Software engineers: would you use a gig economy platform for SE tasks?
8 comments
Neat initiative, good luck! Your TAM analysis will probably have to exclude any enterprise software clients (i.e. anyone that has SOC2 requirements)
I think people in foreign countries would want to do this.
I think companies in the US would want this for REALLY simple tasks.
I think companies in the US would want this for REALLY simple tasks.
I've used elance (many years ago) to pick up extra part-time work and that worked out well. However, my general feeling these days is something like, I am paid pretty well at my day job already so I want to do fun things in my time outside of work. Others may be in a different position where they want to work for themselves here and there, but the people I know who do this tend to establish a client base that comes back for repeat work pretty quickly and then they don't go looking for new stuff any more.
Great point. Thanks for the honest feedback!
It'd be great if you had a page on the site explaining the process for clients. I see the strong.network tie-in, but how do projects work? What if you get a dud? Is this like Upwork or not?
I'm missing a huge piece of the process it seems. Sounds like I need a "how it works" section.
Thank you so much for your honest feedback!
Thank you so much for your honest feedback!
Seeing how awful the gig economy is for every worker driver, deliver person, etc... absolutely not.
Nope. There are two big problems with moving software from the freelance/contractor model to the gig model.
1. Problem definition. Getting a business problem to the point of just sitting down and solving it like it’s leetcode is really hard. That’s a big chunk of the job of software engineers. (Side note, I think most software engineers would be better served by considering these tasks separately. Write the problem then read it). Other gig work is much more clear. I’m at home and I want to be at the airport. I want this list of groceries. I want my apartment cleaned.
2. Order of magnitudes differences in time between projects and between abilities on the same projects. I have a good idea how long it takes to get my groceries, let’s say within a factor of 2. Can someone estimate that accurately on a problem they don’t know how to solve? Even on one they think they could solve but haven’t solved yet? My experience points to no. It’s still an issue even if you charge by the hour. Come in under time and my fixed opportunity costs of finding this job don’t get amoritized properly. Go over and even if I get paid they might leave a bad review. On the client side, how do I know the problem is 10x bigger than I thought and not that the dev is 10x slower? I hope it’s uncontroversial that the performance distribution on software engineering is much greater than that of driving.
1. Problem definition. Getting a business problem to the point of just sitting down and solving it like it’s leetcode is really hard. That’s a big chunk of the job of software engineers. (Side note, I think most software engineers would be better served by considering these tasks separately. Write the problem then read it). Other gig work is much more clear. I’m at home and I want to be at the airport. I want this list of groceries. I want my apartment cleaned.
2. Order of magnitudes differences in time between projects and between abilities on the same projects. I have a good idea how long it takes to get my groceries, let’s say within a factor of 2. Can someone estimate that accurately on a problem they don’t know how to solve? Even on one they think they could solve but haven’t solved yet? My experience points to no. It’s still an issue even if you charge by the hour. Come in under time and my fixed opportunity costs of finding this job don’t get amoritized properly. Go over and even if I get paid they might leave a bad review. On the client side, how do I know the problem is 10x bigger than I thought and not that the dev is 10x slower? I hope it’s uncontroversial that the performance distribution on software engineering is much greater than that of driving.
I wouldn't personally want to do this because gig economy work seems inherently risky and low-paying and my current job is not those things, but I see a more fundamental problem with the very idea of gig economy for software engineering. Gig economy is ideal for short-term, one-off tasks that are well-defined and effectively commoditized. Give me a ride somewhere. Buy my groceries. Carry a bed upstairs for me.
Once you get into something as complex as install a new shower for me, let alone design and install a new shower for me, there is a lot of value in things like long-term client/contractor relationships, word of mouth referrals, and the jobs need to be individually scoped and price negotiated. It gets further and further away from making any sense to have it be app-mediated. You could, of course, have software that aids contractors in managing customer relationships and scoping work, but that's just plain old-fashioned CRM and ERP, not gig economy.
Where I could see this maybe working is for very small discrete tasks, like lend me your computer for a consensual botnet to stress test my infrastructure, help with reviewing this single documentation pull request to ensure it's readable to an average person who isn't on the team already, but I don't think customers are largely in a position right now to be able to individually purchase something like 4 hours of work. The overhead isn't worth it from enterprises with convoluted and expensive purchasing processes. But there is not a sufficient value add for small businesses on top of just "hire a single tech guy to do everything for you," but then it's not a "gig" any more. It's just normal contract work.
Once you get into something as complex as install a new shower for me, let alone design and install a new shower for me, there is a lot of value in things like long-term client/contractor relationships, word of mouth referrals, and the jobs need to be individually scoped and price negotiated. It gets further and further away from making any sense to have it be app-mediated. You could, of course, have software that aids contractors in managing customer relationships and scoping work, but that's just plain old-fashioned CRM and ERP, not gig economy.
Where I could see this maybe working is for very small discrete tasks, like lend me your computer for a consensual botnet to stress test my infrastructure, help with reviewing this single documentation pull request to ensure it's readable to an average person who isn't on the team already, but I don't think customers are largely in a position right now to be able to individually purchase something like 4 hours of work. The overhead isn't worth it from enterprises with convoluted and expensive purchasing processes. But there is not a sufficient value add for small businesses on top of just "hire a single tech guy to do everything for you," but then it's not a "gig" any more. It's just normal contract work.
Potentially yes but they would have to be actually good jobs, at good rates, with guarantees that payment can't be reversed(and other similar scenarios that upwork and the like are famous for).
So basically potentially yes, but the reality is that it's unlikely for anything to ever live up to this standard.
So basically potentially yes, but the reality is that it's unlikely for anything to ever live up to this standard.
Check out the landing page for an overview: https://joincottage.com
I'm happy to add anyone to the beta if you want to sign up.