Sourcers and recruiters are often tasked with finding candidates that have experience in a specific industry. I’m sure that such a thing seems easy to the hiring managers and clients making the request, but it’s actually not an easily accomplished feat to perform exhaustively.

Sure, finding some people who work in a specific industry is easy – simply target one or a few major companies/competitors and you’re off to the races, right?

Not so fast, unless you’re happy only finding some people and you’re not really concerned with finding the best.

Most industries are comprised of many companies, and some have several hundred to over 1000! How can anyone say for sure that if they targeted 10 or fewer companies in an industry that they were exposing themselves to the best talent available?

Some people (and companies) think that the best talent can only come from a short list of companies they’ve identified, which seems both absurd and short-sighted in my opinion. The most talented “game changers” don’t always come from a blue chip Fortune 500 company.

However, even if a sourcer/recruiter wanted to identify people who worked at any one of a large number of companies in a particular industry, they are stuck to only searching for a few companies at a time because most search engines/interfaces have limits to the length of the search string that can be run. This can make for an extremely tedious and laborious search process, which explains why most sourcers and recruiters only search for a handful of companies or make use of built-in industry search functionality.

Industry Search Limitations

One way to search for people who have experience in a specific industry is to use an industry filter/selection, such as LinkedIn’s:

However, a while back I wrote about the intrinsic challenges and limitations of searching for people based on an industry selection which can actually prevent you from finding the people you’re looking for.

A large part of the problem lies in the fact that many people have experience working in more than one industry, yet they can only select one on their LinkedIn profile. The other issue at hand comes from the fact that people can and will identify themselves in any way they want to – which may have nothing to do with the way YOU would identify or label them. 

For example, a marketing professional working for a pharmaceutical company can just as easily think of themselves as working in the “Consumer Goods” or “Marketing and Advertising” industry when making the selection on their LinkedIn profile rather than “Pharmaceuticals.”

Of course this challenge isn’t limited to LinkedIn or social media in general – this phenomenon also occurs in job board resume databases and corporate ATS/CRM systems with similar functionality.

Is There a Better Way to Search for Industry-Specific Experience?

Let’s say you are in need of finding people with SAP experience who have worked in the Food Production industry. If this were something you were asked to do on a regular basis, you’d probably have a list of companies that you typically target for these folks. However, even if you didn’t you can easily create one.

Using LinkedIn’s company search functionality, you can browse for your industry and make appropriate selections amongst the various search options to end up with a list of companies that fit your criteria.

In the example, I refined the search down to 116 companies. Most sourcers and recruiters only search for a handful of companies at a time (or at all), typically due to the limitations of most search interfaces/engines.

However, because a basic LinkedIn account has for all intents and purposes “bottomless” search fields, we can take those 116 companies and, through some creative use of Excel, Word and find and replace, create a 350 word OR statement that we can put in the Company field:

Try doing that with a job board database, an Internet search engine, or your ATS/CRM.

If you take a look at all of the industries in the search results, you’ll see that most of the people who have worked for one of the 116 Food Production companies we searched for did not choose “Food Production” – only about 18% of the results from the top 10 industries represented in the results come from “Food Production” (572 out of 3,181).

If someone was searching for people who mention SAP on their LinkedIn profile and who selected the Food Production industry, they would only be finding a small fraction of the total available and relevant results!

Thank You LinkedIn!

The ability to search for large volumes of companies in a particular industry (I’ve had no issues with 300+!) gives you a more exhaustive and complete method of identifying potential talent with specific industry experience, regardless of which industry the potential candidates selected when creating/updating their profile. 

Of course, when crafting company searches, you’ll have to keep in mind that there are often many ways that people will write company names – you can only find exactly what you search for in most cases.

Happy hunting!

If you ever have a need to perform diversity sourcing, I’m going to show you a trick on LinkedIn that goes beyond the obvious and “everyone’s doing it” methods of searching for fraternities, sororities, specific universities, and of course groups, societies and associations.

Let’s say you were in need of identifying people with specific skills and experience that are also women (software engineers, CFO’s, etc.), and you’ve already tried the standard methods of identifying them. One tactic some people and organizations utlize is searching for common first names for women. However, with most search engines, you’re limited in the size of the search string you can run (sometimes as few as 100 characters!), so you can’t search for many names with a single search. Plus, limiting yourself to only the most common first names is, well…limiting.

While I’ve written about the fact that LinkedIn’s search fields appear bottomless (I have yet to find a limit to the number of characters/terms that can be entered and searched for), I don’t know of many people who try and take advantage of LinkedIn’s limitless search fields.

See where I might be going here?

Imagine if you could search for a large number of names at once – not just common names, but practicaly all names.

You could go to a baby names website (or any names website – take your pick, there are MANY), look up girl names, and build huge OR strings.

For example, here is a search for 385 girl names starting with C:

(Cade OR Cadee OR Cadence OR Cadrian OR Caelyn OR Cai OR Cailey OR Cailin OR Caimile OR Cairine OR Cairistiona OR Cait OR Caitlin OR Caitlyn OR Caitrin OR Caja OR Cala OR Calandra OR Calantha OR Caledonia OR Calendre OR Caley OR Calida OR Calista OR Calix OR Calixte OR Calla OR Callan OR Callia OR Callidora OR Calliope OR Callista OR Callisto OR Caltha OR Calypso OR Cambria OR Camelai OR Camellia OR Cameo OR Camilla OR Camille OR Camira OR Canace OR Candace OR Candice OR Candida OR Candide OR Candie OR Candra OR Candy OR Cantara OR Caoimhe OR Capri OR Caprice OR Cara OR Caragh OR Carajean OR Caralee OR Caralyn OR Cardinia OR Caren OR Carensa OR Carenza OR Caresse OR Carew OR Carey OR Carha OR Cari OR Carin OR Carina OR Carine OR Carissa OR Carla OR Carleigh OR Carlen OR Carley OR Carlie OR Carlin OR Carlina OR Carling OR Carlota OR Carlotta OR Carly OR Carmel OR Carmela OR Carmelita OR Carmen OR Carmine OR Carmita OR Carna OR Carnation OR Carnelian OR Carol OR Carolena OR Caroline OR Carolyn OR Caron OR Carreen OR Carrie OR Carrieann OR Carrigan OR Carrington OR Carryl OR Cary OR Caryn OR Carys OR Cascadia OR Casey OR Casilda OR Casondra OR Cassandra OR Cassia OR Cassiel OR Casta OR Castalia OR Catalin OR Catalina OR Catalonia OR Catarina OR Catava OR Caterina OR Cathay OR Cathee OR Catherine OR Cathie OR Cathleen OR Cathlene OR Cathy OR Catrin OR Catriona OR Cauvery OR Cayla OR Cayleigh OR Ceana OR Cecania OR Cecia OR Cecile OR Cecilia OR Cecily OR Ceinwen OR Celandia OR Celandine OR Celena OR Celene OR Celeste OR Celestyn OR Celia OR Celie OR Celina OR Celine OR Cerdwin OR Cerelia OR Ceres OR Ceridwen OR Cerise OR Chaitaly OR Chaitra OR Chalsie OR Chana OR Chanah OR Chance OR Chandani OR Chandi OR Chane OR Chanel OR Chantal OR Chantel OR Charis OR Charisma OR Charissa OR Charisse OR Charity OR Charla OR Charlee OR Charleen OR Charleigh OR Charlene OR Charline OR Charlot OR Charlotte OR Charmaine OR Charmian OR Charo OR Charu OR Chastity OR Chavi OR Chay OR Chaya OR Chaylen OR Chelle OR Chelsa OR Chelsea OR Chelsi OR Chelsia OR Chenoa OR Cher OR Cheri OR Cherie OR Cherilyn OR Cherise OR Cherish OR Cherlin OR Cherry OR Cheryl OR Chesna OR Chevonne OR Chhavvi OR Chhaya OR Chiara OR Chika OR Chilali OR Chimalis OR Chipo OR Chiquita OR Chitra OR Chitrangda OR Chizu OR Chloe OR Chloris OR Cho OR Choden OR Chow OR Chrissy OR Christa OR Christabel OR Christal OR Christea OR Christel OR Christelle OR Christian OR Christiana OR Christie OR Christina OR Christine OR Christmas OR Christy OR Chruse OR Chrysantha OR Chrysilla OR Chyou OR Ciannait OR Ciar OR Ciara OR Cicely OR Cicily OR Cilla OR Cinderella OR Cindy OR Cindylou OR Cinnabar OR Cinnamon OR Circe OR Cirea OR Cissy OR Claiborne OR Claire OR Clara OR Clarabelle OR Clare OR Clarice OR Clarinda OR Clarine OR Clarissa OR Claudia OR Clematis OR Clemence OR Clementine OR Cleo OR Cleomenes OR Cleopatra OR Cleta OR Cleva OR Cliantha OR Clio OR Clodagh OR Clodia OR Clorinda OR Cloris OR Clotilda OR Clove OR Clover OR Clymene OR Cochiti OR Cody OR Colenso OR Colette OR Colleen OR Colletta OR Collette OR Columba OR Columbia OR Columbine OR Concepcion OR Concordia OR Conncetta OR Connie OR Constance OR Consuela OR Consuelo OR Coorah OR Cora OR Corabelle OR Coral OR Coralia OR Coralie OR Corazon OR Corbin OR Cordelia OR Coreen OR Corentine OR Corey OR Corin OR Corina OR Corinda OR Corine OR Corinna OR Corinne OR Corliss OR Cornelia OR Cornelian OR Corona OR Corowa OR Corrine OR Cory OR Cosette OR Cosima OR Cossette OR Courtney OR Crescent OR Cressida OR Crisiant OR Crispina OR Cristin OR Cristina OR Cristine OR Cristiona OR Cristy OR Crystal OR Curissa OR Cuyler OR Cybele OR Cybil OR Cybill OR Cynara OR Cyndi OR Cynere OR Cynthia OR Cypriana OR Cyprien OR Cyrena OR Cyrene OR Cyrilla OR Cytheria OR Cyzarine)

Yes, LinkedIn can handle that 3800+ character (incl. spaces) search if you cram all of it into the First Name field:

All is Not Perfect

When you couple a first name search string of that size with a keyword and/or title search, I’ve found I sometimes get whitescreens – no results, but no errors either. I’ve also found that sometimes I get names that I definitely didn’t search for, such as “Chris,” when using the first name string I listed above combined with a keyword search. I’m not exactly certain what’s going on there (but I do have my theories…).

Although the huge first name searches definitely work (the size of the search string doesn’t prevent search execution), it seems to help to chop up the names searches somewhat when combining names searches with keywords and titles. Not only does it help with search performance, it also seems to help with relevance (less/no results creep into names not searched for). 

For example, here is a combination of 112 first names beginning with C and a keyword of “Java:”

One Step Further

Although the site I pulled the first names beginning with “C” from was somewhat of a mixture of nationality/ethnicity, you can certainly search for names from specific countries/ethnicities – Hispanic, Indian, Chinese – whatever suits your particular need.

Limitations

Sourcing by name can and will yield some false positives – it would be impossible to be guaranteed to only be returning results of a specific gender or ethnicity. It’s also quite challenging to compile a 100% complete list of names – can you ever be certain you’re searching for them all? I think not.

However, given LinkedIn’s ability to process very long search strings, it is quite easy to search by large collections of first names that have a high probability of returning results of a specific gender and/or ethnicity/nationality.

One Last Trick

When searching LinkedIn for any combination of keywords and/or title, if you were in need of targeting gender – you can add (she OR “she’s” OR her OR herself) to your keyword string. 

While the keywords won’t highlight in your search results – they are most certainly there, and they are pulled from hits in the LinkedIn profile and the recommendations as well, where they are especially telling.

Of course, this technique is even more limiting than searching by first name and only finds a fraction of the total available results, but it can be used as a quick “first pass,” and it does have a high degree of specificity.

Happy hunting!

LinkedIn now shows some new search functionality to the masses who use LinkedIn for free. Perhaps of greater interest, some previously free search options are now premium filters.

I first noticed the changes to LinkedIn’s advanced search page on Saturday, June 19th. Did they appear earlier and I just didn’t notice them? Perhaps I missed a press release?

For those of us without a premium LinkedIn account, the advanced people search interface used to look something like this:

Now it looks like this:

What’s New

While premium account holders have had access to these filters for some time now, the new search options visible to everyone now include sorting results by Function, Seniority Level, and Fortune (50 to 1000).

You can also sort by Company Size:

You won’t be able to take advantage of those search options without upgrading your LinkedIn account.

What’s No Longer Free

Previously, you did not need an upgraded account to search for people and choose to filter your results by Group and “Interested In.” 

If you don’t currently pay for LinkedIn and you click on these search filters from the advanced search interface, you’ll get a pop up window like this:

Although previously free, they are now premium filters, so you have to pay to play with them. 

Or do you?

After you execute a search, if you expand the Groups search filter, you will notice that you can type in the “Enter group name” field. 

 If you begin typing, LinkedIn will show you available options, and you can in fact use the filter to sort by group.

Interesting, yes? Give it a try.

You can also search within LinkedIn Groups the old-fashioned way – by going to Groups, selecting one, and searching within it. You can even click “Advanced Search” and the group you’re searching will be highlighted in the Groups list of the Advanced Search Interface.

As for filtering results by “Interested In,” I’d like to remind you about LinkedIn’s Advanced Search operators. I wrote about them in January 2009, but I don’t know many people who make use of them. However, if you want to be able to sort your search results by “Interested In” without upgrading your account, you’ll want to get to know them.

You can use the interest: operator to search LinkedIn and find people who have specified a particular interest – at least 1 of the 6 listed in this chart:

If you expand LinkedIn’s “Interested In” filter on the left rail, you’ll notice a total of 8 interests (see below), whereas the LinkedIn Advanced Search Operator Chart above only lists 6. 

While it wasn’t difficult to figure out that r = Reconnect, I haven’t yet figured out what = Reference check (if anything). Let me know if you do.

You can run a simple search such as this one using the interest: operator and successfully return results with the specified interest:

ccompany:cisco ctitle:engineer interest:p

If you mouse over the search results, you can see the interest is configured into the search:

You can also see that it works by looking on the left rail:

What the Future Holds

LinkedIn’s new filters are handy – they can definitely help a person quickly and easily refine their search results. I’m sure that more people and organizations will gladly pay to be able to utilize these search features, especially now that they are dangled in front of everyone who is currently using LinkedIn for free.

Charging for more advanced search options makes sense - LinkedIn is a business and they need to make money. I just hope that LinkedIn doesn’t continue the trend of taking free search options and converting them to premium filters.

I’m pretty sure some folks at LinkedIn will end up seeing this post, so don’t expect the Group and “Interested In” search tricks I revealed to last very long.

LinkedIn – you’re welcome for the free user/functionality testing. Can I at least get some LinkedIn schwag, or an honorary upgraded account? 

When it comes to searching LinkedIn using Internet search engines such as Google or Yahoo, there are many different ways to construct your search string (”X-Ray” or otherwise) and get results.

Ultimately, the goal of any good sourcer or recruiter is to find all of the best available potential candidates that a particular source has to offer.

A short while ago, Gary Cozin sent me a link to an article recommending to “forget complex Boolean strings” when searching LinkedIn using Internet search engines.

I read the article and appreciated the advice to go with the easier, simpler, more elegant search solutions, as well as the suggestion to try “Public profile powered by.” I had never thought of taking that approach – but more on that later.

First, a Little X-Ray History

If you do some research online, you’ll find that the LinkedIn X-Ray search strings that experts in the industry were recommending for years looked something like this:

site:www.linkedin.com intitle:linkedin “current * test manager” -intitle:answers -intitle:updated -intitle:blog -intitle:directory -inurl:jobs -inurl:megite.com -intitle:profile -inurl:jobid

That string has 8 exclusions – effective, but far from simple and elegant.

A while ago, I came up with and recommended a simpler, more elegant approach to the LinkedIn X-Ray search that focused on inclusion rather than a long list of exclusions by specifically targeting profiles using (inurl:in OR inurl:pub), which allowed me to get away with only using 1 exclusion: -intitle:directory.

This approach yielded a simple, short and effective “core” X-Ray syntax of:

site:linkedin.com (inurl:in OR inurl:pub) -intitle:directory

Short and sexy, right?

And Then….

Sometime in April, many people (including myself) noticed that when running a “standard” LinkedIn X-Ray string with Google, that rather than returning the familiar results of individual LinkedIn profiles, results would come back organized by first or last name. Additionally, many noticed that jobs would come up at the top of the results.

I checked into the matter and suggested to add a couple of exclusions, in the form of -inurl:dir -inurl:jobs to get back to the desired results of individual LinkedIn profiles.

That worked, but my simple, short and effective “core” X-Ray syntax of site:linkedin.com (inurl:in OR inurl:pub) -intitle:directory got a little longer and uglier with the addition of -inurl:dir -inurl:jobs. I wasn’t particularly happy about that, but it seemed the only way to get rid of the new garbage, non-individual profile, results.

However, it’s still shorter and more elegant than the X-Rays of yore with massive multiple exclusions.

A New Approach

Going back to the idea of using “Public profile powered by” when using an Internet search engine to look for public LinkedIn profiles – I was intrigued by the concept of dropping the X-Ray string altogether and targeting a phrase that is found on every public LinkedIn profile.

It simply never occurred to me to try it, and I sincerely appreciate the novel approach.

Gary Cozin asked me what I thought of the idea. My first impression was that I found it interesting and I respected the mini-breakthrough in thought that produced it, but I told Gary that I needed to do some testing to be able to tell if it is as effective as other methods of searching for LinkedIn profiles on the web.

So I Performed Some Tests

First – remember that when doing any comparison testing of Internet searches , you have to configure them so that they return a manageable number of results – ideally a few hundred or less. Anything higher than that, and you can get into “estimate” territory.

Here is a simple comparison of the “Public profile powered by” vs. a “standard” X-Ray search on Google.

First up – the “Public profile powered by:”

Microsoft “Public profile powered by” “greater atlanta area” “software engineer” java

That yields 7 results. You have to click on “repeat the search with the omitted results included” to get to 103 results. If you click through to page 10 of the results, for some reason there are only really 100. But you get the point.

Next we have a “standard” X-Ray search with the same criteria:

site:linkedin.com (inurl:in OR inurl:pub) -intitle:directory -inurl:dir -inurl:jobs “greater atlanta area” Microsoft “software engineer” java

That yields 287 results. But not really (thanks Google!). If you click through the results pages, you’ll be stopped on page 24 for a total of 232 results.

In that one search comparison, there is a difference of over 100 results, and I didn’t see any duplicates or bad results from my sampling with the “standard” X-Ray.

As a side note, searching for the word “Microsoft” using Google to search LinkedIn will get you many results with hits of common Microsoft software (e.g., Office), as well as the abbreviation “MS,” which even gets hits of Master’s degrees (I found at least 1 instance of this). If you want to force Google to only return the word “Microsoft,” you have to use quotes or the plus (+) sign.

So What Causes the Difference in Results?

That’s a great question. The phrase “Public profile powered by” does seem to be on every public LinkedIn profile, so why would searching for it find fewer results, let alone produce any difference?

Honestly, I am not really sure. Irina Shamaeva has some ideas as to what could cause the difference in search results.

Maybe some people at Google could shed some light on the matter? <not holding my breath>

The Bottom Line

Regardless of why Google isn’t finding all of the available public LinkedIn profiles by searching for “Public profile powered by,” the bottom line is that the “standard” (inelegant and longer) X-Ray string finds more.

Even if the results gap wasn’t as dramatically large as I found in that one test scenario (I did find some larger and smaller results variations), the fact that the results are different AT ALL is what makes me concerned. I don’t like to miss any results.

To make sure you don’t miss any available search results, I strongly encourage you to do your own testing of the various ways to search LinkedIn using an Internet search engine using relevant search terms for whatever you typically look for.

While I love the simplicity and appreciate the approach of searching for public LinkedIn profiles by not having to use the lengthy site:linkedin.com (inurl:in OR inurl:pub) -intitle:directory -inurl:dir -inurl:jobs, I would never abandon the uglier, longer string at the expense of missing any available search results (not even 1!).

Some More Google Fun

Taking the approach of targeting a phrase common to every public LinkedIn profile a step further, I also experimented with searching for the phrases “Create a public profile” and “professionals already on LinkedIn.”

Here’s a comparison:

Microsoft “Public profile powered by” “greater atlanta area” “software engineer” java

Microsoft “Create a public profile” “greater atlanta area” “software engineer” java

Microsoft “professionals already on LinkedIn” “greater atlanta area” “software engineer” java

Approximately 103, 124, and 150 results respectively.

Did that get your attention?

Why are they different at all, when all three phrases appear on all public LinkedIn profiles? I don’t have the answer.

If Google allows you to click on a cached result without apologizing for thinking you’re not human (am I bitter?), you can clearly see Google indexes the other phrases:

So What about Yahoo and Bing?

Everyone has their search engine preferences, and my first choice has always been Google. However, with the recent changes in LinkedIn X-Ray search results as well as the issue of Google thinking everyone has suddenly become a malicious bot sending automated queries to their servers, I’m looking more at Yahoo and Bing.

Interestingly though, as Irina reminded me recently, Yahoo is giving up their search technology and in the near future will be using Microsoft’s. So technically – Yahoo search may not yield any different results than Bing in the future. Enjoy the results while you can.

Anyhoo…Yahoo does currently seem to do a much better job of searching public LinkedIn profiles with the phrase of “Public profile powered by:”

Microsoft “Public profile powered by” “greater atlanta area” “software engineer” java

416 results! But don’t get too excited – if you click through to the end/last page (to keep Yahoo honest), you end up with 208.

Which is less than the 232 we found using a “standard” X-Ray string on Google.

Interestingly, Yahoo also does a better job of getting consistent results when using other phrases such as “professionals already on LinkedIn” (210 results, if you were curious).

What about Bing? Bing apparently doesn’t like searches like Microsoft “Public profile powered by” “greater atlanta area” “software engineer” java.

Final Thoughts

There is always more than one way to run a search, and there is no single “right way” to find people. Use multiple approaches, and always test them to understand and appreciate differences.

When searching any source, your goal as a sourcer or recruiter is to find all of the best candidates the source has to offer. Finding some people isn’t enough.

Regarding search string length when X-Ray searching LinkedIn – if you’re worried about the 32 word search limit that Google imposes, I’d argue you’re not taking into account that a large percentage of people with LinkedIn profiles don’t even flesh them out with (m)any keywords.

Always appreciate, understand and respect how the non-sourcer / recruiter / HR professional uses LinkedIn…most don’t use it as a full-blown resume.

So if you’re loading up your Google/LinkedIn strings with anywhere close to 32 words (e.g., site:linkedin.com (inurl:in OR inurl:pub) -intitle:directory -inurl:dir -inurl:jobs “greater atlanta area” “software engineer” java unix oracle linux application design develop C++ MySql J2EE Lead Senior SQL data javascript hibernate blah blah blah…), you are actually making it impossible to find all of the qualified candidates on LinkedIn.

Contemplate this on the tree of woe.

I just updated my free sourcing and recruiting resources page!

You can always navigate to the free resources page at any time by using the top nav bar which will escort you to a library of nearly 50 knowledge-laden links!

All in one page you can find a collection of articles related to sourcing and recruiting, including the 5 levels of talent mining/candidate sourcing, candidate sourcing best practices and mistakes, social recruiting, how to search LinkedIn, Twitter, Facebook, Spoke, ZoomInfo and Jigsaw for free, Boolean logic, semantic search, Lean / Just-In-Time sourcing and recruiting, and automated candidate sourcing and matching solutions.

Click the image below to be taken to the free sourcing and recruiting resources page:

Be sure to bookmark or tag this page as I will continue to add more free sourcing and recruiting resources. Also – please “pay it forward” – share this page with anyone who might benefit!

Thanks!

Imagine that you were just assigned a position to recruit for and that you needed to present 2 fully screened, highly qualified and well matched candidates within 5 business days.

Your manager/client is requesting candidates with:

• 3-5 years of related work experience (your choice – something not too vanilla/easy/common, but not ”purple squirrel” either)
• Experience in a specific industry, and experience working in a similar environment (size/scale/team/software, etc.)  to the manager’s/client’s
• Bachelor’s degree in a related discipline

Furthermore, let’s say that you don’t have any qualified candidates in your pipeline, so you are essentially starting from scratch.

Under those conditions and assumptions, if you were limited to only 1 method/specific source for identifying candidates to contact, engage and recruit, which would you choose, and why? 

Please Respond to the (100% Anonymous!) Poll

Why Would You Choose That Method/Source?

Okay, this part isn’t so anonymous.

I’m interested to know exactly why you would choose that specific source/method of talent identification.

I believe the question ultimately boils down to which source/method do you believe has the highest probability of enabling you to produce the 2 fully screened, highly qualified and well matched candidates within 5 business days.

Would your answer be the same if you had to produce the candidates within 48 hours?

What about 24 hours?

How about same-day?

What – does that sound crazy to you? If you’ve never had a manager or client ask you to produce candidates same-day before, it actually happens quite often (to me, more times than I can count!).

I’m looking forward to your responses and supporting arguments as to which source/method you would use and why, and you can expect me to challenge them. If I get a good number of responses - I’ll share with you what MY answer would be.

Bring it!

The game is afoot.

It appears that something’s happening behind the scenes at LinkedIn. Have you noticed anything odd when running a typical LinkedIn X-Ray search on Google and Yahoo?

I have. For example – try running this search:

site:linkedin.com (inurl:pub OR inurl:in) -intitle:directory (controller OR CFO) “Charlotte, North Carolina Area”

You’ll get results that look like this:

Notice how instead of returning results of individual LinkedIn profiles – for some reason, the results come back organized by first or last name. 

Clicking on the first result will take you to a page like this, where only the first result matches the X-Ray search parameters:

Hints in the URL

Notice the URL of each result:

The URL is truncated – so we’re missing some info.

If you’re using Firefox, you can drag your cursor/pointer over the result title to view the entire URL (at the bottom of the browser). Simply clicking on an individual search result will also show you the entire URL:

If you’re a search geek like me, you’re no doubt intrigued by the URL, and you’re trying to figure out exactly what LinkedIn is doing behind the scenes to take a “standard” X-Ray string and return results organized and grouped by first or last name.

I certainly didn’t search for “Larry.”

OR vs. AND

I’ve initially found that these new and different results seem to be more prevalent with X-Ray search strings using OR statements as opposed to simple AND-based queries.

For example, this 2 keyword X-Ray string brings back a mix of our more familiar individual profile results from LinkedIn as well as results grouped by first AND last name:

site:linkedin.com (inurl:pub OR inurl:in) -intitle:directory java oracle ”Charlotte, North Carolina Area”

I haven’t yet figured out why an LinkedIn X-Ray searches with OR statements seem to have a much higher percentage of the new organized-by-first-or-last-name search results than X-Ray searches only using ANDs. 

Have any ideas?

Current Title Searches

For those who are curious – configuring X-Ray searches to try and target LinkedIn profiles with specific current titles (e.g., “current * controller”) are also similarly affected.

Actually, this kind of search seems to suffer more, because the result that actually matches the current title condition isn’t typically the first search result. In fact, in one of my tests, when clicking the first search result and searching through all of the results returned by first name  the only matching result on the first page was the 23rd!

site:linkedin.com (inurl:pub OR inurl:in) -intitle:directory “Charlotte, North Carolina Area” “current * controller”

The Fix

The good news, at least temporarily, is that we can easily get back to our more familiar X-Ray search results containing only individual LinkedIn profiles.

We can actually do this in at least a couple of different ways, but my preferred method works because all of the new and strange search results seem to have “dir” in the URL.

That means all you have to do is add -inurl:dir to your X-Ray search strings.

For example,  adding -inurl:dir to the X-Ray search I used at the beginning of this post:

site:linkedin.com (inurl:pub OR inurl:in) -intitle:directory -inurl:dir (controller OR CFO) “Charlotte, North Carolina Area”

Bye bye annoying non-individual, organized-by-last-name search results! 

And if you run into a result or two that happen to be jobs, you can simply add -inurl:jobs to your string to make them vanish.

Is This The End?

For the time being, the -inurl:dir seems to get us back to our familiar individual LinkedIn profile search results, even for current title X-Ray searches.

However, many people (including me) have been wondering for years as to if, when, and how LinkedIn would make changes to the ability to effectively search for and find specific LinkedIn profiles using Internet search engines.

The folks at LinkedIn have always known that people can use Internet search engines to find LinkedIn profiles that are not in the searcher’s LinkedIn network. It’s a byproduct of publishing profiles to the web, which LinkedIn does by default unless the user changes their account settings.

Are we seeing the beginning of LinkedIn attempting to reign in X-Ray capability? I’m not so sure – I would think if they were going to make a such a move, it would be more effective.

Perhaps they are testing the waters to see how they can continue to publish profiles to the web yet affect X-Ray search results?

Keep a weather eye on the horizon.

Strictly for the Search Geeks

Do any of my fellow sourcing hackers have any ideas as to:

1. What exactly is going on with the search results being returned and organized by first or last name?
2. What determines whether a search result is organized by first as opposed to last name?
3. Why do X-Ray searches with OR statements seem to have a higher percentage of the new search results than simple AND-style searches?

In the original LinkedIn Current Title Search: Internal vs. X-Ray post, I found that there was a large disparity in results between identical internal and external (Google X-Ray) current title searches on LinkedIn. 

At the end of the post I asked if anyone had any ideas as to where the missing 288 profiles were, why they weren’t returned using the basic Google X-Ray current title search technique, and how can you find them?

I received a few interesting and insightful responses - I will review them in this post as well as provide my own insight as to the difference in results experienced when running an internal LinkedIn search and a Google X-Ray search of LinkedIn using the same search criteria.

I’ll also show you some interesting LinkedIn search phenomena.

Does Results Sorting Matter?

Kathy Breitenbucher noticed as I did that LinkedIn seems to be taking some liberties with the “Director of Accounting” title, and she decided to sort the results by keywords rather than the default setting of “relevance” (which really isn’t relevance based on keywords, in case you didn’t know).

I decided to sort the results by keyword as well – but I did not see the result she referenced (”The first person on the list is an Executive Director, HR & Accounting”) in the first 10 pages. However, I did notice that on the main results listing, some in fact do not mention “Director of Accounting.”

For example:

However, if you click on the result to view the profile, you will notice there is some reference to “Director of Accounting:”

So be careful not to judge a search result by what is or is not displayed on the main results listing. As I discovered in all cases, even if some form of “Director of Accounting” was not present on the main results listing, when I opened each result, there was in fact at least one relevant hit of the phrase (regardless of order – more on that later).

Does LinkedIn Ignore Quotes?

Medhavi Misra believed that “…quotes are completely ignored in linkedin search and thus the number of results is more.” 

I am not so sure that LinkedIn completely ignores quotes. In every result that I reviewed during my research, there was always at least one instance in which “director” and “accounting” were within close proximity, which leads me to believe that LinkedIn is honoring some form of general phrase proximity.

On the other hand, Medhavi was very observant to notice that a search for “director of accounting” on LinkedIn also returns results with phrases like “Director – Accounting”, “Director & Accounting”, “Director, Accounting”, “Accounting Director.”

This in fact is one of the major factors that causes an internal LinkedIn search for “Director of Accounting” to return more results than a Google X-Ray search of LinkedIn for the same phrase.

However, that is only half of this particular phenomenon. 

Google is Strict, LinkedIn gets Loose

Yogesh Kumar and Irina Shamaeva both pointed out that Google seems to strictly honor phrase searching with quotation marks. In other words, when you use Google to search for “Director of Accounting,” it will only return results with the exact phrase of “Director of Accounting.”

As you can see below, LinkedIn isn’t so strict.

This is also evident on individual LinkedIn profiles.

For the most part – this is a good thing! LinkedIn is “smart” enough to realize that people can represent titles and phrases in various ways while the meaning of the concept remains intact.

LinkedIn’s flexibility with phrase searching and Google’s strict adherence to exact phrase searching likely accounts for a large portion of the missing 288 results referenced in my original post. 

Yogesh Kumar and Kameron Swinton pointed out that LinkedIn appears to ignore common words (such as of, or, at, for – as many search engines do), even if they are contained in what most people would assume is a strict exact phrase search using quotes, such as “Director of Accounting.”

Proof of this can be seen here, where the word “of” is not highlighted:

To compensate for LinkedIn’s phrase search flexibility and how it ignores common words, Yogesh suggested the following Google X-Ray search string to more closely approximate what LinkedIn is doing:

site:linkedin.com (inurl:pub OR inurl:in) -intitle:directory “greater new york city area” “current * * * director * accounting”

That search yields nearly 13,000 results- quite a bit more than the 393 that LinkedIn returns!

When you examine the results, you can see what’s going on here – the replacement of “of” with an asterisk [Director * Accounting] allows Google to return results with titles such as:

• Director Corporate Accounting
• Managing Director at DSZ Forensic Accounting
• Director – NAC Accounting & Planning
• Associate Director, Finance & Accounting
• Managing Director at NetWorth Tax Accounting
• Director/Vice President, Accounting & Finance
• Director Corporate Reporting & Accounting

This search actually represents an improvement on the original search of “Director of Accounting” because in addition to returning that exact phrase, it also returns results that do not match the exact phrase yet still match the intent of the search – which is to find people at the director level who oversee accounting functions.

As such, this is a wonderful example of semantic search – targeting the meaning of the concept of director level responsibility over accounting without being restricted to an exact phrase.

However, all is not perfect with this search.  Remember those 13,000 results? When I navigated through the results pages on Google to review them for accuracy and relevance, I noticed that the results stopped at page 6:

I tried running the X-Ray search in IE and Firefox and got the same results – I could not navigate past page 6, and the actual total number of results ended up being 57. A far cry from 13,000!

Any ideas as to what’s going on here?

What about Indexing?

Gov’t Recruiter took a totally different angle than anyone else who responded to my challenge. His answer as to why the Google X-Ray search returned fewer results than the equivalent search inside of LinkedIn is that ”LinkedIn indexes with its RDBMS immediately as a new profile has been created, granting LinkedIn an Instant Refresh. Google’s results are based off spiders and metatags keying and registering  typically taking 45 days to index for most search and meta search engines.”

Now, I am not certain of the accuracy of his claims of it taking around 45 days for Google to find and index pages (such as public LinkedIn profiles), but I do think he is onto something here. Certainly if someone creates a new LinkedIn profile today, it is not likely to be indexed by Google (or any search engine) in the same day and thus searchable and retreivable via Google immediately, whereas new profiles ARE instantly retrievable when searching inside of LinkedIn.

Do you have any thoughts, additional insight or facts (such as exactly how long it DOES take for Google to crawl and index a new LinkedIn profile) to share on this concept?

Overlap and Private Profiles

Irina Shamaeva offered a few ideas as to the variance in results between an internal LinkedIn search and a Google X-Ray search: 1. The results of using one, two, and three asterisks may be different, so we need to combine them all for an X-ray answer. 2. Private profiles won’t show up in Google. 3. It seems LI may decide that “director accounting” is the same as “director of accounting”; Google takes words in quotation marks as they are; am I right?

While the possibility of combining the search results from the 3 separate searches is a sound theory, when I wrote the orignal post, I carefully reviewed the results of 1, 2, and 3 asterisks in the phrase “current  */* */* * * director” and I found there to be significant overlap in the results. Thus we cannot simply add the results of the search with 1 asterisk (85) to the results of the searches with  2 and 3 asterisks (98 and 105 respectively) to explain the variance in results.

However, Irina correctly pointed out that any LinkedIn profile that is not published to the web would not be indexed and thus searchable/retrievable via a Google X-Ray search. I don’t have the stats on how many people actually change their account settings from the default/recommended “publish profile to web” to “None off,” but from my experience it is likely a small percentage of people.  Nonetheless, this is a legitimate explanation of why you may not get as many results from an X-Ray search as you would when searching within LinkedIn. 

Irina also nailed the difference in how Google and LinkedIn process phrase searching with quotation marks, as did Medhavi and Yogesh. Well done! 

My Thoughts

I enjoyed not giving away any answers in the original post. Although I was hoping to have more people try and figure out exactly why there can be such a huge difference between equivalent internal and external LinkedIn searches, I was pleased to receive insightful thoughts from a number of sharp sourcers around the world.

I have found that the best sourcers are people who are curious, analytical, investigative problem solvers. This exercise has proven something that I continually espouse – that it is critical to ask “why?” when you encounter something that doesn’t seem to work quite right, to experiment via the scientific method, and that a significant portion of learning comes from examing the results – especially for what isn’t highlighted.

In addition to what was accurately revealed by all of the people I’ve referenced in this post, I wanted to share a couple of other phenomena I encountered while exploring the differences between internal LinkedIn searches and X-Ray searches.

LinkedIn is Rebellious and Inconsistent

First, I noticed that when searching inside of LinkedIn with a phrase of “Director of Accounting” in the current title field…

…that LinkedIn would highlight one of the search terms (accounting) by itself, not in a phrase, and not even in the title section, as you can see in the “Specialties” section! What gives?

Although I clearly searched for the phrase “Director of Accounting” in the current title field, “accounting” is highlighted by itself outside of the title fields – and yet I noticed that neither the individual search terms nor the phrase “Director Accounting” were highlighted when they were in fact present elsewhere in the profile (see above). 

Anyone from LinkedIn care to explain?

LinkedIn Puts it in Reverse

Second, I also found a few examples in which LinkedIn would actually reverse the order of the keywords in the phrase “Director of Accounting,” which I found quite interesting:

LinkedIn Searches for Relevant Abbreviations

Third, LinkedIn also returned results that did not even spell out the word “Director:”

Kudos to the search team at LinkedIn – nice work!

Thank You!

Thank you to everyone who responded to my original post with thoughts and ideas! Please let me know if you can provide any additional insight to the questions I raised in this article.

I’ve complied a list of the top 25 most connected recruiters on LinkedIn.

To generate this list, I searched LinkedIn for anyone in the world who had one of the following terms in their current title: recruiter OR recruiting OR recruitment OR sourcer OR sourcing OR talent, and I sorted the results by connections. It looks as if you will need at least 20,000 1st degree/direct connections to even crack the top 10!

If you click on each result, you will be taken to their public profile. You might want to connect with them if you haven’t already.

Click here for the full search results on LinkedIn – there are at least 301,384 of them.

If you happen to add “human resources” OR HR to the current title search string, the top 25 results don’t change much, with the exception of the top 4 (2 new results show up there). Also, the total results climb to over 820,000.

So you know how to X-Ray LinkedIn, and perhaps you even know how to target current titles with an X-Ray string. However, did you know that you’re not finding all of the available matches within LinkedIn using this technique?

Did that get your attention? Would you like to know approximately how many people you might be missing when you run a current title X-Ray search of LinkedIn?

Unless you have a premium Linkedin account, you will likely have to resort to using an Internet search engine to X-Ray into LinkedIn to find and view profiles of people who are not in your network, and being able to search by current title can be extremely useful at times to reduce the number of false positive results.

While I am usually not a big fan of title searching, a large percentage of LinkedIn profiles don’t have any text entered for each work experience entry. As such, title searching becomes a necessary evil as anyone who creates a LinkedIn profile doesn’t have to enter anything in the ”position description” field, but they MUST enter a company and a title.

In this post I will show you the discrepancy between identical current title searches conducted using LinkedIn’s search interface and an X-Ray string – and it’s HUGE! At the end, I also throw in challenge for you. Are you up to it?

The Search Criteria

For this exercise, I am going to target people with a current title of “Director of Accounting” in the “Greater New York City Area.”

LinkedIn Current Title X-Ray Search

Here is a “standard” Google X-Ray search string targeting current title:

site:linkedin.com (inurl:pub OR inurl:in) -intitle:directory “greater new york city area” “current * director of accounting”

This yields 85 results at the time I ran the string.

Remember – the asterisk on Google is processed as a “fill in the blanks,” where it serves as a placeholder for any unknown terms. This can be 1 or more unknown terms (you cant nail Google down to a specific number on this).

In the case of the single asterisk, the current title X-Ray string appears to work because the asterisk “fills in the blank” of the bullet that separates the word “Current” from the actual title entered for the most recent position on the LinkedIn profile.

Are Two Asterisks Better Than One?

Because I am curious, I decided to see what using 2 asterisks yields:

site:linkedin.com (inurl:pub OR inurl:in) -intitle:directory “greater new york city area” “current * * director of accounting”

That produced 98 results at the time I ran the string.

The additional results appear to come from instances where ”Director of Accounting” is preceded by a word, such as Asst Director of Accounting, Executive Director of Accounting, Regional Director of Accounting, etc.

We’ll take those, thanks!

Are Three Asterisks the Charm?

Going one step further (hey, why not?), I decided to shoot for 3 asterisks:

site:linkedin.com (inurl:pub OR inurl:in) -intitle:directory “greater new york city area” “current * * * director of accounting”

That yielded 105 results at the time I ran the search.

The results gained come in the form of Google taking some liberty in interpreting my search, actually altering the exact phrase of ”Director of Accounting” with titles such as Regional Director of Finance and Accounting, National Director of Project Accounting, etc. At this point, we’re starting to mess around with the purity of the intent of the search. Close enough though?

Just to see what Google/LinkedIn made of it, I decided to shoot for 4 asterisks, which only returned 46 results at the time of the search. While we can see a few familiar and accurate results from the previous strings, you can also see things got messy, so shooting for 4 asterisks in this case brought diminishing returns, both literally and figuratively.

And Now for the Same Search Performed Inside LinkedIn

Using LinkedIn’s search interface, here is what I configured to find people with a current title of “Director of Accounting” using the zip code of 10001, and going out to a distance of 50 miles which should essentially cover the same area as the X-Ray location of “Greater New York City Area.”

Wait a minute – did I hear someone ask how I can I possibly know if a 50 mile radius of 10001 is equivalent to all of the people categorized by LinkedIn as being in the ”Greater New York City Area?” I’m glad the question was raised!

It’s actually not equivalent to ALL LinkedIn profiles with a location of “Greater New York City Area,” but it’s pretty darn close. Let me show you.

When you run that search, you should get somewhere around 348 results. That’s over 3X as many results as our best X-Ray string!

Scroll down the left rail where you see “location” and click the “+” sign:

That shows us there are AT LEAST 348 people with “Director of Accounting” as the current title on their LinkedIn profile that are also classified by LinkedIn as living within the “Greater New York City Area.”

But wait – there’s more. I decided to push the radius out to 75 miles and then check the location details at the bottom of the left rail:

Yep – we’re up to 389 results in the “Greater New York City Area,” and you can see we just started to bleed over into surrounding metro areas recognized by LinkedIn.

You’re probably not surprised to find out that I decided to push the radius out to 100 miles just to see what happened.  As it turns out, we actually squeeze out 4 more search results (393 total) for a current title of “Director of Accounting” and a location of “Greater New York City Area.”

All LinkedIn Current Title Searches are NOT Created Equal

I’m sorry if I rattled your faith in X-Ray searching LinkedIn for current titles – but this little exercise has shown quite dramatically that all LinkedIn current title searches are not created equal.

I thought you should know that you cannot run a current title X-Ray search of LinkedIn and be confident that you’re finding most of the relevant results. In fact, it appears you’d be missing the majority of the available results!

The best “standard” current title X-Ray string returned 105 results for people with a location phrase of “Greater New York City Area,” whereas we were able to scoop up 393 results of people with a current title of “Director of Accounting” and a location of “Greater New York City Area.” 

Can You Solve the Mystery?

So where ARE the missing  288 profiles, why aren’t they returned using the basic Google X-Ray current title search technique, and how can you find them?

I have some ideas. Do you?