The lead-up
Our team has generated $269,256,833 from engineering job offers at 1,658 different tech companies. It took over 57,000 interviews to get there. Unless stated otherwise, we are providing the numbers in this article by sampling from our data across 57,000 interviews. We are a mixture of engineers, coaches, product, and leadership folks, and we’re excited to demonstrate data-based insights you can apply to help get a job you love.
In the darkest night
According to LinkedIn, the number of Software Engineering (SWE) jobs in 2023 has gone down: 250,000 in May, 135,000 in August, and 122,000 in September. USBLS details the highest unemployment rate since Feb 2022. You can’t leave the house for a Yoo-hoo without hearing someone say “The job market is tough right now.”
We won’t mention it again because it’s frankly, unhelpful. Most candidates don’t control when they enter the market. Like it or not, you’re here, so instead of focusing on some massive force you can’t change like the market, gravity, or the fact that you can’t find Yoo-hoo anywhere anymore, instead, focus on what you have power over, namely, the data-backed strategy of your job search.
Your mileage may vary
If you’ve got less than 2 years of experience, assume you’ll fare worse than the conversion rates discussed later on in this article. Fret not, one day you’ll be a big bad senior who gets their pick of the litter of companies. And until then… godspeed!
Lastly, interview skills between programmers can vary widely. If interview skill is quite high or quite low you’ll fall outside of the normal distribution, calling for different tactics. Our data points are best viewed as a rule of thumb, for the 80% of engineers out there. These conversion rates are averages, you may fare better or worse.
The flawed numbers game
The average engineer: takes 2-5 minutes* to fill out an application, sends out 70 applications to get 1 interview, and has to take ~20 interviews before getting a job. That’s an estimated 46 – 116 hours spent applying to 1,400 jobs, and that’s just to get ONE offer! (And that’s not accounting any time for interview prep!!)
Note: The 2 – 5 minutes to fill out an app does not account for the time it takes to source the role, read the role, and cross the mental hurdle of putting yourself out there. Depending on your speed, this could have each application reaching the 10+ minute range.
That formula assumes you will:
- Send 70 apps per week consistently
- Send applications only if the role is relevant to your background
- Use a variety of job search platforms
- Avoid shortcuts like “Easy Apply”
- And move extremely quickly at a pace of 2 minutes per app.
Most engineers do none of those things.
As much as engineers may find modern software engineering interviews unsatisfactory, there is significantly more love lost for the act of getting the interview. The royal court of doing interviews is studying, but for getting interviews it’s basically mindless work. To make the mindlessness harder, for many engineers the hardest part is getting interviews.
Getting interviews (the traditional way), and interview preparation are conflicting skill sets; ApplyPass will get you 4 – 8 interviews per month and save you 8-12 hours a month while doing all those things we just said most engineers don’t do. Use that extra time to prepare for those interviews, hunt down a Yoo-hoo, harvest butterflies, or whatever gets you going.
The numbers game reimagined
A crucial moment in the early days of Facebook was learning if a user racks up seven friends they’ll stay on the platform. It became their north star: get every user 7 friends. When you can distill growth down to one focal point, unsurprisingly, it makes it easier to focus.
The job search is noisy: “Should I redo my resume? I’m thinking of going after Fullstack roles from now on. Who was that colleague who worked at Adobe? Maybe I really should redo my resume.” Done in a silo with no guide, the noise of a job search can be greatly amplified. A singular focal point is one way to cut through the noise.
The magic numbers
We’ve got a lot of data on interviews, in fact over 57,000, but these are the data points with the most practical application. This post will first demonstrate the data, and then teach you how to strategically use these data points as a guide.
Infamy in recruiter screens
On average, engineers pass 32% of recruiter screens. However, the top and the bottom of the distribution curve are living completely different lives; the top 25% pass recruiter screens at a rate of 64.48%–and the bottom 25% pass only 5.54% of the time. An average failure rate is ~70%; due to the lack of domain knowledge a “technical” recruiter has in software engineering, one wonders (though companies don’t track it), what is the average false negative rate? And for the folks failing 95% of the time, those rejections may be happening for reasons so small you might not even notice.
Five is the magic number of practice interviews
Pathrise’s dataset showed most engineers fail their first 4-5 onsites. Triplebyte found you have a 40% chance to get at least one offer on your first onsite, and an ~80% chance on your fifth onsite. But after five onsites, the probability stays consistent at 80-85%. Interviewing.io also found that chances doubled after 5 repetitions of practice. After completing 5 mock interviews you’re twice as likely to pass a Facebook tech screen.
Hardest to like, hardest to get
It makes sense to be oriented around the milestone of 20 interviews, since it takes on average 20 interviews to get one offer. However, most candidates don’t like their first offer, in fact, it’s not uncommon to be the worst offer they receive. Controlling for the outliers and counting all first rounds equally (recruiter screens, coding challenges, etc), engineers who get 2 offers take an average of 32 interviews.
If it takes 20 interviews to get your first offer, and only 12 more to get your 2nd, that means the quicker you get to your first offer the fewer interviews it takes to get to subsequent offers.
It also suggests your first offer is the hardest to get. Anyone who’s gone to an interview with another offer in hand knows this; the 2nd offer is easier to get because you don’t need it, you’re confident with nothing to lose, and have the skill to perform at a high level! Speculatively, I’d imagine that trend continues; the 3rd offer is easier to get than the 2nd, the 4th easier than the 3rd, and so on. But this analysis focused on folks with up to 2 offers.
The clock is ticking
It takes an average of 142 days (4.73 months) to receive the first offer. And the median days till the first offer (86) are lower than the average days (142), so there are more outliers on the right of the distribution (engineers who take a much longer time to get an offer.) That makes sense because the time it takes to get that first offer varies widely. For those who get their first offer ~twice as fast as the average, they require on average 27 interviews. The ones who are ~twice as slow averaged 48 interviews to clock their first offer.
Since the data set varies widely, skews to the right and the twice-as-fast engineers are more than 2X as efficient than the twice-as-slow ones, this suggests a compounding effect. An actual name has been put to this by coding bootcamps; if it takes longer than 6 months to find a job, an engineer can fall into what’s called the “doom loop.” Which is a cycle of rejections and lowered confidence, where the more time that passes, the more productivity drops.
How to use the magic numbers
Three tactics will help reach our data-based strategies:
- Pick an interview threshold
- Set a time boundary
- Get unstuck with heuristics.
Pick an interview threshold
We all want offers, but getting interviews is more controllable than getting offers, so it makes sense to orient your focus to getting interviews. An interview threshold is a milestone and focal point to cut through the noise. Once you achieve an interview threshold, the data show you’ll have gotten an offer or be very close to landing one.
Pick an interview threshold, and then make hitting your interview threshold your singular point of focus, as Facebook did with “get each user to 7 friends on the platform.” If you don’t know what number of interviews to pick for your threshold, give your best guess. Factor in if you’re aiming for one or two (or more!) offers, and how many interviews you did in your last job search. If you can’t remember, look back in your calendar as most interviews have an invite or review the data from whatever tool you used in your last job search.
If you don’t know what to pick, go with 35 as it’s smack dab between average and a worst case to get one offer. Plus, it’s quite near 32 which is how many interviews it takes to get 2 offers on average.
Set a time boundary
When do you want to hit your interview threshold? Don’t say “tomorrow”, it’s unrealistic, and don’t say “ASAP” because it can’t be measured. Remember, 4.7 months is average, ~2.3 months is fast, and things start to get hairy after 6 months. If things don’t go as planned, never be consistently looking for more than 6 months. After every 5-6 months (at most) take a real break, for at least a few weeks, and most importantly, if you’re stuck, it’s essential to uncover what’s going wrong.
Use heuristics to know when you’re stuck
Most of the time, unwanted outliers happen in a job search because someone gets stuck. Whether they take 48 interviews to get one offer, or take 7 months to get one offer, usually part of the problem is their lack of awareness that there is a problem in the first place. They usually don’t have benchmarks to which to compare themselves. If you’ve ever wondered “Am I on the right track?” in your job search, or “What am I doing wrong?”, then these heuristics are for you. Use them as a benchmark to track progress, or as a warning sign for job search health.
A sign you might be stuck at the top of the funnel is passing less than 30% of recruiter screens. Or, if you want to compete with the top candidates, passing anything less than 50-60% of recruiter screens. A sign you might be stuck at the bottom of the funnel is failing more than 5 onsites without receiving an offer. Whether you’re stuck at the top or the bottom, the optimal solution is to have someone who knows what they’re talking about give you feedback.
This heuristic lets you answer one of the most common questions in the engineering job search: “Am I ready to interview at EvilMegaCorp?” If you’re unsure of whether or not you’re ready for the next round, it’s a simple formula:
The important piece is to make the practice interviews relevant to your goals. For example, if you want to get a job at Google, it won’t help you to do practical take home projects because Google’s process has none. Google’s coding rounds are complex algorithmic-based questions you probably won’t find on leetcode. A relevant practice round doesn’t need to match 100%, but at the very least it’d be a coding screen with an algorithmic based question.
Self-doubt can make it so you never feel ready for an important interview, but the best way we know to actually measure readiness is to complete 5 similar practice rounds because after completing 5 onsites, or 5 mock interviews, your chances of passing double.
The skill of interviewing
Courtesy of The Change Kit.
The reality is that each time you start a job search, you’re starting over. You have to build your competence up from the baseline. As the data shows, getting an offer is easier if you do lots of interviews in a smaller time frame.
How you go about getting those interviews is up to you. We’re obviously biased, but we see ApplyPass as the best option. For a few reasons, the most important being our team has expertise in engineering hiring from more than 57,000 interviews. We baked that knowledge into our product, an AI with a human touch, to get interviews for you; you fill out a form and we do the rest. Try our 1-week free trial, which on average gets users 1-2 new interviews.
