Given how wildly the framework usage [1] differs from the long-standing and well known State of JS survey results [2], I'd be fairly wary of anything else they claim. Not saying the results are wrong, probably just different samples, more so that it seems the variance is way too high to get any useful insights.
This is a report for agencies. They cover a completely different set of customers and most of the time are called in to maintain. Maintaining leaves a little space to introduce new technologies and processes, which leaves you with older technologies.
Also, these companies have no reason to switch to other (newer/shinier) technologies since the software they deliver is tied closely with the chosen language. Switching would mean finding new customers.
State of JS is also biased to the group that opts to fill the State of JS survey. I do think it's a great survey with real value, but it's quite likely it's own distinct group too.
The survey in this article claims to have reached about 1000 times more people (though not all JS devs) than the stateofjs survey. I would expect some differences due to selection bias
The first pages made me very skeptical, so I stop to read this. Beyond that there is no data intelligence on these reports, the numbers are weird. It's showing Colombia's hourly rate more expensive than Sweden's. Also, I know very much my homeland market, I am sure that it's wrong. I did not go further.
It’s a clickbait article... any article that claims something about rates, salaries, sales is basically that. They are often completely inaccurate, with data coming from either self reported numbers (best case), or a guess based on a sample of job postings they scraped from somewhere.
Yeah, the whole thing looks to be basically an ad for agencies that have registered with them.
Nice for the agencies, not all that useful to this crowd for data collection. Alot of the numbers are just way, way off from industry standards so it makes me wonder where they got them.
I oversee the hiring of dev shops for a mid-sized company in the US and the rates for e-commerce/web development are all in the $120-$150/hour range.
The funny thing is that some of these companies outsourced to the Ukraine and India and the rates didn't change. I would have thought that we would at least see some of the cost savings.
Both companies that outsourced ended up being complete disasters (we are still cleaning up the technical debt 3+ years later), which happened before I was involved.
Since this is the case, it made no sense to hire any company that didn't hire all developers directly in the US or Canada (which is now a requirement).
I suppose I should thank these companies, because it's the reason I have a job today.
The concept of outsourcing for a writing job is weird. You don't see The New Yorker outsourcing their essays to Pakistan, Hollywood outsourcing their scripts to Ukraine, the NYT outsourcing their reporting to the Philippines, etc.
Tech seems to be the only industry that thinks that hiring writers from overseas to produce content is somehow going to result in anything viable.
If you read the articles, I would say that the companies listed above (like the NYT) are not the ones outsourcing, the outsourcing is quite limited in scope, and mostly is design jobs, not writing jobs.
It's not that weird if you consider that a large chunk of software engineering is CRUD apps and digital plumbing, which doesn't require a lot of creativity. Nevertheless, I'm sure India, Ukraine and other countries have tons of talented developers that due to global economics are willing to work for a lower pay. Maybe the problem with the outsourcing model in these cases is that it's being poorly executed. It's already difficult enough to build a strong team within your own city, it wouldn't be any easier to do it remotely across the world, much less if you delegate it to an HR department purely focused on costs reduction.
> a large chunk of software engineering is CRUD apps and digital plumbing, which doesn't require a lot of creativity
The issue is more precision and lucidity of thought and language, and the ability to communicate ideas and emotion to others from a given culture and socioeconomic background.
Consider that when development velocity grinds to a halt because of "technical debt", this is almost always just a euphemism for developers not being able to read each other's code. (I understand there is an actual idea behind technical debt, I'm just talking about what's almost always happening in practice.)
as a 3rd world developer now living comfortably on the 1st world this is exactly my experience, people in the developing world are in general more ignorant on technology, so even if you are an excellent developer it is still going to be a hard sell to ask for a US dev rate. The emerging economies are not technology driven in most cases. In my case I had to look for work outside my own country first to more reputable South American countries and eventually in North America.
I think the problem with outsourcing is it adds a layer of indirection between the stakeholders and the developers. As a developer myself, I find it almost impossible to work effectively unless I have direct contact with the stakeholders.
However if you have architects that design software properly in the U.S. and pass those specifications to outsourced staff, then it's akin to a NYT writer dictating his notes to be typed in and proofread by someone overseas.
My point being; there's an efficient way to utilize cheap labor.
Right, but that proofreader might not totally understand nuances in language and how words take on different meanings to different audiences, even if they are the same words used.
Back to the programming reality from the analogy, I've seen outsourcing happen, for businesses that revolve around American customer bases and rules and regulations- now the code is outsourced, but the basic understanding of how business logic related to the real world use was gone. So the offshore team, while technically competent, was not equipped to make decisions about software with the consumer in mind. If you were working on something and it didn't really make sense in a real world sense, as a dev, you could go to your product person and raise concerns. Without the cultural context- those concerns never get raised.
Agreed. So in that case, I would not use an outsourcer for that particular piece. The problem is American developers treat outsourcers like they are American developers. Don't do that.
Treat them like 1) there is a language barrier and 2) they will not know there is a language barrier and they will forge ahead in the wrong direction.
Some of the best developers and PMs I have worked with were born in a country you barely know about, it just took a little time to get to know them and their skillset.
The best way to combat that is to not design anything too far into the future. This is where agile and sprints can be very helpful. It also means you are checking the work of the subcontractors on a regular basis. After a while you will learn how to communicate very well, you'll see the strengths and weaknesses of the staff and you can operate within those constraints.
That is the real work here; discovering acumen. Once you know that, you can hand out work that fits within their wheelhouse.
American developers excel at this because there's less of a communication gap. If an American can not do something he will tell you. If you are wrong about the way you are doing something, an American developer will tell you. Outsourced subcontractors will blindly do what they understood they were supposed to do without questioning anything resembling rationale.
This is not the way to look at it. Software engineering and content writing are not even close to being similar in any sense. You don’t need cultural context to write good software
I manage an outsourced team having worked in Latin America. I agree with the other commenters here. It will fail if there's no quality control and no filtering. Just like in the US, there are good devs and bad devs. It's up to the agency to actually manage and help them grow.
> The funny thing is that some of these companies outsourced to the Ukraine and India and the rates didn't change. I would have thought that we would at least see some of the cost savings.
Why would the rates be lower if the market is one where demand for labor exceeds supply? It's not like lowering the rate is going to allow them to sell more hours, since they are still constrained by human capacity. If the quality of the work was not good, that's a different issue.
In this kind of discussion context, typically the OP's rates are the bill rates to clients (the OP's company). I normally see the smaller agencies clear 40-50% margin, and the big names are clearing 50-80%, so I wouldn't be surprised if those agencies in Ukraine and India were paying their teams in the $60-75 per hour USD range at most, which is considered really good in those parts of the world.
Margins are much higher. Large scale IT Services pay around 40$-60$/day for skilled specialists.
I briefly worked in a specialised consulting role where the charge to company was $2000 per day and my cost to the company was $55/day.
I used to do margin calculations at a mid-size US agency that staffed a lot of roles. In terms of actual project margin, we rarely got past 30% and was usually more like 15-20. The trick is that there is a lot of cost in operating an agency beyond just salary. The cost to the agency of employing a developer (or a designer or a tester or whatever) is more than just what they're paid.
We had a rate card that included both the cost to the agency per hour for a particular role and the agreed rate we had settled with the client. We had a fixed rate for each role/locale even though we paid people different amounts. I knew my own salary and handful for my direct reports and the "cost" side of the rate card was a lot more than I was getting paid. And that cost was the base on which padded to get profit from the client.
And that profit itself wasn't pure profit for the company because a lot of paid for our fixed costs (office space, hardware) and non-billable roles (HR, receptionists). And of course a lot of it ends up in the hands of our holding company and their stock holders.
It would be interesting to know whether the same issues would exist outsourcing to a country such as New Zealand with less of a language and cultural barrier or if it is still largely the nature of remote agency work.
Page layout is not great, got to the end and saw founders pictures so I thought that was the whole article. I'm interested so I reopened to double check and there's more on the left.
Yeah i was feeling so out of the loop wondering wth everyone is taking about.. where is the carousel? Checked the page 2 times before I saw that button under everything else. It looks like the sort of thing companies do when they don't want you to click on something! It's such a bad theme.
edit: after checking out the entire article now, it is quite insightful to be honest. Lots of interesting research and data. I think they should just make a simple html version like the craigslist website and it will be a super hit.
Yeah, I would have missed it too, if you hadn't mentioned it. The sidebar is not visible unless the browser is fullscreen. That's fine, but then the button on the bottom that continues to "II. Mobile dev agencies >>" should be above the founder pictures. That way, it'd be clear to those who the sidebar is not presented to that the content continues. If that's not appropriate because the founder pictures are meant to be part of the content, then perhaps they should be on the last page, because it does kind of signal the end.
I’m confused ... this says Andorra has >15k people working at dev agencies. The country has a total population of <100k, most of whom work in the tourism industry.
Yeah, I'm thinking the simpler answer here is that this data is basically trash. Based on some of the other comments in this thread, it really seems like it is not a good quality survey with pretty poor quality control.
It reports fewer agency developers in China than in Andorra. The difference between the two represents 1% of Andorra's total population and 0.7E-4% of China's.
As another commenter pointed out, their numbers don’t line up well with other broad surveys like State of JS or Stack Overflow, so I would say it reflects a low-quality survey.
Because these are agencies. They're often called in to maintain legacy systems, which don't get to use shiny new tech. Hence low or missing numbers for things like Ruby or React, despite their prevalence otherwise.
Sadly, that would actually point the other direction. Ruby was red-hot circa 2010, and while it’s still widely used, it’s not the new shiny thing anymore. I would expect to see more “legacy maintenance” Ruby work than brand-new ruby work these days.
I only have anecdata but all that I have of it points to agencies being heavy Ruby/Rails users and React as well.
Both of them work well since they impose a structure and programming model that basically works for a lot of business software, so you can focus on the domain problems and move quickly over the standard tasks like an asset pipeline or account management.
Ruby is definitely not new and shiny. There are plenty of long-lived Ruby on Rails projects that need to be maintained.
Makes sense not seeing React, though. It is far more complicated and fragile to get SEO-friendly pages rendered with React (and there aren't many legacy systems that use it, either).
It'd be really interesting to see more than just averages for rates. Many dev shops I've interacted with are all right around $250/hr here in the US / on coasts, I wonder if there are similar standards elsewhere, or different tiers in different areas.
Hourly rates seem very off for Italy as well, as the real average is much much lower to the declared $80/hr (web&software). It is especially suspicious that France ($64/hr) and Germany ($55/hr) have a much lower hourly rate, when in reality salaries and rates in these countries are at least 20-30% higher than Italy.
Agencies like Reply, Accenture, Avanade and Engineering are quite overcharging other companies for their consultants in Italy, taking advantage of the fact that a lot of big companies are hostile to keep devs in house (almost all of the big banks for example). I had the chance to see for how much I was being sold as a consultant some years ago and it was almost 3x the salary I was going to get (3 x 32k before taxes).
Lot's of commenters in this thread are trying to draw conclusions about dev salaries from this data, but you really can't do that for exactly the reasons you described. It gets even more disproportional for some of the bigger, high price tag consulting firms. In the Big 4 it's common to have entry level devs making $80k/yr ($~40ish/hr) while being billed out to customers at $200-300/hr.
I work in one of the companies you mentioned in Italy and, although it's true what you said, the rate paid by the customer still is around 350-550 Euro/day, which is $46-$73/hr, and these are the rates at top tier consultancy companies. $80/hr average still seems very high to me.
These rates are quite interesting. After I did one freelance project (managed team of 3devs, created prototype of 5G API network component), I am thinking about starting my own small software agency. I was expecting to charge 30-35EUR/hour (Central Europe region), but according to these rates, my expectations are quite low.
Can anyone give me any advice, books or podcasts, how to get better in business negotiation, to be more confident to ask a proper charge?
I did a stint as a "contractor" at a body-shop consultancy back in the 90's, only a few years out of college. Although they did pay well compared to what I could get as a perm employee, they were charging THREE TIMES what they were paying me. The friction that this caused was a deal breaker: the project manager knew that I was costing as much as quadruple what their ordinary employees were costing them, and expected 4x the output.
I've subcontracted out to folks, and even at a 2x markup, with the coordination, review, and other unbillable activities I made almost no profit. As a percentage, it was barely worth even having the initial leverage and I rarely did it (at 2x markup) unless it was strictly to meet a deadline. In large orgs where profit is a requirement for just about every interaction I can see how 3x or 4x would be necessary to justify doing the work in the first place.
IME this is very common and in some situations can be beneficial for everyone involved. The company hiring the consultants get the flexibility of being able to quickly hire you/fire you if needed, rather than going through a lengthy recruiting/hiring/firing process. They pay the premium for it, but it can be worth it.
Then the consulting agency obviously makes a nice profit off of it.
And then the consultant themselves can usually benefit from it as well, because typically they get a higher wage than the average employee to begin with, and also it becomes a very easy conversation to have with the company "just hire me as a FTE instead of paying the consulting company. My salary is currently $50/hr, you're paying them $150/hr for my work. Instead you can just pay me $100/hr, and we both benefit." I've seen a lot of consultant -> FTE transitions happen like that.
That only does work if the company is actually looking for a long-term FTE, though. Like I mentioned above, a lot of the time the company is paying the premium in the first place is so that they can quickly and easily drop you off their payroll if needed. They lose out on that by bringing you on as FTE.
Early job in the 90s - first 'web' job (1998) - was earning around $40k - $20/hr (more or less). I was being billed out at $125-175/hr. Factor in a loaded cost for me with some 'benefits' (not much) and it was ... $30/hr? Being billed out at 4-6x your wage eats away after a while. I was given some motivational stuff about teams and overhead and such. I asked for an office with a door and $60k salary, and was rejected, so I left. Had they even just given me the office with a door, I'd probably have stayed, and life would have been a lot different. They had a new office building built to spec - with a cube farm, then a wall of offices with doors. Those were empty. "For future managers". I wasn't allowed in them - they sat there empty. :/
Sure, but that applies to all the numbers. The average agency developer in a country that shows up with $30 in this survey isn't making that either (India is at $36, and I've heard talk of $15 or lower wages there).
The percentage that is overhead will vary, but only by so much. Otherwise, someone will come along and find a way to undercut your prices and take your business. It's reasonable to assume that if one hourly cost is double another, the developer's pay is higher.
(Not denying someone might find a way to really pull a scam in the short-term, and find a kid they can pay minimum wage to, but that's not a scalable or stable arrangement).
>The percentage that is overhead will vary, but only by so much. Otherwise, someone will come along and find a way to undercut your prices and take your business. It's reasonable to assume that if one hourly cost is double another, the developer's pay is higher.
That's really only true if you assume that all of these dev agencies are equal. In reality, a lot of them have different specialties (both tech wise or industry wise), better (or worse) reputations, or certain partnerships. All of these things can greatly change the amount that the dev agency can charge.
I've personally seen plenty of agencies charging $100/hr while another agency with a different "specialization" (but otherwise similarly skilled and paid devs) charge $250/hr. The bill rate hardly ever corresponds with salary, IME.
an early consulting contract in the USA for me was, to retrieve the code and data from a State University department.. the contract was for >$10k per month, yet the graduate students who built it were paid miniumum wage.. records were in the collection. The school would not release the code+data for years, apparently, with infinite excuses and stalling, never a "no"
The whole site offers a poor UX with weird redundancies, terrible navigation, grammatical errors, etc. It's probably sensible to conclude the methods used to gather, clean, and present the data are correspondingly untrustworthy.
I wish someone would do this for legal work. I always felt there could be similar plot charts to help reveal the outcome of wildly different needs, and still give a feel for what to expect
Something really, really annoying about this site: it uses automatically-progressing carousels that cause content reflow below them in the page.
I encounter this sort of thing so often, and it’s really, really bad. Positively awful. Super disconcerting.
In most cases like this, it’s just that one slide is longer than the others. In this case, there are a couple like that, but also for the rest, image tags have been used without specifying their dimensions, so they take up no space as the carousel slide enters, and then once the image starts loading they reflow instantaneously as the dimenions are known.
Please, if you have a carousel, you must make sure that its height never changes as it progresses. (And please always put width and height on your images, too, so the intrinsic aspect ratio can be applied properly.)
Oh, and auto-progressing carousels as a whole are a blight, but a sometimes-just-barely-tolerable one. So long as they’ll just stop progressing once I interact with them! Please!
Does anyone like carousels? I encounter software developers railing against them, I encounter normal people railing against them (from recently, a non-techie had one request about a new site, that the carousel inserted by the developer of the prototype—because it was a part of the WordPress theme used—be removed), I honestly can’t think of having encountered a single person that likes carousels. They’re just all-round irritating.
Especially for a site filled with charts. I'm trying to compare some data, and all the sudden the data has moved to a new place. I stopped reading after 2 pages, even though I was very interested in the results. Easily one of the most frustrating sites I've used recently.
haha oh man im going to click a link.. is the page done loading..? think so.. here goes... content loads in and moves the target.. clicks wrong thing goddammit!!!
[1] https://devquarterly.com/report/q2_2020/web#developers_worki...
[2] https://2019.stateofjs.com/front-end-frameworks/