Irina posted an interesting piece on discrepancies in search results between LinkedIn Recruiter and a free LinkedIn account which prompted me to do a little digging as I don’t think I’ve ever come across materially different results in actual use.
While the discrepancies are definitely interesting, and I would love to know exactly what’s causing them, I don’t find them particularly troubling. Read on to learn why.
Even if you don’t have a LinkedIn Recruiter license, you will likely still find this post interesting, as it examines search logic and strategy which can be applied to sourcing via any site/resource.
In the C++ 3D iOS “computer games” example, where a free account returns 150 results and an LinkedIn Recruiter account returns 43, I wouldn’t recommend anyone to search for “computer games” as a keyword as it is too limiting. If the goal is to find people who develop computer games, I would run a broader, more inclusive keyword search than exact phrase of “computer games,” which many people who actually develop games would not use in their LinkedIn profile.
If you were to search for (games OR gaming OR game) instead of “computer games,” you get 494 results for both a free account and LinkedIn recruiter. You could also search using the industry filter in both LIR and a free account – in this case they return the exact same number of results (133).
One thing to notice so far is that “computer games” is also an industry on LinkedIn, and when you use it as a keyword, you get massively different results between LIR and a free account. However, as soon as you drop “computer games” and use other terms to search for the experience, or you actually use the industry filters, you get the same results across both types of accounts.
I didn’t find the hospital/healthcare example alarming, as I am not sure why someone would search for [hospital health care] when they are already searching for current employees of New York Presbyterian Hospital (they’re already working in the health care industry to some extent as they work at a hospital). Also of note is that in the example search, there is no search by title or a real skill, so I am not sure how effective that search would be in helping identify people with any specificity.
If you remove the [hospital health care] keywords and search for current employees of New York Presbyterian Hospital in the NYC metro area, LIR actually returns slightly more (6,911) than a free LinkedIn search (6,865).
If you search for the current title of [Nurse] in addition to the NYC location and current employer of New York Presbyterian Hospital – both LIR and a free search return exactly the same amount of results (1,180), which I believe is a more accurate and practical representation of what a LinkedIn user might search for.
The pattern I can see from all of the examples provided is that they all search for keywords that are also industries (Computer Games, Hospital & Health Care, Research, etc.), and in all cases I am not sure why someone would search for those terms, as they would represent very weak searches IMO. Another way of saying this is that I don’t run into the issue of discrepancies in “real world” searches between a free LinkedIn account and LIR based on the queries I run and train others to construct.
For the health care search, I would advise people who were really looking to identify candidates to use [hospital OR hospitals OR health OR healthcare] as keywords in LIR, which would result in 6,675 results, significantly reducing the gap found. However, as I have previously stated, I wouldn’t actually ever search for [hospital OR hospitals OR health OR healthcare] when searching for current employees of a hospital, as they are implicitly working in that industry by the very nature of their current employer.
Of note, for the Research Intel Labs China example Irina provided, using LinkedIn’s free search, if you exclude people who mention Intel as a current or past employer from their profile, you find 28 results. If the searcher’s intent was to find people with Intel Labs experience, these 28 are false positives littered throughout the 283 – they mention “Intel” and “Labs” somewhere in their profiles, but not as an employer. If the searcher’s goal was to actually find people who either currently work for Intel Labs or have previously worked at Intel Labs and currently live in China, I would recommend either searching for “Intel Labs” or simply searching for Intel Labs using the current or past employer search field.
Additionally, I am not sure why someone would search for a keyword of [research] when they could simply find and review all of the current and past employees of Intel Labs regardless of the keywords they chose or did not choose to include on their LinkedIn profile – a very manageable <300 results.
Although I am pretty sure the discrepancies in search results across free and paid LinkedIn accounts, at least from the examples I have seen, has something to do with searching for keywords that are also industry terms, I can’t explain exactly why searching for a simple term like [research] produces such huge differences in search results between LIR and a free account (119 vs. 283), other than perhaps LIR doesn’t search the industry field when you use an industry term as a keyword. As soon as you remove [research] from the search (which, remember, is also an industry in LinkedIn), search for a current or past employer of “Intel Labs” and switch to a current title search of [engineer], you pretty much get the same results between LIR and a free search (92).
If the interesting response from LinkedIn posted by Joseph M provides me with any educated guesses, I would say LinkedIn’s free search is “dirtier” and less precise than LinkedIn Recruiter’s, allowing more, but not necessarily (and not likely to be) relevant results to be returned. On other words, until we learn more, my position at the moment is that LinkedIn Recruiter license holders need not worry that they are missing out on relevant search results provided they don’t run very basic/imprecise searches using keywords that are also LinkedIn industry terms.
Have different observations/advice? Please share.