Show HN: Startup jobs of 2018, visualized | Hacker News Reader

Show HN: Startup jobs of 2018, visualized

startup.jobs

77 pointsmarc38 comments

It's great that you have lots of data to be able to analyze, but I think you've missed the point with that you're analyzing. I'm not sure what the visualizing is doing to offer any insight? Just graphs of months where it goes up or down at a slight varied rate, and a geo graph showing that (obviously) CA and NY have the most job openings?

I think the only interesting "insight" (If you could call it that) is the popularity of different technologies (though not how its changed in 2018 vs 2017, or how it differs by company, or salary range).

Who's the audience - those looking for jobs in startups? Those who research the industry? Job posting sites use their data as a leading economic indicator - can you point to anything in this realm from this data?

Analysis that would make this more insightful and actionable would include:

a) Show the data compared to previous years, to say for example if hiring by month was actually cyclical

b) Jobs by startup - ie, what percentage of Carvana jobs are eng, sales, finance etc

c) No salary information analyzed - probably the most interesting data is missing

d) A table of top 10 countries by no. postings vs their previous year would give more info to your geo graph

e) Is startup appetite for remote workers growing over time? As a number of jobs and as a percentage of jobs - and what areas are there higher % of jobs going to remote workers?

A good example of a valuable data company is Statista [1] (links to how a geo graph works with a table), where they have this sort of data over a longer time horizon - which provides more actionable insights (eg, is my industry growing, what is the penetration rate?).

If there were more ways this pie was sliced then I could see many interesting insights coming from such a useful dataset!

[1] https://www.statista.com/outlook/15000000/109/consumer-elect...

These are some great points. I appreciate you taking the time to write them down. I'll see which of them I can implement.

a) Comparison with previous years will be totally doable. As a quick hack you can simply compare with https://startup.jobs/trends/2017 – but I understand ideally you want the graphs to show the comparison.

b) Jobs by Startup. This is indeed something I'm planning to add to the Startup pages like http://startup.jobs/@carvana

c) I agree salary information would be super interesting. Unfortunately very little of that is public information so I simply don't have it.

d) Data goes back to mid 2016 so we don't have a ton of historical data. But for next year I could look into historical line graphs for top 10 companies.

e) I like the idea of cross referencing sector/skill and remote work. I do have the data to compare e.g. engineering jobs versus communication/PR jobs in terms of remote job openings.

Thanks again!

Thanks for this, yes I think these could be good ideas. Just quickly on d) I meant it would be visually useful to see a table showing the data for the top 5 or 10 countries too, or else you have to hover over each country to see, and then you could show vs 2017 to see if there has been a large increase in a given country for example.

Understandable about salary data, a shame though! Thanks a lot!

I was curious as to why neither C# nor .NET appeared on the "Top Technologies" graph. Then I realized it apparently combines C, C++, and C# into a single bucket, given that the same tag [1] also shows jobs for all three.

[1] https://startup.jobs/c

This website is a great example of the challenges with being data-driven. It leads to a conclusion which is terribly flawed because of a failure to account for relevant nuance in the data.

C# is definitely not C++ or C. It would be better to lump it with Java.

But how many developers are now going to walk away thinking they should bone up on C since there are apparently so many positions out there for it?

And how many other statistics out there are similarly flawed? How much of our day-to-day off-the-cuff knowledge is based on poor data entry?

Feels weird to see Microsoft as a technology in the top technologies diagram. Also it is surprising that there are many job posts for assembly.

I was also surprised that the word Docker or Containers didn't make the list in that section.

If you drill down the assembly jobs seems related to manufacturing assembly.

Ha, that makes sense but also doesn't, manufacturing assembly as a technology?

Remember this is a list of startups which doesn't always imply technology. And even technology companies need people to manufacture and assembly stuff.

I'm getting 404'd on many pages

https://startup.jobs/@spacex

https://startup.jobs/@shopify

I'd be curious to know what the ratio of location dependent remote jobs compared to world wide remote jobs are. I suspect that the latter are practically non-existent (i.e. 10s total at best), but you never know...

Over 90% senior positions is very interesting to me.

I'm pretty sure this is a result of non-technical people calling the shots on what roles they're hiring for.

  ROLE: Sr. Full Stack Engineer Level 5
  MUST HAVE:
  10+ YEARS JAVASCRIPT EXPERIENCE
  4+ YEARS JSON EXPERIENCE
  7+ YEARS PERL ON RAILS EXPERIENCE
  Linux
  Sketch / Adobe Photoshop / InVision
  PhD in Machine Learning

And not to soapbox here too much but but sadly I actually have come across a job ad demanding "4+ years JSON experience"

Don't you love it when companies signal that you really probably shouldn't work there

I wouldn't count out a company based solely on a poorly-written job description. Those are often put together by less-technical recruiters or the HR department. And yeah, maybe the team that's hiring should be more involved in the actual job posting, but the whole hiring process is such a huge PITA and time-suck that they're too happy to hand off whatever part they can.

> 7+ YEARS PERL ON RAILS EXPERIENCE

What's "Perl on Rails"? Isn't it supposed to be "Ruby on Rails"?

"Job posts not specifying a seniority level are excluded."

13824 senior positions / 120548 total openings = 11.5% senior. If we assume "senior" only applies to "dev" and "engineer it becomes 13824/(15484+9723) = 54.8%. Somewhere between those sounds right for ratio of senior position ads I've seen.

West Virginia really had no postings?

I double checked our database and there were indeed no postings for West Virginia. There were 2 in 2017.

This is great. What tools/languages have you used to build it?

Unfortunately I'm getting a 404 not found. Looking forward to see it!

Hugged to 404?

Afraid so. The 404 is happening because the server cannot find a 50x.html file.

I'm adding caching as we speak. Expect the site to be back up in a few minutes.

…aanndddd we're back!

Thanks!

Linkedin and Netflix are still considered startups?

Developer of Startup Jobs here. We use a pretty broad definition of the word startup. In future analysis I'll see if we can differentiate between early-stage, later-stage, etc.

LinkedIn is straight up a Microsoft property now, so it's not even considered "late-stage" anymore.

Netflix was founded in the late 90s, so if you consider them a startup, then the term is pretty much meaningless.

This really puts your overall methodology in doubt.

I agree that Netflix isn't a startup, but I'd say that looking at their founding date is misleading -- Netflix-the-streaming-video-service is an almost completely different company from Netflix-the-DVDs-in-the-mail company.

Some good points. We don't concrete criteria for which companies we consider startups, but it's something along the lines of "is this a technology company that is disrupting the marketing, and/or rapidly evolving its business model"?

I do think that's the case for Netflix, even though technically it was founded a while ago.

LinkedIn is a bit harder to defend as its business model seems to have stagnated.

I wouldn't think any publicly traded company, nor any subsidiary of a non-startup, qualifies as a startup, even if they previously did before they exited. The tough categorization calls in my opinion are for slow-growing or stagnant long-private tech companies.

With your definition, doesn't Amazon still qualify? It's majorly disrupting markets and rapidly evolving its business model. And it's definitely a tech company.

Netflix is still pre-profit and growth oriented in an attempt to capture/develop a market, and its value is based on speculations about it succeeding. It isn't quite a startup, but its also a bit of a startup... Correct me if I'm wrong.

Not to mention that Netflix is also publicly traded. Personally, I would not consider any public company to be a startup.

A definition that includes Playstation, which is a 24 year old brand created by Sony, a 72 year old public multinational corporation. It seems like your broad definition is any company that hires web or mobile developers and isn't Facebook, Google, Amazon or Apple (can't say FAANG because Netflix is on there)

was surprised to see java as the #3 technology behind iOS and android, but this explains it.

I was surprised to see the top technologies be what they are. It differs from some other lists I've seen.

120,000 jobs is almost within itself the size of Apple.

I mean unless these jobs are serving totally unique products and market otherwise I can envision a future where all these jobs overlap in some form and coagulate into a single giant startup.

Reply on news.ycombinator.com