I’m all for optimism – my coworkers comment all the time on how positive of a person I am. But sometimes when it comes to sourcing, you have to be pessimistic.
What do I mean by that? I mean that when it comes to writing a Boolean string, you can’t look through rose colored glasses (I’ve often been guilty of this). Instead, it can be beneficial to take a pessimist’s view.
Least Common Denominator
Look at your Boolean string and think, “What’s the least on-target profile that could come up with this search string?” I’ve found that examining your Boolean with pessimism will give you a better search. Plus, it will save you time in the end.
Thankfully it doesn’t end with pessimism. The next important step I’ve found is to actually resolve the issue. I ask, “What can I add to my Boolean to eliminate that kind of off-target profile?”
You could add a “NOT” statement, or maybe another “AND” statement – whatever is the most effective way to resolve the problem.
We also have to be careful to avoid being too strict (I have also been guilty of this). The last thing you want is to not limit the spread of your profile-elimination spree. Then you could be removing candidates from your search that would have lined up well for the role.
If you’re a Boolean nerd like me, you could get lost in the rabbit hole of repeating this till the end of time. But we don’t have till the end of time!
Since we all have limited time, we have to determine where the rabbit hole stops. It’s important to do a cost benefit analysis of time spent sourcing vs. time spent sifting through profiles. Examine how you can save time (aka: be thoughtfully lazy) to be the most efficient.
Before you get lost in the clouds of airy methodology, let me show you what this really looks like. This is an example of a partial search with some basic Boolean using SeekOut’s field search.
cur_title:((software OR application OR Go OR Golang OR backend OR “back-end” OR “back end” OR Python OR ML OR “machine learning” OR engineer) AND (engineer OR developer OR programmer))
AND (“machine learning platform” OR “machine learning infrastructure” OR “model serving”)
So I ask, what kind of profiles could come up that I don’t want? Besides the obvious Manager or Intern titles, this could bring up people with the title of “Machine Learning / Data Scientist”. It’s likely that I don’t want to see those kinds of profiles. This is a very junior Machine Learning Software Eng role.
Since I’ve determined what could come up that I don’t want to see, I ask, “how can I fix this issue?” In this case, the unwanted profiles can be eliminated with a simple NOT cur_title:”Data Scientist” statement.
It’s also possible that people with the title “ML Engineer” could come up that have nothing to do with Machine Learning! It is just a two letter abbreviation. Candidates could mean a “Messy Language Engineer” or some other nonsense.
I ran the search before changing any of my Boolean and scanned the results. I looked at what kind of profiles had “ML Engineer” as a title and determined there weren’t enough off target with that title. No need to spend excessive time to eliminate the nuanced differences.
Article Continues Below
AI and Automation: How They Will Impact the Future of Recruiting?
Let’s look at another example, this time it’s a search for a Supply Chain Manager.
cur_title:((“supply chain” OR procurement OR sourcing OR purchasing) AND (manager))
AND (“procurement support” OR “supply chain management”)
I scanned through the results and found several people that seemed too focused on arranging the transportation of inventory. Let’s say for this role we don’t want to see that type of candidate. Instead of scrolling through and eliminating them one by one, we might try knocking some of them out by adding this to our Boolean.
NOT (transportation OR freight OR 3pl OR truck*)
It’s possible that adding the above “NOT” phrase eliminates candidates that could be a good fit. Ultimately it’s up to you to decide what you should eliminate from your pool or not based on the candidate you know you need, and what you are seeing show up in your search results. Focus on examining how you can be stricter with your Boolean to increase your sourcing efficiency and save you time on the back-end of going through hundreds of unnecessary profiles.
When removing a group of keywords (or profiles) from your results, make certain that they do not overlap with profiles that are a fit for your search. For example, removing words like ‘manager’ from an individual contributor search could remove profiles that just happen to mention the word ‘manager’ but are really individual contributors. Think of a Boolean diagram for 3 possible combinations of traits: 2 traits you want to have and 1 you do not want. The words you remove should only affect the profiles that do not have both of the desired traits.
This approach wields pessimism and efficiency in your problem solving. My aim is that you’ll be able to spend more time on the good candidates and less on the ones that are not a fit.
And that is a wrap on the power of pessimism in sourcing! I hope you’ll be looking at your Boolean strings with more criticism in the future. All to save yourself time in the end.