I worked on a startup that received seed funding to tackle this. I helped build the platform (using Twilio video) and then answered the first $5000 of calls. Things I learned:
- Many entry level programmers at big companies (think Walmart Labs, Target, Goldman Sachs, etc) are too scared to ask managers/team for help. If you work for a large company, make sure new hires feel safe asking questions and seeking help or they will seek it elsewhere.
- The really hard questions they're willing to pay $240 an hour for have too much context for you to grasp. They want you to replace their ORM layer or tell them why their 600 line test case is not working.
- Getting customers is not too hard by poaching them from Stack Overflow, Reddit, CodeMentor.
- Developers make good money working full time and idly sitting at their desk part of that time. Here you only get paid for active time and the idle time in-between will kill your salary. Your motivation has to come from helping other people and watching them grow, not from $$$.
Yeah, the idling thing, or looking from the other side, the on-demand expectation pretty much kills it. Or makes it hard to organize.
I used to be active on CodeMentor a few years ago, and the bad thing for these problem solving kind of requests is that they are usually prompt requests. Now working remotely, in theory, you can be available almost any time, the problem is, that after a while you just feel like not wanting to leave what you do. At least not on a short notice.
Agreed upon mentoring sessions that can be scheduled work a lot better from this regard.
Actually, when I joined codementor, I was working on a startup that offered psychology consulting in a similar way. (And one of the reasons me joinging codementor was to get first-hand experience of the dynamics of such a service.) Our psychologists didn't like either that they had to be around and that they had a lot of non-meaningful inquiries. (Similar to what you see on codementor.)
There’s also the awkward feeling of not being able to solve the problem and still charging something. I actually preferred “tipping”. I solve your problem and then you pay me instead of running a timer.
> ...awkward feeling of not being able to solve the problem and still charging something...
I think that can be addressed by setting the right expectations. A discussion with an expert doesn't have to end with "the answer" to be super helpful.
A lot of times people have just gone down the wrong path and need someone to get them going in the right direction or even just confirm that they're doing the right thing.
This requires a empathy from the provider and grit from the person seeking help. This kind of work is definitely not a "stackoverflow" kind of thing. It's also can't be about the money, because anyone with enough skill to help people out with dev problems is likely already well paid and doesn't really "need" to do it.
Well, on codementor you can give a refund. And that would erase the session (and the rating, if you received one) from your history. I think it's better than tipping. Tipping, at least to my European ears, sounds like this voluntary thing as opposed to paying for a service.
BTW, what happened to your service? Or, probably better to ask, how far did you get and why did it fail?
> Getting customers is not too hard by poaching them from Stack Overflow, Reddit, CodeMentor.
I have contributed to the SE for more than 6 years, beginning as a greenhorn and ending so far with more than 400 posts(most are answers). I find the value of SE is threefold: 1, it provides me a communication platform; 2, it provides me a place to take notes; 3, it keeps track of my improvements partly by a reputation system which I can monetize elsewhere. I only spend time writing answers there in my idle time.
Yeah I have up mentoring because it doesn’t pay super well and the way things are the mentor often knows more about the problem then me because for example they are using a specific thi g like redux thunk so then I have to learn about that in spare time for free to then go help them for an hour for $60 and also just having to work on a specific time basis makes it less convenient than coding. So all in all it’s not great for the mentor.
Have you experimented with a more constrained niche like Hackerranking? My take is that people would pay for automating the useless interview rounds or at least to have some handholding. Easier on the scheduling as well. Win-win-win for all sides considered.
edit: conversely I would be interested to know if such an option already exists
I believe there is HackerRank training. I think it's offered by some of the "interview training" companies, but I don't have links offhand. I know I've seen a startup on HackerNews that was supposed to connect you with an ex-FAANG engineer to help teach you how to "beat" their interviews.
You'll need some mechanism to raise rates to attract top developers.
Even without the first free minute and time spent reading questions, the most a dev could hope to earn with this service is ($0.75/min)(60 min/hr) = $45/hr. Perhaps for students or devs who barely meet the 100 rep minimum requirement, that might be attractive, but for the experts who've written the best Stack Overflow answers, that's likely to be a (small) fraction of their bill rate.
Free contribution is rewarding as a means of helping.
Well-paid contribution is rewarding for the $$$.
Poorly paid contribution often loses both incentives.
I would also argue that possibly doing it per time may be the wrong incentives.
I have had issues that took me several hours to try to debug, only to have someone else show me how to fix it in under 2 minutes. So that dev gets paid $1, but I would easily pay much more for the fix in that issue because of how much time I sunk in it.
The Graybeard engineer retired and a few weeks later the Big Machine broke down, which was essential to the company’s revenue. The Manager couldn’t get the machine to work again so the company called in Graybeard as an independent consultant.
Graybeard agrees. He walks into the factory, takes a look at the Big Machine, grabs a sledge hammer, and whacks the machine once whereupon the machine starts right up. Graybeard leaves and the company is making money again.
The next day Manager receives a bill from Graybeard for $5,000. Manager is furious at the price and refuses to pay. Graybeard assures him that it’s a fair price. Manager retorts that if it’s a fair price Graybeard won’t mind itemizing the bill. Graybeard agrees that this is a fair request and complies.
The new, itemized bill reads….
Hammer: $5
Knowing where to hit the machine with hammer: $4995
I had a plumbing problem in my house. I had a sink that was randomly filling up with clear cool water. When I used the sink it worked fine. Now I'm pretty handy, but I had no idea what was causing that to happen.
So I called a plumber. I showed him the problem, and he immediately knew the cause. The condensation drain line for my AC ran to that sink. He suggested snaking the drain. He was there 10 minutes but that fixed the problem. He charged me $175 which was his minimum. I gladly paid it. I didn't view it as paying $175 for the fix, I viewed it as paying $175 for his expertise.
you know that some developers intentionally add "time bomb" bugs and introduce bugs that trigger after certain period of time - just so that they knock it off with the hammer in one minute an justify their "maintenance support" contract.
there was a story about one european company and an Excel spreadsheet with VBA code that would stop working after 3 months and a developer who would "unlock" it for another 3 mo if he has a maintenance contract
> On Monday, David A. Tinley, a 62-year-old from Harrison City, Pennsylvania, was sentenced to six months in prison and a fine of $7,500 in the scheme.
Initially, the offending code was password-protected. However:
> While this worked for about two years until May 13, 2016, Tinley's scheme was discovered when he was out of town and he had to give his password to Siemens' employees because of a time-sensitive deadline that required the spreadsheets to work.
I experienced something similar early in my career. I was stunned when the old guy who'd been brought in for one day of consulting explained that he'd renamed his parameters with single characters to guarantee future work.
but the GM had already given us a stern warning to show him respect. also he got around without the use of his legs. If I'd reacted as I wanted to it would not have gone down well...
I was recently having a conversation where I speculated that with how bad some programmers are at naming things, some code reviews might actually be easier if you ran them through a code obfuscator first, so you can focus on what the code does rather than being tricked into making invalid assumptions by misleading names.
This gives me a great inspiration for a programming puzzle and learning method:
Replace all variable/function names with generics. Your only job is to give names to every variable (trying to figure out what each does). After you're done you can compare to de-obfuscated names to see if they do what you expected (if they're named sensibly).
Back in the days of Visual Source Safe you could roll your system clock forward and check something and it wouldn’t go into effect for everyone else until their clocks caught up. For example, 6 months after you leave the company.
I agree -- the ability to put a flat rate bounty (similar to what is done with non-monetary bounties on Stack Overflow) would be a better way of fair price discovery. Maybe even allow the devs to place bids on problems and auction the price down from a maximum price bounty.
> Free contribution is rewarding as a means of helping.
> Well-paid contribution is rewarding for the $$$.
This is crucial. Changing incentives changes the social contract and relationship and does that irrevocably.
I recommend Dan Ariely’s book Amazing Decisions on this topic.
Example: You don’t pay for the thanksgiving turkey dinner at your inlaws’ and if you did that’s be rude no matter the amount.
If you give your neighbor some fruit from the tree in your yard that’s one kind of relationship. If next year you try to sell it that’s a very different one. Transitioning from one to the other is significant.
Lastly, financial incentives kill generosity. The giver now starts to worry if they’re providing too high of a value.
45$/h will attract a LOT of attention online. And not from the folks who should be answering questions. Just look at what happened with Hacktoberfest [0].
Thanks for sharing, I apparently missed the story the first time. Unfortunately, Twitter suspended the @shitoberfest account (https://twitter.com/shitoberfest/) Would have loved to have a look.
TLDR: Imagine every single major repo being inundated with tens or hundreds of thousands of absolute rubbish PR's that do nothing but change one line and break the build.
Like fork React or Linux and then open a random file in the web editor and type "GIVE ME SHIRT" in a random position.
Then save it and commit and submit a PR.
Followed by opening a dozen issues demanding that it be accepted so it counts for the free shirt.
I contributed to a small (<5k star) repo at the time and after telling the issue/PR-creators to stop, started getting newsletter signup spam and hate mail directed to my git commit email too.
I don't even know who I pissed off, but a couple months ago wound up on every dating site one could think of's mailing/user list... I mean whoever did it obviously spent a fair amount of time just finding all the damned places...
Exactly. My usual hourly rate is a minimum of $70/hour. (I only recently became a freelancer, and am still learning to maximize my rates– I suspect my current rate is a small fraction of what I could earn.) The only thing this provides for me is a more liquid/elastic way to earn money.
I can and do regularly solve problems that save people hundreds of hours. I would gladly pay to have someone solve my technical challenges that cost me hundreds of hours.
(I am an expert in Python and a deep Generalist with experience across many domains and types of software.)
Isn't the thing to do to charge $200ph and offer "significant discounts" for interesting work. Now people have a reason to pitch you their projects and you can still charge $70ph for the interesting ones.
If you're an expert in Python and a deep generalist I suspect 70 is too low, depending on where you are at.
I follow the easy to understand rule that my price for freelancing must be at least double what I make as a full time employee, and thus a full time employed position must be at worst half of what I make as a freelancer.
This also allows me to see approximately where my best position in the market is. For example right now I am getting offered full time wages above half of my freelance rate (which is a little over $100 an hour) which implies to me that for my profile it might be more worthwhile to take a couple years full time employed.
Someone once told me “Take your salary and double it, and that’s your minimum contracting rate”. It seemed impossible to me for years, but ended up working just fine when I finally did it. In fact I more than doubled my salary, and no one really bats an eye. If you have skills people need, they’ll pay.
For what it’s worth I was earning a salary around $110k CAD, and my contract rate (assuming I worked full time) worked out to around $240k CAD. No one has had a problem with it yet (although I went back to a salary position since then - I have kids and contracting was hard to organize around family and the pandemic).
I recently started contracting for a consulting shop I had a relationship with, took some back and forth to get them to agree to $100/hr and I felt pretty good about it. Then I got placed a project for a FAANG client where I'm the only contributor and saw what they're ultimately paying for my time...
> The hard part is quoting the rate with a straight face
The best anecdote I heard about that was the guy (wish I could remember his name, but he was interviewed by Matt on Freelance Transformation) who found that it was much easier to quote a high rate if he had a cigar in his mouth because it made him feel like a mogul :-)
I've had the luxury of not needing to ask for money because I need it. When you aren't in need, it seems like asking is a lot easier. I'm not worried about a No, so I'm not fretting about the answer or my delivery.
This has made a huge difference for me because I'm a terrible negotiator. Earlier in my career when more hinged on negotiation and I needed money, I sometimes worked for small fractions of what I could have.
I liked what another commenter said about adding tiers for devs with higher S/O rep. So if the dev has a higher rep, they can make more $ per minute. Something I'll look into adding.
I don't think anyone is going to make much money off of it starting out since I don't have a large supply of questions coming in yet, but there are a lot of underemployed devs out there who would be happy to make an easy $15 bucks an hour just googling answers for people and telling them what to do.
I think a lot of poorly-received questions on S/O are the result of people not knowing where to look for their answer, or being intimidated by reading the docs or source code. So they post the questions hoping someone can guide them.
Call a Dev isn't a competitor to S/O. S/O is a wiki. Call a Dev is basically Clippy in human form that you pay per minute.
> there are a lot of underemployed devs out there who would be happy to make an easy $15 bucks an hour just googling answers for people and telling them what to do.
Maybe there's a reason they are unemployed
> I think a lot of poorly-received questions on S/O are the result of people not knowing where to look for their answer, or being intimidated by reading the docs or source code.
Isn't is the point of engineering? To know how to get information and what to search for? If the 15$/hour dev knows and the guy on my team doesn't, why am-I not employing the person answering questions? My company sure pays more than 15$/hour...
Unfortunately I suspect this may go the way of many freelance sites, and get absolutely flooded with low quality indian developers.
Where the wage is lower $0.75 may seem a lot more attractive and there are a lot of fresh out of university indian devs.
If you've ever tried bidding on remote dev jobs online, you'll practically always get undercut by an indian dev at a price you can't even come close to matching, and unfortunately I suspect that given how often they appear in comments online asking people to do their entire project, I suspect they may often be bidding for jobs they don't know how to do.
This is true, but it illustrates a key dynamic. For most people, as soon as you put a dollar value on a task that someone may have done for free or for fun, you have now fundamentally changed that relationship.
If someone is doing something for only the intrinsic motivation, putting a dollar value on it changes it to an economic calculation. The person may still do it for money, but they're more likely to only do a level of effort commensurate to the economic reward. Whereas if they're doing things for intrinsic reasons they may be willing to do more work.
There are a couple of behavioral economics experiments that bear this out. The Soma Experiment from the 70's gave participants a puzzle game to solve, and measured how long they tried to solve it. One group of participants was paid for their time, the other wasn't. The paid group on average spent less time trying to solve the puzzle than the group that wasn't paid. There are a couple of other experiments in that vein, but the common thread is that intrinsic motivation can be more powerful than economic rewards in multiple contexts.
So someone might be willing to contribute some code to OSS for free, but if you ask them to develop some code for $30/hr they might pass.
Stackoverflow is both asynchronous and asymmetrical. The former lets the dev answer when convenient. The latter means the dev knows that their time is being multiplied and potentially lending help to large numbers of other devs over a large period of time. That is a compelling idea. A phone call is one-to-one and inconvenient. $45/hour is also hugely underpaid in the US at least. So, the proposed system is underpaid in both $$ and personal satisfaction.
I wonder if OP's product would work better if it was more like StackExchange bounties. Post a question alongside a non-refundable dollar amount. When someone posts an answer that solves your problem, you can award them the bounty. If no answers are accepted after a reasonable amount of time, the bounty is forfeited.
This setup contains risks for both the asker and the developer, but I think they're balanced pretty well, and the incentives are set up correctly.
Yep, and I happily spend a fair amount of my free time working on open source, as well as personal projects that I don't make any money from. I'm also currently helping a friend of mine learn to code just for fun and to give back. But when I do contract work, I expect to be paid well. If you said, "well, I'll pay you $10 / hour, and that's way better than what you get for this open source you work on!", I'd not only not do the job, I'd be insulted.
Seems like a silly mechanism. I think a better tactic would be to offer "bounties" on problems, then devs can decide whether or not the problem is worth the effort. As long as the requester defines the input/output for the change, then this should work out well.
There are times when I'd spend 15 minutes writing a small bash script for $30. Sometimes I feel guilty about buying luxuries like delivered food, and I wouldn't feel so guilty if I use half the saved time earning enough to pay for the meal, even if it is well under my typical billing rate.
It's ok for a normal job (guaranteed 40h/week) but for hour-based freelancing (let along minute based!) it's really not. $45 is €37 and you can absolutely get more than that as a freelance developer here in Germany.
People who are used to getting a salary severely over estimate how much money an hourly rate translate to per year when you don't have guaranteed work lined up.
As a rule of thumb $x per hour is equivalent to roughly $xK per year for a salaried position when you account for unpaid time between gigs, paying your own insurances, paying for your own equipment, not getting paid vacation and sick leave, etc.
What if there were really good devs on StackOverflow in developing economies that were willing to take on $45/hr consulting because their day job pays them $20/hr?
Just because Silicon Valley developers want $300/hr doesn't mean the whole world regards that as a standard.
$45/hr, billed by the minute for a bunch of short gigs spread out across the day with the need to do fresh business development between each one will never leave you anywhere near half utilized.
Put the bill rate at $200/hr in this situation and you might replace that steady $40k/year salary. But it’ll be a lot more stressful.
Agreed, plus the day to day answering developer questions is going to be a little more tame than if you were in a job constantly putting out fires or maintaining a tight and stressful schedule
Only for Americans. Everywhere else in the world that's a lot of money. Your average software engineer in Europe can hope to earn maybe €50k per year late into their career at most.
some how if you can leverage reuse of questions so someway for someone paying for an already answered question then it would scale up and keep paying somewhat like SO rep works. I'm currently in the top 0.58% of SO users with a reach of 4.3 million people, if you can monetize that to a small extent it could somehow provide a scalable pay scale.
$45/hour is a good rate for Ukrainian developers, I think. Or Philippines developers.
That’s about USD $90,000/year.
This is good money, if you live remotely in a lower cost of living, foreign country. But it’s a mediocre salary in the United States. You can barely pay your expensive rent with $90k/year. Actually, you’d be at poverty level. And there is zero possibility of affording a house with this salary. But you might be able to eke by if you live in the less competitive Midwest.
I was in KS and I was just paid "highly" eg. $12+/hr although I was multi-role
edit: maybe less was making $11-12/hr and not full time, my math was off. I was freelancing on the side though before I got into the industry but yeah $20K range was my figure at that time eg. just over 3 years ago(it did suck, I was mad broke still am but working on it).
Wow, sometimes this site really needs to exit the stratosphere and come back down to earth. I live in the Midwest and you can comfortably pay a mortgage on a sizable house, all utilities, all necessities, and still have a shitload left over to play with on a salary like that. By no means do you have to "eke by."
I think the issue is a lot less that the Midwest is less competitive and a lot more that the cost of living in a West Coast tech bubble is exorbitant.
I'm sorry, but this is an absurd statement. You can live very comfortably the vast majority of places in the US on $90,000. If you can't you're in one of the most expensive cities, pay a huge amount of debt servicing, or simply have unreasonable standards of what you think "necessities" are.
The only issue is billable hours - will
these be 100%? If the platform has that liquidity, then yes - if not, chasing assignments and lowering prices leads to a race to the bottom.
I'd guess that the people at Stack Overflow might have an issue with someone using their name in this way. It should probably be clearer that this is entirely unaffiliated.
100+ reputation on SO is also a ridiculously low bar, it's almost meaningless. Not that any amount of reputation by itself is a sure indicator, but 100 really doesn't mean anything at all.
This kind of site also requires a pretty large critical mass to work, you don't just want a bunch of random SO users willing to participate, but you need the right users with experience in the right tags to match to the requests in a reasonable timeframe.
There's a good reason Stack Overflow doesn't allow people to pay for answers, people already cheat enough for meaningless internet points, this will get much worse with real money on the line.
I'd rather have your help because you are eager to help and probably the patience to understand my problem, than having an expert who doesn't care and just want to get over with it.
First, the person who brought the game spends an hour reading the rule book. Finally, someone finds a 10-minute video on youtube where the creator of the game walks you through a few rounds of the game.
That was my experience with Mice and Mystics, anyway. Awesome game. Still can't get anyone else to play it with me.
The idea looks interesting, but a couple things of note that make me kinda wary:
- $0.75/min is $45/h and it's awfully cheap for what's ultimately consulting work
- 1 free minute in each call, call ends if no payment source is connected: what about people that would keep getting those one minutes? I know it's not much, but for some problem it might be enough if not to solve it, at least to get input that could lead to resolution.
I still like the idea because I feel strongly about stackoverflow pushing for solving everyone's issue instead of someone's, which often end up solving barely anyone problem as solutions provided are way too generic. The added effect is also that the community has devolved into a hot mess that will close so many questions as dupes even when they're not.
The free minute sounds like a really good idea to me - there might be twenty seconds or so of introduction which younger folks will tend to view as unnecessary time-wasting and older folks will tend to view as necessary pre-connection and the free minute allows these greetings to go uncharged and be mostly uncontroversial. I think it both counters the greeting time and also serves as a pretty good PR move. Also, if I can answer your problem in less than a minute then you didn't even google it.
I am sure a bunch of these sorts of questions will roll in and, if you're trying to continuously get answers within the free minute you're probably going to end up accidentally paying them a good portion of the time.
> Also, if I can answer your problem in less than a minute then you didn't even google it.
Don't undersell this ability, it takes years to develop, and honestly is part of the value a "senior" person brings to the job. If you don't believe me, just watch a beginner searching for the answer to their question on their own. They'll use overly-broad terms, click irrelevant links, and spend several minutes looking at non-useful answers.
As someone more experienced, you can look at StackOverflow answer and instantly see it _was_ the right answer in 2012, but is no longer compatible with some other thing and trying to implement it is going to result in a world of hurt. You'll also know what search terms to use, what to avoid, and little tidbits like "SQL Server" refers to Microsoft's, or that the answer involving a sysv script probably isn't relevant to your issue with Debian Bullseye.
Actually looking at the model setup by the site for contractors to grab questions on a voluntary basis - I sorta doubt anyone would actually take your question.
The site design is pretty opaque so I imagined you had some SMEs out there that a call in would be automatically connected to but the contractors can view a list of questions and pick and choose ones they think they're a good match for.
I think that makes it a valid question to the site designers if there should be either:
1. Some level of payment for the first minute of calls to make sure those dumb questions get answered
2. Some feedback mechanism for contractors to mark specific calls as googleable and for the site to send back a "Did you try googling it buddy?" response via email (though in much more diplomatic terms)
It sounds like this service might actually struggle with questions that are too easy - I'm also really wondering about the unpaid research time potential. Will it be a faux pas on this service to call up the client at the point where you think you know the answer reasonably well but may need to do some on-the-phone research for their specific details or would the expectation be an expert ready to answer your question specifically for the Sun Sparc 8 architecture?
$45/hr _does_ seem cheap, however, you don't have the upfront sales cost that comes with consulting. You don't need to maintain a relationship with a client. Hopefully cashflow and payment would also not be an issue with an intermediary guaranteeing some kind of payment (but it is a new startup so who knows...)
I used to offer tutoring for C, Go and Python programming for $50/hr.
I can’t tell you the number of people who asked me to do work on for their day job... I would always accept at $100/hr.
Almost every week I’d max out my 15 hours I’d set aside for “tutoring”. So beyond that I’d charge $200/hr, and I’d still get people for both tutoring and day job style work lol
$1/min is a fair price, but if you’re good you can make a lot more
This is fascinating to me. Is there any particular kind of profile that hired you as a ghostwriter? Do these people know how to do their work and decide to outsource it or they have no clue and just got lucky in the interviews?
I like this because it's a really off the wall take, but it also feels a little bit like the sort of reasoning that gets you tangled up with a Bernie Madoff type figure.
It's true that unscrupulous people are often very effective. But it's also true that they are unscrupulous. Leaning into lack of scruples probably isn't the most stable of pro business strats, but I imagine it works out sometimes.
I too would like to hear more about the clients. Also would be really interested to know how you viewed the level of difficultly you were dealing with when people were paying $200 out of their own pocket - did you view it as "tough stuff" ?
This interesting to me. I spent the last 4 years teaching programming full time, and also have a decently high SO reputation. I've recently gone back to full time software development, and miss teaching. The main problem is that at $0.75 / min, it's not worth my time. I generally have at least one contract side job making several times that much, and I'm not unusual in that respect.
Is there a similar site where you can put a dollar bounty on a solution for faster service? There have been SO questions where I would have paid for faster attention.
When I use Fiverr to hire a dev, I often have to contact multiple prospective devs to figure out that their skills are a match for the problem I'm trying to solve. I'll chat to 5 (all who advertise python) and give a one line intro about the problem, and it's immediately clear than only 2 can or wish to do it. E.g if I need someone who specifically knows the Selenium package.
Do you have a similar mechanism that can help to match the right dev with the right customers?
Also, $1/minute is more expensive than Fiverr. For my quick projects I would still use Fiverr due to the cost savings. I'd expect to fork out $60 for a two hour project there. On the other hand, others here are saying that it's too low for their hourly rate. Some price flexibility might be a good idea?
Devs ping any problems they feel capable of solving. They can ping with 2 messages, either they KNOW they can help, or they THINK they can help.
Once you receive a ping you can vet their S/O profile and if you like what you see you can start the call.
It's not so much for contract work or to complete tasks or projects like fiverr is. It's about getting personalized help with code your struggling with.
So instead of the customer finding the dev (as on Fiverr), here it's the other way around. Interesting. Lines up conceptually with the StackOverflow model of Q and A.
Ad hoc scraping tool for something experimental that I didn't want to do myself. Would've taken me 2x as long as the dev I hired since I haven't done that before. Quality wasn't so important, I just wanted the output of the scraper on a specific site for a one off thing.
$30/h is a typical rate for a dev on Fiverr, I've been very happy with the value I've gotten from it.
Because I know python and if I needed to make a small adjustment to the script (which I usually do, usually minor tweaks), then it's within my capabilities to do so. Takes me thirty seconds to adjust a small thing in a language I know well.
Also the pool of python devs is so large that shrinking the pool in this way doesn't impact on my ability to find someone suitable quickly
The bigger issue than shrinking the pool of devs is shrinking the pool of solutions and missing better/faster/cheaper options. E.g., many companies did exactly that in the previous two decades by trying to do everything in Java when a higher level language would have served them better.
If it's what you know and you're taking it over, though, that makes more sense.
> E.g., many companies did exactly that in the previous two decades by trying to do everything in Java when a higher level language would have served them better.
Can you give us some real-world examples that you've seen?
I agree but we're talking about a pretty quick and dirty personal project here where quality, maintainability etc aren't important, and I vaguely knew that the Selenium package could do what I wanted in this language.
If I was hiring for something more important I would've put more thought into that aspect
Possibly, but they're only asking for $60, and I figured it would take me a while to learn the package + parse all the data from the various tables and sections (it was rather heterogeneous data across the site) + if this site happened to use anti scraping measures that could mean it takes longer. They've probably done this same thing 30 times (it was their main offering on Fiverr) and can copy and paste a previous project.
So it made time/financial sense to hire them, especially since I don't need this skill for later
I feel like this is horrible both for the person getting help and the person doing the helping.
The person getting help is effectively trained to apply a quick fix mentality to everything because nobody is going to pay by the minute to learn fundamental software engineering concepts, no matter how much they should know them.
And for the person doing the helping, the pay is below average plus they need to review call requests for free in addition to it. They'll probably also have to deal with entitled stubborn people line in every customer service gig.
I predict the result will be that capable developers will avoid helping, because they can have a better working experience and more salary elsewhere. At the very least, I'd avoid wasting my time on such a platform. I would maybe consider it at $300 per hour, but the pay needs to be extremely high to compensate for the free reviews and idle time in between paid calls.
That again makes the service less useful for people needing help, because all the truly capable developers are off the platform. So you pay to get a quick fix from a mediocre developer... Kind of like on fiverr.
I'm a professional developer and I write arduino software for money "on the side." The effective rate varies a lot, but it's safe to say I won't be giving up my day job too soon.
I basically describe it as getting paid to tinker.
It benefits two parties: the person who needs help/needs code written and me. I get paid to do stuff I like and can do it sitting on the couch watching TV. This is how I envision a lot of skilled devs doing this: as a nice sideline that's fun to do. Not everything needs to be thought of as a replacement for your primary income source.
Interesting idea, but doesn't seem competitive for Americans to ever respond, if you only pay them $0.75/min after the first minute and the experts have to search for questions they want to answer, meaning there is guaranteed unpaid time between calls.
A potentially interesting option would be to offer tiers based on stack overflow reputation, and to ensure that a developer has reputation from the desired tags.
Site is down right now. I'm wondering if there's any mechanism to verify the developer you're paying for help. If I'm going to be paying, how do I know this person is actually knowledgable about my problem domain and good at troubleshooting?
On the helper's side, since it's by the minute, it's not much risk if the person is completely over their head and beyond any reasonable help, you're still getting paid while you try to figure that out. Except that the pay is somewhat low versus the going rates for consulting work in the western world.
This is cool from a Dev who needs help. It feels more like a micro-consultancy or more like a call center model. It is really bad from the Dev who takes the call. If I put myself in the position of the Dev who takes the call, It is pretty crap. Let say average call duration is 3 minutes, that means I have to take 20 calls/hr. Thats 20 context-switches per hour. I have worked at call centers, after first few calls in a day, people are just not motivated enough to help. There is going to be mental fatigue for the developer who is helping.
I'm sure there are people that will do this, but I view it as particularly grueling work. If I want to work an 8 hour day to pay the bills, I need to work 8 hours at top efficiency since I'm always being watched, and on top of it, be social in different calls all day.
Some of my best problem solving is while I'm out for a walk letting my mind wander. This is very much not for me.
Oh, on top of this, I'm sure the dispute process after spending say two hours trying to do something and failing for whatever reason is going to be _fun_.
2. Someone elects themselves as a candidate to answer that question.
3. I receive notification of the candidate and approve them
4. A call happens
Do you have any idea around what sort of time lag would be inherent in each step? Do you expect steps 1-4 to happen within a half-hour window and thus be mostly instantaneous as far as phone interactions go or could it go longer?
If you expect a quick response then I don't know if your rates are high enough to keep people actively watching the question board instead of checking in a few times a day - if not then I think the usability of the product may suffer since scheduling the call could become problematic. If I have a meeting in an hour should I bother trying to set up this call or do I need to wait until I have several free consecutive hours?
Ideally you'd post a question and get a ping within the first few minutes. But there is a 1 hour max window for any question to receive a ping. Once the ping comes through, if you decide to accept it, then the dev has 60 seconds to answer the call.
As it stands, there isn't much supply of questions or devs to answer them (just launched today with no early signups). I was thinking one way to get around this would be to let devs subscribe to email or text notifications when a questions is posted with a specific tag.
A diverse notification setup seems wise - otherwise yea, I think solving the balancing problem is going to be an ongoing struggle. The one hour limit sounds pretty wise!
The problem is, that no one can give you (or themselves) a reasonable estimate on how much it will take (and thus how much it should cost) to fix an issue in an unknown codebase.
The mismatch is easily done away with by adding ratings to the system. So much so, than on similar platforms, if you are unlucky and run into a hard issue/prick customer (or some combination of this) too early then it can ruin your chances forever. (It does matter a lot whether you get your 1 out of 100 3* ratings as the first one or the 100th one.)
Interestingly today, I was posting an answer, and I added some polite wordage to the bottom of what was a fairly dry post to offer the [clearly new] user some encouragement. I'm also a committer/developer on the project in question.
I was edited several times by non-contributors removing my desire to be polite. What bothers me is that these words are being attributed to me without my usual polite demeanor.
StackOverflow was pretty awesome at one point in time, but I'm not providing free answers on the site anymore.
I used to provide services like this, but at a much higher rate. $60/hr is not competitive at all, especially if you're charging by the minute without a minimum of at least one hour.
I'd be very happy to answer voice questions about an answer I wrote on SO (or one of my questions without an answer, which is an even stronger sign that I spent time on the problem and perhaps have found a solution which isn't on SO).
Since I already know the answer to the question, the call would simply be me expanding/explaining the answer. It would be very valuable for the person on the receiving end and not much work for me. The value isn't in that I'm "an experienced developer", the value is that I'm a developer that had the exact same problem the asker is having.
Answering some "general" question wouldn't be very interesting to me, and wouldn't have the known value of asking the exact right person. That to me just looks like a gig economy version of Stack Overflow. The mechanical turk version of enterprise development.
I love it, and I think $45 USD an hour is great (it comes to $7,200 USD a month) which is great for the part of the world I am in.
My worry is about the payout, which apparently requires Stripe. Personally I would be happy to receive payment in some crypto (even a stablecoin), or Paypal. But having to setup a stripe account seems overkill for me.
Paypal is f*&^ expensive. They'll enforce a conversion to your native currency which will cost about 2.5% (but no one really knows - at least this used to be the case last time I used them).
$45/hr seems pretty low from here (Eastern Europe) and it definitely DOES NOT translate to $7200/month (which would be an OK income). Anywhere. You can't do this kind of work 160 hours/mo. Though if you have a stable project and you do this on top and only calculate the monthly income for a comparison, then OK. But you'd still have to calculate with at least your holidays.
Not unlike codementor, though the landing page maybe has a bit different value proposition. There they offer mentoring, this one looks more like specialized problem solving. But, of course, the two overlap strongly. E.g. on codementor quite a few of my sesssions were about looking into and helping to solve specific issues.
Those are actually pretty high adrenaline situations, at least in the beginning, because you only have a few minutes to grasp what's going. Or at least it feels like so and at least you need to start showing signs of understanding the problem. It was pretty fun (and stressful, in a positive way) in the beginning.
Cool idea. Where I work it would be much more useful to use it to find someone who can consult on specialised knowledge rather than programming problems. For example we are working on a disaster recovery architecture at the moment. This whole area is new to us so I'm reading a lot about it, but I know that an hour long discussion with someone with expertise in that area could save us months of time and tens of thousands in costs. It would be great if the platform could hook me up with those kinds of people.
$1 per minute? Reminds me of those collect call numbers you would see on a late night commercial, except instead of a "hot girl" you get a condescending asshole.
To be fair, it is sometimes not our fault we are "condescending assholes". It's the environment we are in. It could be from management and/or HR not hiring for the right skillset, to underspecified design details from even more condescending assholes called architects, to leadership not leading but just delegating important decisions (and thus covering their own asses) to engineering teams.
$1/min seems like a very arbitrary threshold. That means, that you as a developer, need to be making no more than $60/hr for your expertise. Following Indeeds hourly to salary conversion that means that if you make over $124,800/yr then you are actively losing value for each minute you spend in this ecosystem.
That may be as intended. In certain states you may never make more than that, in others you may (almost) always make more than that.
In a high cost of living area that would most certainly be the case. Usually these devs will get room mates and do other cost cutting activity to compensate but I imagine they'd be the ones most economically drawn to this solution other than other underpaid parts of the world.
What a great idea. My advice to the OP is to get a few questions going.
I made a similar site for remote work ads but I only had like 2-3 ads on it, expecting the users to fill it up. Well pretty much the initial feedback was that users felt it was a pre-launch/scam site because there was nothing on it.
Look up how reddit was started. It took a long time even for them before actual users started to post.
Does this have screen sharing? Helping people with computer issues with just audio sounds like a nightmare. It is possible to share your screen or share a tab over WebRTC connection nowadays, e.g. Google Meet does it.
If it could integrate with a pair programming thing that let the helper edit the user's code on their machine in real time, that would be even better.
Make this a browser extension that understands what Stackoverflow questions I am digging through. Show me a couple of hihhhly relevant Devs who are _online now_, and let me start a conversation with them from right there in the browser toolbar.
I would definitely use this. Very interesting idea. It's like a 911 for development.
However I do worry that more complex problems cant be solved over a phone call. So it has to be a context less question that's simply over my head. But I have had many!
Cool project. It may or may not work, but with a pivot it definitely could.
Answering questions for money even at $1/min is fine if the payouts work like clock work. I'm not seeing anything that indicates to me how that part works. That's the only USP. The rest is just another non-liquid information market
Users with more than 2,000 reputation and/or at least one Gold badge can give higher-quality help. I would either restrict the site only to these users, or I would put them in a higher-quality section. They would also get paid more, at least $1 per minute. (It currently pays $0.75 per minute.)
I use codementor as well, as a mentor primarily. Got a lot of clients through them, including longer-term clients for my consultancy.
Immediate feedback: If you want me to sign up, 75¢/min is too low, and you do need a mechanism to raise rates. My standard codementor rate is $2.25/min (and goes up to $3/min on some specialized skills).
Feel free to email me if you want some feedback/video chat. I was on hackhands back when it still existed. I'm a sucker for these types of platform. I especially love getting clients looking for actual coaching/mentoring, not just debugging.
Yea things are always a lot faster when you have someone who can tell you exactly what you don't know.
The main difference with Call a Dev is the lack of friction for the devs who need help, and the devs who need work. You just post your question and if someone can help, you'll get a ping with a link to their S/O profile. If you like what you see, you accept the ping and the call starts.
A video is a good idea, I'm putting one together now.
DevOps is another area you could focus on. A lot of people think it's fun to learn programming, but very few want to learn devops (e.g. why does this work locally, but not on AWS?). People will def pay to solve devops problems quickly (source: I've done it many times).
I built this app and want to say first and foremost, thanks for checking it out! I also want to address a few common things I've seen mentioned.
> 75¢ a minute is not enough
I personally don't think it's too low for a job where there's no interview and the only requirement is > 100 rep on S/O. With a critical mass of question flow, answering questions on Call a Dev would beat working at your local gas station making $11.50 an hour.
The job market is ultra competitive for junior-intermediate devs. And there's not a lot of opportunities for them to make money b/w grinding leetcode, sending out resumes, and competing for contracts on upwork in a race to the bottom.
I made Call a Dev as a place to find a companion to help you while you are struggling with something. They may not know the exact answer to your question, but sometimes two heads are better than one and getting a little perspective from someone else is all you really need.
> 100 S/O rep is too high/low
For todays questions about whether aliens exist, where do babies come from, and why do my balls have sperm in them, it's admittedly too high. But until I get some real data points of people requesting refunds (which you can opt to do just by clicking a button after a call ends) it's seems okay.
> Letting people post anonymously/not having a report button
The first app I ever built on my own was all thanks to Michael Hartl and his rails tutorial from 7 years ago. As soon as I deployed that app I was eager to show him what his free tutorial had enabled me to build so I emailed him and he said, to paraphrase, "Cool. I can't really see it though because it wants me to sign up to use it."
Ever since then, I've been relentless about removing friction. That trade off means that a few troll posts are bound to get through and I'm ok with that. This app is still in the interesting project phase so adding a report button is a nice to have that I can add down the line, but after working on this for 3 months I had to get it out the door.
Appreciate all the feedback.
For anyone curious about the stack its an elixir/phoenix/liveview app.
The job market is ultra competitive for junior-intermediate devs. And there's not a lot of opportunities for them to make money b/w grinding leetcode, sending out resumes, and competing for contracts on upwork in a race to the bottom.
I’m not convinced that digging out a new, lower, bottom is the right solution to this issue.
If you were truly stuck on a problem at work, wouldn’t you be willing to spend $2 or even $5 per minute to solve it? That could be the difference between your service paying sweatshop wages versus actually providing a viable option for underemployed devs.
Considering the questions tab currently includes "What the fuck is this?" and "Why is my butt full of poop?" I am going to assume they will turn off anonymous posting soon.
Nice! I was getting tired copy/pasting code from stackoverflow. Now I can pay someone with SO reputation to do this for me. The price isn't too high, so very appealing.
I think it’s a great idea, but it should cost at least $3/min or more to get real talent to join. I’d pay top dollar if you can also guarantee that they’ll solve my issue.
Thanks for checking it out! I could never guarantee that tho as Call a Dev is mostly self-service from the standpoint of the question asker and the question answerer. I wanted to reduced friction as much as possible for both sides of the equation.
If you're willing to pay more money for more guarantee you should check out codementor.io
Many comments highlight the issue of verifying developers. Did you think of leveraging Github Sponsor program?. Developers can be verified via github profile and paid through the same platform.
Relatedly, if you would like help fighting procrastination you may benefit from me as an accountability coach. It's just me personally helping folks stay accountable to their goals with a couple of calls for a couple of hours every week. I'd be delighted for you to sign up. More info here: https://coding-pal.com/
This isn't something I would use to earn any significant amount of money. But I would still want to try this out.
The only feature I need are notifications and subscribtion to tags. Say I want to keep the site in the background and only answer specific questions. I want to subscribte to the `android` tag and get a notification in the browser, whenever an android questions comes in.
I’ve always wondered why StackOverflow didn’t implement a bounty system based on money.
I’ve regularly asked questions on SO that were too specific to get a good answer. I’d have loved to be able to just award $100 for the accepted answer.
It seemed so bloody obvious. But no, bounties are based on karma, you can’t even set a bounty if you’re a new user, and SO tries to make money as a job board...
Do you have any guarantees about how quickly a response will come in? I'm a bit concerned that the low payrate to the dev means that you're likely to miss higher skilled respondants. The payout isn't high enough to justify going FT on it and it feels like it'd be unreliable to try and pickup questions here and there when you happen to be available.
Thanks for the question! Your concern is justified. There are no guarantees. The only thing for certain is that if you don't get a ping within 60 minutes then your question will be deleted.
It's kind of a chicken/egg problem where there aren't enough questions because I don't have enough people to answer them, and there aren't enough enough people to answer them without a steady supply of questions to keep them busy.
I'm reconsidering the pay rate after reading a few of the comments here.
This is a profoundly inefficient way to make money. Since you can’t really serve multiple people simultaneously there’s no opportunity to apply leverage to your work, every dollar you earn has to be 1 to 1 effort, and because the rates are so low very few skilled devs will spend focused time on this.
I wonder how well this idea translates to other engineering disciplines and if there is something about software that makes this idea easier.
What kind of liability is there for engineers asking for help and working in FinTech for example? The attack surface here for social engineering gets bigger.
The only positive use case I can think of, is self taught developers using this to avoid wasting hours being completely lost learning something new for the first time.
Once you're working, you have peers you can ask who will answer any beginner/intermediate questions.
Interesting. I always envisioned a site where more experienced devs could help out newer ones with mock interviews or coding prompts to help quantify self taught coding skills.
I just realized services like this would probably fit the bill if they also offered code reviews.
This reminds me of the premium-rate 1-900 numbers back in the day. I specifically remember seeing one of them which offered tips and strategies if you were stuck in a video game.
I would definitely love to try this out at least once. I could see it being helpful.
I recall seeing a video of the Nintendo support center, where people (kids) would call up for video game help. And these game wizards would help them step through the game.
I would’ve never thought that was a thing, much less a little career while it lasted. But it provided some jobs to people, whose purpose was to just play a bunch of Nintendo video games.
I don’t have the link, but maybe someone can find it on YouTube somewhere.
I worked at the Redmond WA Nintendo of America location. You'd get qualified to do the game support lines. The starting was on setup and installs. There was a 4 week onboarding program teaching you how to easily look up things inside of their custom help database. The database would have model numbers and give pointers on identifying the input select. Identifying the equipment usually involved the son/daughter since they could get behind the entertainment center.
Super Agents were the ones who answered both support lines and used a different part of the same db program. There was some AS400 looking terminal for subscriptions, ticketing, and creating shipping labels for say Pokemon game carts that died. People naturally abused this and would send wood blocks. The desks had TVs and game systems and tons of issues of Nintendo Power magazines.
The floor was huge, hundreds of people grouped in by "streets" named after characters. The cubes were in clusters of six. I recall the game testers, they were in a different area, playing and playing one part trying to replicate issues or discover them. Sounds soul crushing. The neat thing were the marketing displays, like the units that would be installed at stores or showing the old card games Nintendo started with. The employee store was neat, getting that gold controller or other games was a breeze. There was an indoor lounge full of arcade cabinets, usually the cheats listed on the sides and one of the corp guys would geek out on Robotron 2084 and get a score so high it would reboot. Good times.
The Deja Vu feeling was bothering me. Luckily found it pretty quickly..."High Score" Episode 2 on Netflix has the most footage of what you're talking about.
Cool idea - could also hire those people to pass “coding interviews”. Wont reflect bad on candidates as they wont really have to apply interview skills anyway.
Jokes and downvotes aside i support any product or platform that helps devs gain independence.
It’s one of those startups that sounds like a good idea but isn’t... everything that needs this has too much context and is dependant on being able to be answered quickly. The combination is intractable.
So they charge $1/min the customer but only pay out $0.75. That's pretty steep a commission. The usual rate for similar services is around 20% (or between 10-20%).
Why not use Upwork? I’ve found very good devs on Upwork to help me for an hour or two. They make $200-$400 for basically just talking me through something on zoom call.
My current hourly rate breaks down like this. I make adjustments every 6 months.
15 years professional experience + 5 years experience in latest specialized tech stack + 1 hour = $200 dollars/hour for contracts longer than 6 months. Hourly rate is higher for shorter term contracts.
My rate is more cost effective than not delivering the solution.
How/where do you manage to charge this kind of rate? Are you contracting clients directly or through some agency? Also, is this for on premise or remote work? You take on full projects or augment existing teams?
Yeah, so I'm aware that the way to charge more is to tell clients that you'll charge more ;). However, the range people work for is pretty wide and depend on a lot of variables. I was trying to figure out some of those in your case to see whether there is a big difference with the type of the clients you work with.
There are companies that need senior level software engineers to lead in updating legacy enterprise applications to latest tech stack. Development and integration into existing systems are extremely difficult and time consuming. These companies have attempted several updates and most have failed over the years.
These are mission critical systems that process and generates millions and billions in revenue. They have already spent millions of dollars and years of effort in failed attempts.
These are the companies that will pay to get their systems back on the roadmap schedule. There are big names and small names.
- Many entry level programmers at big companies (think Walmart Labs, Target, Goldman Sachs, etc) are too scared to ask managers/team for help. If you work for a large company, make sure new hires feel safe asking questions and seeking help or they will seek it elsewhere.
- The really hard questions they're willing to pay $240 an hour for have too much context for you to grasp. They want you to replace their ORM layer or tell them why their 600 line test case is not working.
- Getting customers is not too hard by poaching them from Stack Overflow, Reddit, CodeMentor.
- Developers make good money working full time and idly sitting at their desk part of that time. Here you only get paid for active time and the idle time in-between will kill your salary. Your motivation has to come from helping other people and watching them grow, not from $$$.