Big companies can't risk going with a freelancer (who might go out of business or be booked when they need updates) so they put out an RFP to software firms.
A month or so of upfront design (wireframes, screen design, user stories). Another couple of weeks of iterations to get signed off by the branding/marketing group.
The development will need a team, four developers is about right (we need to get this app out fast in the ever-changing mobile landscape). We'll need a manager at half time and a team lead to handle client meetings. Throw in a dedicated test engineer (so many phones to test on these days...) and a month for formal test plans and execution (did we mention this app might be audited? test plan is 500 pages).
Make sure it is localized as well - we are a global company with global customers, all of whom we value. We also don't want to be sued, so we'll need a EULA screen that the user must accept. And a help page. And a way to register on our site.
Oh yeah, the way our budgets work, we've only got one shot at getting money for this app - can't do a minimal release and update it after it's in the store.
Conversative costs for an app in the OP's "BMW Tier":
One month upfront design at $100/hr (1 designer, 1 dev): ~$30k
Three months of development at $100/hr (4 dev, 1 test): ~$250k
Three months for PM at $150/hr (0.5 PM): ~$40k
One month formal resting at $100/hr (1 test): ~$15k
Monthly maintenance/update 20hr/month (1 dev): ~$25k
Total: Over $350,000 for the version 1.0 of your mobile app (on a single platform)
You sound like you're disagreeing with the OP, but the article explicitly calls out a $400K+ category of app development if you "do everything right". I think you're on the same page as the author here.
The Bugatti Veyron SS costs approximately $2.7M. That's about the top of the price range for new-production road cars. Owning and operating a world-class race car costs more and tends to be budgeted as a lifecycle rather than a one-time purchase.
I'm sure someone else can come up with witty analogies for apps, but I think the model holds. Outsourcing the development of an app costs as much as a car, for some value of car. You get what you pay for, and the things that cost a lot at the high end aren't immediately obvious to people not in the business.
I second this estimate, though we do it at half this cost as our developers are in India. Typical enterprise applications require 3+ years of maintenance & enhancements that would work out much more than just getting the app out of the door. Also most enterprise applications would require a travel component as requirements gathering and field testing has to be done on premise. Add to this cost of equipment that is required for development and testing. And yes, freelancers cannot be trusted with enterprise development as long term support infrastructure is required.
> freelancers cannot be trusted with enterprise development as long term support infrastructure is required.
How is a company that has employees any better? What are you going to do, lock the original devs in the dungeon so they can't leave?
The reality is that getting the same set of eyeballs that did the first job counts as the kind of special expertise that only comes with a premium. Either the devs are being paid to watch Law and Order seasons for three months in-between your maintenance cycle, which is its own special kind of premium rate. Or, you have to outbid the project they've moved onto, plus any contractual penalties they might incur for moving off the new project, plus the Fred Brooks Tax for playing musical chairs with developers.
A company is a going concern which can provide legally tenable support guarantees irrespective of individual developers which is not possible for a freelancer. Project Management and enterprise processes are precisely there for this purpose. With a freelancer, you have no such guarantee. With enterprise outsourcing, cost is inflated, but business continuity is taken care. Reason why Bank of America would not look out for the next door freelancer while developing their mobile app. There is a reason for organizational forms to exist :)
> which can provide legally tenable support guarantees irrespective of individual developers
There is very little in software that is irrespective of individual developers, least of all support guarantees. It in fact hinges entirely on the current and former individual developers, and the quality of whatever notes or documentation were left behind. The mere fact that two parties negotiate that software should be performant and bug-free, does not make it so by force of will.
As for business continuity, most "app" companies have been around for 1-3 years, whereas the developers that comprise them usually have a decade or two at writing software professionally on the resume. There are more established "consulting" companies that pre-date iOS, but I'll let you in on a secret--they farm work out to freelancers. It's entirely a question of how many middlemen you want between you and the person ultimately responsible for getting the work done.
I don't know where you got the notion that software is just creative art form of its individual developers. In that case, software would have been a guild activity and the industry will not exist. Software has certain creative elements, but there is also ample engineering and planning involved. Don't fall prey to hubris of lone-wolves of yesteryear - the software universe is manned by millions of faceless workers who create, design, engineer, plan and produce complex software. It is not one or two individual developers who matter, it is the team. This is coming from a Software Architect who has designed and implemented dozens of complex projects with large teams. Nothing would be farther from the truth if I said, the success of my projects were solely because of me or two three lone wolf programmers! Software as an industry depends on the notion that the process of developing software is controllable and replicable. If you do not get that notion, then you can as well dream industrial revolution did not happen and software universe do not exist.
That's a lot of bloat for a single-platform, first version app. A designer + developer duo can put a nice polished app out in a month or two, and could cost way under $100/hr. You certainly don't need a team of four + a PM. The process you're thinking of fits the "I want the best" price range (1 month testing?).
Way under seems like a bit of a stretch. The average developer salary is $92,000 per year, which works out to around $45/hr if we assume a standard work week. The rule of thumb is that you need double or even triple of what you are paying your employees in order to be able to afford to employ them. Even if you are the business owner and employee, you still need to recoup the costs of running the business for yourself so the math remains the same. It seems the bare minimum you could charge for the average developer is $90/hr, which is just shy of the $100/hr. quoted. Above average developers will require much more per hour.
You could, perhaps, charge way less if you hired below average developers, but that seems to fall under the first category and not the one the parent is referencing. This, of course, assumes that you are operating in the USA. Numbers may differ in less costly parts of the world.
I feel it still sets a decent baseline for the purposes of discussion. We will never be able to accurately model every last developer's situation.
Though entry-level developers are even more costly to employ. They require more training, more time to solve problems that are old-hat to experienced developers, etc. You still need to bill them out at, say, $100/hr. with the expectation that for every productive hour, they have an unproductive hour that isn't billable.
Hi, I am from another country. When you say that the average developer salaray is $92000 does that include all costs associated with that employment? Including insurance and possible taxes?
I am just trying to compare.
EDIT: I don't mean _all_ costs (not like desk and computer) but salary + benefits + insurance + pension + taxes.
$92,000 is what the employee gets paid by the company. It would include any costs incurred by the employee (income tax, transportation costs, etc.), but would not include things like benefits, insurance, etc. that are paid for by the employer. That is, in part, why you need, at least, twice the rate of your employee in order to be able to afford them.
Oops, I missed the part about not being able to use freelancers. $50-$60/hr is a common rate for freelance developers. There also excellent devs that don't live in countries as expensive as the US.
This is a recipe for disaster. I have been full-time freelance for the last 4 years, and scaling into a full-service consultancy with all the accoutrements (like an office for my employees), and I can tell you that $100/hr is really a bare minimum for meaningful profit margins on developers. I have a few contracts under that (for various reasons), and with the new hires I'm contemplating, I am going to have to raise their rates or drop off and replace them with better paying clients.
As a freelancer, you might be able to get away with $70-$80/hr minimum, but $100 minimum still is easily within a reasonable price range.
I would expect a fair market rate for great developers to be something more like $150-$200/hr. Unfortunately, I have a hard enough time getting clients around here at $100, but in order to really have the business firing on all cylinders, I would need to be charging $150+ for each developer hour.
The margins at $100 and below are barely enough to make payroll; there is almost nothing left to pay for things like significant marketing or PR, investment in equipment or facilities, loose time to spend on internal projects or internal maintenance, etc.
I think the point was, that even if you are a freelance - you still have employ yourself, with all the overhead that implies: running costs, pension, vacation, accounting etc.
So basically most above-average developers in US (and areas with similar wages / living expenses) would charge at least $100 per hours.
Your point about developers in countries with lower living expenses than the US is valid. But that would also depend on the customer being in said country.
Where the developer lives has nothing to do with what they cost (at least it shouldn't). Cost is a function of value. So if you get a really cheap developer then either you've found someone who has no idea what their market rate is (what else are they ignorant about? Anything that will hurt your product?) or they're just not as good as someone more expensive (the majority of the cases, though people don't want to believe this. They want to think they've found "a deal").
So let's assume these guys are on staff working at market rates for experienced pros in a major metro area. Say salaries of $90k per year (realistically they cost much more and we'll get to that)
They each then make $7500 a month...multiply by 2 (2 guys working for a month) = $15k. 2 Months = $30k.
Either way we're still in Hyundai Accent to Honda Accord range there, and the hourly rate is a bit over $40/hr (or $80/hr for both guys).
So you are right that it can be done for under $100 per hour. But it still costs about a car.
More realistically,
If you estimate it out the right way, benefits and all included, you start to climb up the hourly rate scale very quickly. And the cars you can buy with that money get a hell of a lot nicer.
A good rule of thumb is that you assume skilled experienced labor wants to bring home about $100k. A good conservative estimate of what it costs to employee somebody at $100k is about $200k (benefits, overhead admin expenses, etc.). And if you follow that, we end up with about $100/hr per person pretty quickly.
You are assuming that your single developer is capable of everything, and has time for everything - writing mobile code, writing backend code, DB, DB optimization, profiling of all the code, setting up analytics, documentation, etc etc etc.
If you can find a guy like that, and if he can write maintainable code, all "way under" $100/hr minus designer's salary, you've struck gold. People like that are extremely rare.
I really doubt a v1 of an app would have much if any "optimization", profiling, analytics, or documentation.
I would argue that the majority of the easy apps are done. Anything more complicated than a fart app will require profiling and optimization. It's easy to forget that iOS and Android are embedded systems with a finite amount of resources.
Well, you get what you pay for. That's the difference between some random dude and a pro. I deliver my v1.0 products optimized and documented. Analytics is pretty much a requirement these days, executives absolutely need to know how their app performs.
The article didn't appear to be speaking to a "big company". Those people already know the cost makeup of getting their $300k+ app done.
I think this article was speaking more to the smaller business that has no experience in app development but wants to take a plunge into the market. I deal with these smaller businesses here and there. They're often very unrealistic when contemplating the costs associated with app development. They're thinking $500, like that ad they took out in the local paper and they're often shocked when I let them know that their product vision is a couple of orders of magnitude different than their budget.
This article would probably be one I'd point them to in order to help them understand the investment that they're making, since it is an investment.
Big companies can't risk going with a freelancer (who might go out of business or be booked when they need updates) so they put out an RFP to software firms.
A month or so of upfront design (wireframes, screen design, user stories). Another couple of weeks of iterations to get signed off by the branding/marketing group.
The development will need a team, four developers is about right (we need to get this app out fast in the ever-changing mobile landscape). We'll need a manager at half time and a team lead to handle client meetings. Throw in a dedicated test engineer (so many phones to test on these days...) and a month for formal test plans and execution (did we mention this app might be audited? test plan is 500 pages).
Make sure it is localized as well - we are a global company with global customers, all of whom we value. We also don't want to be sued, so we'll need a EULA screen that the user must accept. And a help page. And a way to register on our site.
Oh yeah, the way our budgets work, we've only got one shot at getting money for this app - can't do a minimal release and update it after it's in the store.
Conversative costs for an app in the OP's "BMW Tier":
Apps ain't cheap.