Boolean Search Strings Anyone Can Use to Find Candidates Now!

Though many articles on the Electronic Recruiting Exchange have talked about advanced search queries using Boolean logic online and how they narrow in on just the right candidates, I haven’t seen anyone explain exactly how to construct a great Boolean search string step-by-step, and adapt it to your specific needs. While various recruiting software products (e.g., AIRS Search Station) simplify the process by letting you type keywords and use pulldown menus instead, you don’t need to spend a cent for similar results. Nor does it require knowledge of HTML or computer programming to write a Boolean search string that zeroes in on who you want. Let’s go after resumes residing on Web pages. This is the biggest pool to draw from-at least 10 million more resumes than the 5 million total resumes contained in proprietary databases of career sites (e.g., Monster.com). To start, we need to limit the search results to resumes. Unfortunately, hardly anyone puts the word “resume” in the resume itself. Therefore, if you simply search for “resume” along with appropriate skill keywords on your favorite search engine, the percentage of actual resumes in your results will be quite low. Most of the time, however, candidates put the word “resume” in the title of the document (i.e., what displays as white text against the blue background at the very top of your Web browser), as in “Resume of John Smith”, “Jane Doe’s Resume” or “Resume: Unix Administrator.” If not, they often include ‘resume’ in the Web page address (URL). For example, www.recruiting-online.com/resume.html or members.xoom.com/clobdell/resume.htm. So go to AltaVista, click the “Advanced Search” folder tab, click in the “Boolean Search Query” box, and type (in all lowercase letters, exactly as follows): (title:resume or url:resume or resume) In AltaVista Advanced, when you type title: immediately in front of the word(s) you want (with no space after the colon), it searches only in that Title bar at the top. Ditto for url: immediately preceding your search term(s) to search only in the URL. But we also included resume with nothing preceding it, meaning search for the word in the normal body, because we don’t want to miss the few candidates who might put it there. But that’s where trouble could arise, because many other Web pages that contain the word “resume” are NOT resumes (resume-writing services, employers’ job descriptions, etc.). We want to weed those out. These pages typically say something to the effect of “To apply for this job, send us your resume…” So we need to add more to our search string, keeping in mind the types of words found on these pages that are NOT on actual resume pages, and eliminate those with the AND NOT command. Your complete search string should now read: (title:resume or url:resume or resume) and not (“resume writing” or job or “career advancement” or “employment opportunity” or “human resources” or eoe or hr or preferred) Note the use of quotation marks: they only surround phrases; never place them around a single word or any group of words unless you are searching for pages where those words appear in consecutive order as typed. AltaVista searches for upper and lowercase of words when you type in lowercase, which is why we didn’t write EOE or HR. However, if you only wanted to find words when they appear on pages as all CAPITAL LETTERS, then type them in all CAPS. The above string is the core of most any Boolean search. What follows is the part that varies based on your particular candidate needs. The only important exception to the above is if you are searching for human resources personnel. Obviously, having “human resources” and hr in the “AND NOT” statement could eliminate many good resumes. Now let’s say we’re an Eastern Massachusetts employer looking for a site designer who knows JavaScript. Think about words and phrases typically in these resumes, but not other types of candidates’ resumes, that will let us narrow down by skill. Remember synonyms: what one candidate calls something might be named something else by another. Either way is acceptable, so we’ll put that in an “OR” statement below: (“web designer” or “graphic designer”) and javascript JavaScript is a separate term, and we must have that skill, so we use the Boolean operator “AND” between them. “AND” indicates that both things on either side of it must be true in each search result. Applied to the above, the parenthetical phrase must be true (i.e., at least one of those skill titles must be in the resume), and javascript must be in the resume. If your candidate must have additional skills/terms in their resume, you can append additional AND statements. In my opinion, the toughest part is limiting the resumes to local candidates. Unlike proprietary databases where the candidate’s location is specifically logged and searchable, the open Internet currently forces you to search against what’s in the resume. Searching by town and/or state is problematic for various reasons (e.g., the location name or its abbreviation is often in the name or location of colleges people graduate from, so your results may include people from all over who happened to graduate from a school in that state). I prefer area codes, though zip codes can work. For example, all Massachusetts zip codes begin with 01 or 02. We can take advantage of the Boolean operator known as “the Wildcard” for this. The Wildcard (symbolized by the asterisk character on your computer keyboard) stands in for any character or characters. So (01* or 02*) would be a Boolean string to find any such zip. Unfortunately, AltaVista requires at least three characters before a Wildcard, so you’d have to use (011* or 012* or 013* etc… 021* or 022* … etc.), making for a very long search string that might crash your browser or cause an AltaVista error. It’s cleaner with area codes. Eastern Massachusetts has only four area codes, so you can search that whole area with (781 or 508 or 978 or 617). This isn’t perfect, however. The problem with both zip and area codes is that occasionally these numbers mean other things on a resume. For example, 508 might be a computer course number that someone included in their “Education” section, or 781 might be the first three digits of a Texas zip code. You may have to try your search a few different ways until you find the location criteria that works for you. Sticking with our example, our search string is now: (title:resume or url:resume or resume) and (“web designer” or “graphic designer”) and javascript and (781 or 508 or 978 or 617 or Massachusetts) and not (“resume writing” or job or “career advancement” or “employment opportunity” or “human resources” or eoe or hr or preferred) Note that the “AND NOT” portion is at the end. This is important: Search results are often inaccurate when AND NOT statements are in the middle of a search string. Just before you click that Altavista “Search” button to generate your results, in the “Sort by:” field just below the Boolean Query box, type: title:resume All results containing the word(s) or Boolean phrase in the “Sort by” box will be at the top of your results. In other words, while the search string you created should generate good results, you might have to scroll to the 10th or 20th page of results until you saw all the resumes. By typing title:resume, all the pages that have resume in the Title are conveniently pushed to the first part of your results. Not bad, eh? By the time you’re comfortable with these Boolean searches, my next column (in about a month) will show you additional refinements and the follow-up steps used by top Internet recruiters! Important note: The examples above were tailored for use with AltaVista’s Advanced Search. These can also be used with other search engines that support Boolean queries (e.g., Hotbot.com, Infoseek.com, NorthernLight.com) but the syntax may need to be tweaked slightly. For specifics, click the link for Search Help on whichever sites you wish to use.

Article Continues Below

Glenn has been developing innovative sourcing and recruiting strategies, techniques and tools in scalable, cost-effective ways since 2015 at State Street, one of the world’s largest custody banks, focusing on diverse talent for North America.  From 2010-15, he was Group Manager, Sourcing Center of Excellence at Avanade, a $2 billion IT consultancy owned by Accenture and Microsoft. He led an online-focused offshore team and junior onshore calling team, plus some global training and talent sourcing initiatives.  In 2009, he conceived and implemented the Sourcing Lab series at SourceCon, which soon became its most popular track.  In 2017, he devised and proposed the Programmers track, which debuts at this SourceCon.

In the 1990s, Glenn created Recruiting-Online.com, a pioneering Internet recruiting seminar which remains the world's longest continuously-running, self-paced online talent sourcing course.  He has trained recruiters from hundreds of companies from the Fortune 500 to small staffing firms.  His popular "Beyond Job Boards" presentations have helped job-seekers tap the hidden employment market via innovative methods around search, social networking, and personal brand enhancement activities.

Glenn was a senior Internet researcher for Microsoft from 2005-2008, focusing on competitive intelligence and proactive international recruiting, following 2 years in a similar role at IT solutions firm, Getronics. Besides presenting at recruiting industry conferences, he co-founded the Boston Area Talent Sourcing Association (BATSA) in 2014 which he still runs.

A Yale University graduate, Glenn discovered recruiting in 1996 by founding the first newspaper chain-owned regional resume/job board in Massachusetts, JobSmart, which won the industry’s two most prestigious awards in 1998: the EPpy Award (Editor & Publisher) and the Digital Edge Award (Newspaper Assn. of America).

Topics

Leave a Comment

Your email address will not be published. Required fields are marked *