I recently encountered the LinkedIn profile of my old SWE colleague I hadn't heard from in a long time. With curiosity I checked his profile's "Experience" section to see what he's been up to since our paths diverged. As I was eyeing his job history, I scrolled down to the entry about our common past employer. To my great surprise, under it he had listed a bunch of impressive things I know for a fact he had not committed. To give you an idea, things like:
- Implemented and optimized an Apache Kafka solution which improved data reliability by 40%.
- Designed a data warehouse which improved data retrieval speed by 40%.
- Built ETL pipelines with Snowflake, which reduced data processing time by 55%.
I was flabbergasted! This was a junior engineer who was let go because even with a lot of hand-holding he struggled with basic tasks, so that's not just some cute "rounding up" or upselling in his resume, but completely made-up stuff. I had heard of people embellishing their resumes a little, but I didn't know that anyone I know did it that blatantly! That's some con artist level baloney! Coming from a culture which values honesty above all other characteristics, I was bewildered.
As my initial shock wore off though, I started questioning my reaction. Is there much harm done here? A hiring organization should always do its due diligence by verifying a candidate is as advertised, so whatever they've written in their CV should anyway be taken with a grain of salt (or, evidently, in some cases with a spoonful of salt). If you can get more interviews by luring in recruiters with a honeypot of a resume, then good for you! While telling lies on the Internet does play a part in the enshittification of society by eroding general trust in each other, and it punishes the candidates who choose not to lie in their resumes, it's the game we should hate, not the player, right? I myself wouldn't want to do this kind of thing though, as I'm a bad liar and I've learned that life is much less stressful when I don't have to keep a mental track of what lies I've fed to whom. Like the kids (maybe) say, I'm trying my best to keep it 100% 24/7 fr fr 🔥💯👌.
Anyway, while I was contemplating what I had just found, I realized that aside from all the lies, there's a more profound issue with padding your resume with measurements of the impact you've had. In almost all situations, the numbers reveal no useful information! You need more context to draw conclusions on the engineer's skills! For example, let's say I accidentally fudged up this website by doing something silly like loading unminified script files in the <head> of the HTML document. If that made the FCP (First Contentful Paint) take 500ms on average, and I then go fix my own mess and bring FCP down to 100ms, can I justifiably claim that I've optimized the page to load 80% faster? The numbers certainly suggest so, and without context, that improvement might sound impressive. But once you know how trivial the work I did was, it'd be kind of embarrassing to even bring it up.
Even if we're talking about something not as trivial, like "I built a data warehouse which improved data retrieval speed by 40%" it's hard for an outsider to really assess the impact made; did the data warehouse impact all data processing or only certain types of operations? How big of a scale are we talking about here? Was the original data processing pipeline unnecessarily slow and you could have achieved similar impact with less work (i.e., did you kill a fly with a bazooka)? Sure, statements like this can work as conversation starters — let's say an interviewer asks you to expand on some of these tidbits on your resume — but, again, without the added context figures like this are trivia, no matter how impressive they might sound as standalone statements. The thing that irks me though is that if you know impact percentages are flawed, you might still be better off adding them to your resume than not. Some recruiters tend to love signals like this, because I guess it makes it easier for them to justify pushing a candidate into the recruitment pipeline (don't even get me started on the whole "data-driven decision making" meme1). So even if you know resume padding like this is bullshit, the game is rigged in a way you might be forced to add numeric impact data in order to gain a leg up against your competitors.
In all honesty though, job seeking is full of BS and jumping through hoops like this — it's not unlike mating rituals of certain animal species, who have to complete all sorts of trials to find a match.
For people interviewing candidates, I suppose the takeaway is that you should never blindly trust in what a candidate says they've done — only trust what you can verify, whether it's their past accomplishments or the skills they showcase during the interview process. That's nothing new to seasoned interviewers. But going back to straight-up lies in a resume; is there a way to catch them in initial screening, or is it inevitable that brashness improves one's chances of moving up in the candidate stack? From my perspective, technical recruiters seem to lack the skills to do that kind of verification, so it's up to the engineers who interview the candidates to catch the lies.
One might think that recruiters talk with each other though, and that if a candidate gets caught lying too much on their resume they'd be blacklisted for a lot of firms. I've interviewed candidates but never done talent sourcing, so I'm unsure if this is common practice; a quick round of googling suggests it's not, at least not in the US. Maybe there should be a common registry of lying candidates then? I doubt that'd be legal in the EU under the GDPR though, but come to think of it, there are probably very good reasons not to have one elsewhere either — think of the consequences of pissing off a recruiter with the power to add you on a national no-hiring list. To quote Kanye: "No one man should have all that power"2.
Footnotes
-
Interpreting statistics can be notoriously difficult (correlation vs. causation and yadda yadda), even if you have the technical knowledge to do it right (which I don't, really).
On one hand, if you know what you're looking for and trust the quality of the data, leaning in on statistics can be a useful way to streamline decision-making: "we've A/B tested between a blue and red button, and people tend to click the red button 50% more often, and that's really the only thing we're interested in, so let's go with the red button". (This is what most people endorsing data-driven decision making think they're doing).
On the other hand, in real-world business situations data integrity can be dubious, and most often you're using weak proxies to infer the actual phenomenon you're interested in. For example: "we launched a campaign asking user feedback on their favorite colors, and 70% of them said it's red, so that means if we make the button red vs. blue, it's sure to drive more clicks".
And those two scenarios don't even cover the case of people pulling for their own agenda. For example, the manager of the team who implemented the red button might perceive its success as an important factor to getting promoted, so they might come up with any kind of wonky (but plausible-sounding) statistics that support the choice of red instead of blue. ↩ -
Interesting fact of the day: as I was writing this, I learned that this hook in Kanye's POWER is actually a quote by a police officer on Malcolm X. ↩