, such as hl (specifies the language used in the Google interface), lr (language restrict; uses standard two-letter language codes), ie (input encoding), and oe (output encoding).
A few more query operators: wildcards, synonyms, NOTs and number ranges
You might recall from Part 1 that quotation marks in a search query signify an exact phrase, and the pipe symbol (|) signifies an OR in Boolean logic. But did you know you can use the tilde operator in front of a word to have Google match on synonyms of that word? For example, ~car repair would also match on auto repair and automotive repair in addition to car repair. Try it and see. You’ll notice all three phrases are bolded KWiC (keywords in context) in the search listings that are returned.
A minus sign directly in front of a word acts like NOT in Boolean logic, eliminating from the search results the subsequent word or quote-encapsulated exact phrase. For example,confidential business plan -template will only return results where the word “template” is not present in the page.
To specify a numerical range, use two dots between two numbers, which could be years, dollar amounts, or any other numerical value. For example, confidential business plan 2007..2010 will find documents that mention 2007 or 2008 or 2009 or 2010.
Another favorite operator is the wildcard, i.e. the asterisk character. It can substitute for a whole word in an exact phrase search. Imagine that you can’t remember the exact words of an expression; you’d simply substitute asterisks for the words you can’t recall.
For instance, “* of my * have been greatly exaggerated” — is it supposed to be “news,” “announcements,” or “rumors” in the first slot? “death” or “demise” in the second slot?
Let’s put what we’ve learned so far in Parts 1 & 2 about Google operators and URL parameters to some practical use…
Estimated number of results
The estimated number of results is just that: an estimate. And a wildly inaccurate estimate at that. You can often get a more accurate count of results than this estimate by going to the last page of results and clicking the link to “repeat the search with the omitted results included.” A shortcut to this is to append &start=990&filter=0 to the end of the SERP URL.
For example, searching Google for site:redenvelope.com returns “about 5,010 results,” but appending &start=990&filter=0 to the end of that SERP URL takes us only as far out as page 84. So the true indexation number is 836. That’s a far cry from 5,010!
Google won’t let you go beyond the 1000th result, so if you are checking indexation and it goes to page 100, then you’re at the end of the line and can’t see beyond that. In these situations, you can work around the limit by looking at just a subset of the site by using the site: and inurl: operators. For example, site:www.example.com/articles/ will return only pages in the articles directory of this hypothetical site. The site: operator used in conjunction with inurl: allows you to refine your search even further, for example: site:www.example.com/articles inurl:print-friendly.
Finding sites to link to you
Finding folks who are inclined to link to your site is a common task in SEO, but it’s a laborious and painful one — IF you aren’t a Google power user. It’s like looking for a needle in a haystack.
Try this trick: get Google to return a list of targets that are topically relevant and that already include attribution links to their web developer or web host. Like so: site:edu | site:gov [some industry] intext:”website * by”.
In this example query, we’re limiting results to .edu and .gov sites that have content related to a certain unspecified industry / keyword market and that have an attribution such as “Website developed by” or “Website marketing by” or “Website hosting by” somewhere on the page. The thinking is that such site owners would be more predisposed to include other attribution links to in-kind donors/sponsors.
Monitoring inbound links
Want to see which competitors are successful in acquiring keyword-rich text links that specify your targeted search term? It’s easy with the inanchor: and allinanchor: operators. For example,inanchor:”neopets cheats” will return pages that have garnered links with “neopets cheats” in the anchor text (there’s my daughter’s site at #3!).
Host crowding
On a page of results, Google will group listings from the same site together. Googlers refer to this as “host crowding.” Recently Google started showing more than two listings per page in certain (relatively rare) circumstances, all host crowded. It can be really useful to determine the true (original) position of an indented (host crowded) listing.
If you customize the SERPs to 9 results per page using the num=9 parameter, and then consequently the indented result disappears, that meant the indented result was actually at position #10 and was merely visually promoted to appear at a higher position.
With that information, you could target other listings on page 2 to promote to page 1. It would take only one new listing added to page 1 to displace your competitor’s indented listing. More on this in my previous article Deconstructing Grouped Google Results.
Keyword competitiveness
I never liked the KEI (keyword effectiveness indicator) score as a metric for keyword competition. For one, the denominator is based on a number (the “estimated number of results”) that, as already discussed, is completely bogus. Besides, how well does that number correlate to ranking difficulty anyways — even if it were accurate?
A better way to gauge the competition for a particular keyword is to compare searches for the keyword you are targeting, then the exact phrase version of the term (in quotes), then that exact phrase (in quotes) within the title tag. It’s the last one that really represents your true competition.
After all, if they didn’t specify that exact phrase in any of their title tags, how seriously were they going after that search term anyways? Consider white noise at about 11,400,000 results, versus“white noise” at about 3,120,000 results, versus intitle:”white noise” at about 79,100 results. The serious competition is represented by the 79,100 number, not the 11.4 million.
