Attention HN: What's the Best Way to Find a Good Dev for Your Startup?
I’m a startup founder who recently secured VC funding. As a technical founder, I understand the realistic and unrealistic expectations when working with developers. Over the past few weeks, I've been searching for my first dev, specifically a Next.js Full Stack Developer. I posted the job on LinkedIn and received over 1200 applications in the first week (I manually went through each and every one!). After screening, I narrowed it down to about 80 candidates for interviews (Again I was being very generous with screening and anyone I thought who even remotely had a chance got a interview).
However, my interview experience has been quite disappointing. Out of the 80+ interviews, only 4 candidates had portfolio websites, which is quite surprising for a Next.js dev role. Even more frustrating was discovering that all 4 had copied the same tutorial for their portfolios. It's not hard to find out that people copied a tutorial when they all look the exact same...
When interviewing I try to avoid asking Leetcode questions because in my opinion they are scary and intimidating and unrealistic. My interviews are scenario-based, where I present a problem and ask the dev to walk me through a solution verbally (NO CODE!). These problems are usually straightforward but help me test the persons problem solving skills, often involving scenarios like retrieving data from an S3 bucket in AWS but when bringing them on the website having issues with rendering. Despite this, less than 5 candidates could solve the scenarios. The meme of putting "Software Engineer" on your resume after watching a hello world tutorial became a little bit more true after that experience. I am not trying to be mean or anything. I gave everyone a very fair chance.
I’m starting to wonder if the era of good developers is over, or if my approach to finding them is flawed. Some other founders have suggested using take-home assignments that are 24 hours long instead of traditional interviews for this role.
For those of you who have been hiring React/Next/Vue/JS developers for a while, how do you find the "good" devs? Any tips or strategies would be greatly appreciated.
Thanks!
40 comments
[ 2.1 ms ] story [ 87.3 ms ] threadAdditionally, I know many great developers who do not have portfolio websites.
I'm not sure if it's the wrong approach, though. Some like doing things end to end, and these will find their place in small companies.
I haven't used it myself, but I consider myself a "good" dev and get to be picky about what projects I work on. This is the kind of job opportunity I would respond to.
Notice Dane's entire approach is centered on what the "good" devs value and are interested in: their fears and desires.
Also Dane approaches specific "top" devs directly (vs blasting out job postings and wading through mediocre applications.)
I really liked this part:
I demand the best work from you. I expect the best from you. You will be challenged. You will grow. And your skills will sharpen. If you like to settle on being average and do average work, this is not your project.
Expects the best, pays $50 per hour.
That dev is now the CEO of the business:
- https://www.linkedin.com/in/rajesh-dhawan-68702388/
- https://www.paperlesspipeline.com/
So it must have worked at least once.
Dane used this process over ten times. Paperless Pipeline is probably the most successful example and semi-documented.
---
I think if you offer an appealing opportunity, you don't have to offer the most money. There are many other things you can offer (as outlined in that PDF).
Recently I almost took a project for 40% of my normal rate. Fortunately things worked out and I negotiated 80% of my rate, even though I should probably increase my rate.
The document itself I agree has some parts which are a bit over-the-top but that's probably the case for all hiring advice. When you list the qualities of the perfect candidate out on paper it's guaranteed to look ridiculous.
https://web.archive.org/web/20131208173349/http://thefoundat...
For example extremely experienced engineer with wide market exposure and hi empathy can manually review CVs that passed some basic filters for quality.
Someone capable of quickly putting himself in each candidate shoes based on the line in CV.
That will produce a small but hi quality bunch of candidates to talk to.
Another thing (besides lack of portfolio being a weak data point) - some very strong engineers are very nervous and suck with live coding (rightfully so) they prefer take home assignments where they have time to carefully think about the solution.
Also best engineers are often worst in selling themselves (because they spent their 10k hours engineering, not selling or leetcoding).
Best engineers are also stack agnostic. Emphasizing specific new shiny thing in job posting can be a deterrent, and in their cv you would rarely find “nextjs”, because for them it’s just another library. Not something to brag about or even mention in cv.
To attract them - spend more time describing the problem, the challenge, the value you gonna create and less - tech stack and constraints.
What and why, not how.
Proof or it didn't happen. I have never met--or even heard of until now--a developer that preferred this.
What’s so surprising?
And what other alternative can even remotely represent actual work conditions that a candidate would have?
Id say anything ~1 work day is a good upper limit.
2. ~8h of work should be an estimated task scope for average developer on position you hiring for, or for candidate’s experience
3. Ideally it is an option to choose from (“you want live coding sessions or take home assignments?”)
Why not look for promising devs and develop them?
Here are some of the lessons I have learned over the years (Hiring on and off for almost a decade now):
- Portfolio sites are crap. They are built by people who have no real experience and need it to try and break in the tech world. Theya re mostly "bootcampers" or self taught devs who aqre too junior. You mayfind a few gems but extremely unlikely.
- Most "good" devs are taken, let alone great ones. To add to it, most good/great devs are not interested in risky startups. So your candidate pool is extremely small already. You cannot just post a job and expect to find the gems that would be a fit for startups. You most likely won't not for the first few hires that matter.
- The only way is to tap into your own network and experiences. For example, I was able to convince an ex-co worker of mine to join my company as I knew he would a great fit. He wouldn't have applied to any jobs but he gladly took me up on my offer as we worked together for a few years in our corporate jobs way back.
- You also need to sell your company/product/vision. Get online everywhere. Have a twitter. Be social. Talk about your company/product/goals etc. It may attract some good people. It may.
- Your first 5-10 employees need to be dreamers, hustlers, romantics. They cannot be someone who jut wants a job. Won't work. Never does. Good luck.
It's also surprising that only 4/80 had a website to show you. I would expect that most web devs have a personal site or Github to show you. I did that prior to my last internal team move going from mostly backend work to mostly front-end work. It seemed to make a good impression for the manager that I had a working site with a variety of components and some customized CSS. It's also a real site for advertising my honey and other apiary products. I also wonder if the sites you mentioned were really identical, or if the framework and css templates just made them look similar? It's really easy to see a material or bootstrap site and say it looks like any other material of bootstrap site.
However, this might also mean that OP is using a bad pipeline, probably craigslist or something similarly cheap/free where good devs don't hang around.
There is a higher bar for developers on Upwork or Toptal but they are very expensive.
But since you want great developers (ideally senior developers) and have the VC funding to do so, this shouldn't be a problem.
NextJS/React covers a massive talent pool, so you will inevitably get less experienced candidates going through your pipeline until a good one shows up.
You might not spend as much time if you were tapping into the markets of say Clojure or Elixir, maybe even Go. Not telling you to change your stack, just highlighting the differences in talent pool size and quality.
Be transparent with the compensation from the get go. If you are trying to tap LATAM markets, good devs are reaching (close to, but still a bit far) US salaries, so offering 70-80k for a Senior role won’t guarantee a Senior dev.
Second, is the job description a turn off? I've had friends ask me why they can't get applicants. They showed me the job desc, which sounded a lot like looking for a part for a machine. I pointed out everything I didn't like about it and she said that everyone else does the same so they'll keep it to not seem weird.
You can't do this at a startup. Be weird. VCs expect you to be building things OpenAI can't and won't; you're already insane by definition. There's hundreds of founders out there in this era of AI who are looking for adventure and you have to stand out to the applicants as well.
I actually checked out your product and sorry to say, it's not something I want to be doing either. If I wanted to do finance & AI research, I'm well funded and well salaried where I am. Try to make it seem attractive. There are also people who are happy to work 6 day weeks, but you have to prove that you're a winner.
Third, check that you're not fishing in a puddle. Job sites are absolutely terrible, 1 in 200 can't do basic things. LinkedIn is okay, but you're fishing next to the biggest companies in the world. There's indie sites that prioritise remote or 4 day jobs, try those. HN will have a Who is Hiring thread every month; the quality is very high on both sides here.
If you're looking for a code-monkey - don't