marcamillion

I love this question, but I would like to extend it a bit. Even though the body of the question the OP asked is about software dev specifically, how do you learn other non-software dev best practices when no one is there to teach you?

Specifically, how do you learn how to build a company around your product? Or how to build a board of advisors/directors? What are the best practices around managing those?

How do you shift your cash flow management thinking from spending out of the reserves built up, to spending out of future cash flows?

In other words, aside from building product, there are so many other skills that a founder has to learn as they grow their company. I assume most YC companies learn this stuff from YC or their batch mates that have done it before them.

But for those of us not in the Valley or in YC, how do we learn this stuff?

Also, I guess the main premise behind my question was how do we learn best practices in company building in general? Ie how do we build the machine that builds the machine using best practices for everything, or as many things as possible (for both tech and non-tech companies)?

When you say handle sales, do you mean handle product development for each project that you have landed? So basically an account manager? Or do you mean someone that goes and lines up new sales? Just want to get specifics, so everybody understands what is being discussed.

Lol.....that's why I asked for clarity, because I didn't want to assume :)

Partially a cop-out. All "lifetime access to" responses, aren't realistic responses.

Care to expound on this further?

Definitely a copout :)

Can you explain a bit more....do you have a way of gathering those leads now? A way that can potentially be automated?

Hi WingH, Can you update your profile with contact info, for anyone that would like to contact you about potentially providing any of the above.

Thanks.

I see in a lot of threads that people talk about "many companies that do third-party fulfillment". As someone actively in the market for such services, can anyone provide some links to a few services please.

I know about the YC Company - Ship Bob.

What are some alternatives?

Thanks!

I literally lolled at this comment.

I assume this is sarcasm...just want to verify :)

I think one of the biggest points being missed here is that Uber may very well be an iceberg. Meaning moving people is literally just the tip, but to get to the rest you have to subsidize the hell out of moving people.

The fundamental problem with what Benjamin pointed out with the lack of benefits with network effects, only applies when you are thinking about moving people and only thinking about 1 singular network. Sure, you and I may only want to go from point A to B within my city.

I think the real value for Uber is when they have this huge network of constant activity between most points within any city/country, the value proposition to carry cargo goes through the roof.

Imagine being able to send a document (or a package) anywhere else within the city in 20 minutes.

Amazon is experimenting with Drones for quick delivery, but just imagine being able to purchase something on Amazon, Walmart, Target, and get it within 30 minutes from an Uber driver.

Their network effect looks different than most others, say Facebook, because the real value is a network effect of networks. i.e. they have highly concentrated networks within cities, and they have a high concentration of city-networks within states, and outward.

The clear value proposition there is one can easily move a package from your house through your network within your city, to another network in an adjacent city, and on and on to say the next state.

Right now, sure you can ship something 'overnight' via FedEx relatively long distances within the US but technically it's not REALLY overnight. Technically, you have to reach the FedEx store before some cut off time (say 12 noon), so that package can then be taken to their sorting facility and make it out on the flight that night.

Imagine if there is the a real-time network where at any moment any package can be placed on the network and be on the most efficient route to the destination immediately. That's obviously the holy grail, but no longer do you have packages sitting in sorting facilities and waiting on bulky planes to take off.

They may not have these plans, but I have no inside knowledge and that's one clear advantage I can see of having a network where something is always being delivered between almost any 2 points within the network.

I assume that all of these investors are not dumb and neither is Kalanick and his team, so I suspect there is a much larger logistics play than we can imagine.

Just like Tesla isn't just a car company, but is also both a commercial power (Southern California Edison) & oil company (Exxon) plus maybe an autonomous delivery fleet all in one, I assume Uber is something similar we just can't see it yet.

Yeh this is exactly the problem I am trying to solve.

Not quite sure how to do that though, and if it is even possible.

Not sure if Upwork does it, but Upwork is the antithesis of the type of network that I am interested in building.

The projects you get are usually bottom of the barrel.

I have no interest in those types of projects.

I hear you. The term 'spamming' is quite subjective.

Either way, the fact of the matter is sending cold emails is actually a powerful way to build sales -- there is a simple reason so many people do it, it works.

Like it or not, many startups do it.

So I was simply pointing this out for the OP's benefit.

I appreciate the fact that you guys made product decisions, but this statement was a little red-flag to me:

For instance, a startup of 3 would not need to send 500 emails a day, that is a clear red flag.

There are many startups that do cold outreach, via cold emailing, that could do roughly 500 emails per day. I am not sure how typical this is, but as someone that has started doing sending cold emails as a direct sales tool, it seems to me that you just haven't had any customers like that yet.

If you do though, don't assume they are spammers.

Yes, same here.

I was surprised actually. I do usually see this in some FOSS projects, but never in a commercial project.

It was actually quite refreshing.

Octopress + Github Pages + CloudFlare = Free, Stable that can handle a good amount of traffic.

Yeh I think everybody is saying that if you are already hosting the source on GH, there is no need to host the static files on S3 -- you may as well just host it on GH:pages, because that will be free and you get all the benefits of S3.

None...which is the same number as the number of interviews I have gotten.

So I think this the article perfectly sums up a thesis I have developed about these economics indicators and the way Economic Growth & Productivity is measured.

All of this is intuitive, and I have no hard data to back up my theory, so I would love for someone with much more econometric experience than I to prove/disprove this thesis.

Thesis: The global economy and US economy in particular is going through a radical productivity growth spurt and shift that is not currently being measured properly.

This article is the perfect example of this. You mean to tell me that during a period where there has been an explosion of independent contractors of many different types, we have seen "Startups Weighing on US Economic Growth"? Surely that can't be true and fully reflective of the reality.

There are two aspects to this. There is direct job growth within startups (so Employee headcount) and there is the issue of what is a 'new company'.

Direct Job Growth.

It is hard to argue that the average startup isn't much more productive today than it was 10 years ago, much less 20 years ago. Particularly tech startups. For tech startups it's most glaring, because you have things like AWS, Heroku & the App Stores that allow you to deploy a relatively easily scalable, product that can reach hundreds of thousands/millions of users/customers as a team of 1 - 5.

No longer do you NEED someone just to manage 1 server, or even add additional server capacity and deal with Colocation-related issues and all of that stuff.

You also no longer need to pay huge licensing fees for development software platforms and developer tools. So the barrier to entry to shipping has dropped to 0, basically.

So whenever a startup raises a nominal amount of money, it can go into much more high-value jobs (like customer acquisition and customer support) for which there isn't always a direct correlation between each incremental dollar in revenue earned with the number of people you hire to support that revenue. In some cases there is, but in many cases there isn't. Or rather, the up scale hiring process is horizontal rather than vertical. Each customer support specialist can handle more support tickets today than they used to 20 years ago, for much cheaper. Aka, the support systems that multi-national companies have always used are now available for much cheaper and often much better to startups at $50/employee in many cases.

When you think about the various aspects within a growing tech startup, you can see this same principle across all disciplines (even including HR and Employee benefits via services like Zenefits and its competitors). So the productivity that can be bought with each marginal dollar invested in a nascent startup is so much greater as a result of these highly, specialized and in many cases very economical services that can be leveraged from third-party providers, than had been the case 20 years ago....yet these articles and current econometric models would have us believe that productivity growth has flatlined. Really?

What is a 'new company'?

While there may be significantly less direct job-growth (as a result of the issues I highlighted above), there has been an explosion in the number (and types) of marketplaces that have sprung up that allow customers/users to be independent contractors. Not just typical web dev/writing/etc. But your excess space (AirBnB and all its clones), your excess vehicle (Uber, Lyft and all clones and derivatives), your excess time (Instacart, TaskRabbit, Mechanical Turk, etc.), and any other service that has sprung up that allows random people to do random gigs from a marketplace of gigs of different kinds.

So yes, the Ubers of the world no longer add significant employees to their payroll to service increasing revenue as a part of operations, but by creating marketplaces where random people can earn a living, doing things they previously never did (or even considered doing), surely has contributed significantly to economic growth in ways that aren't currently being measured properly.

Those people likely haven't registered legal entities, they probably just have a bank account, so they won't show up in "new company" data. But I bet if there was a way to measure those non-registered, independent contractors across all of these problems and you contrast that figure with the same category 20 years ago, you would get a much different picture of the US economy and productivity growth.

I could be wrong with all of this, but every time I read one of these articles....that's what jumps out at me. The disconnect between what is being measured and reported in articles like these, and all the products/services we see being launched on TechCrunch, Product Hunt and HN that significantly improve the average person's earning power by both being able to sell said product/service or sign up to be a participant in that marketplace, has always been jarring.

Let me know if this makes sense to anyone and if I am missing anything.

I would love to crystalize this thesis and ideas some more, to do a full write-up in a blog post so please poke as many holes in this as you can.

Thanks!

What's interesting about the timing of this story is that I have recently been playing around with Upwork for some freelance work and the issue I am having is actually a different one.

I have come to realize that there is a fundamental problem with the marketplace itself. I don't think it matches clients to freelancers properly.

I did two exercises. I posted a few positions as a 'buyer', and I got a lot of spam (i.e. non-personalized, crap postings to my position/gig/job that was obvious they never read it). I got more than I expected, which makes it difficult to weed through and find a freelancer I want to work with. Granted, I didn't want the typical "low-ball" freelancer. I was looking for a freelancer that knew what they were doing. Alas, I was unsatisfied with the results and ended up not finding what I was looking for.

I also responded to gigs as a Ruby developer. What's remarkable is that it is literally very, very difficult to get any work, much less the type of work I would like (high-value work with a handful of clients, potentially doing on-going work).

I first started off with a relatively high-ish hourly rate for UpWork ($80/hr for someone with 8 years of Ruby & Rails experience and 15+ years of web development experience overall). Because I had no 'history' with the platform, that didn't work. I filled out my portfolio, and responded to each job in a very custom way detailing the specifics of how I would tackle each job I was submitting a proposal to. This took much longer than just spamming, and was more mentally taxing, but I figured I could make up for my non-Upwork-track history by putting more into my proposal. No dice.

I then dropped my rates (down to as low as $40/hr) just to test, still no dice. I didn't even get responses.

Then, I assumed that maybe my proposals weren't robust enough or maybe I wasn't communicating my capabilities in my portfolio properly enough, aka I was being hit with a 'portfolio tax'.

So to get over this, I decided to actually bid on fixed budget tasks that were very specific in what they want and overlapped with specific stuff I have done in the past -- specifically "B2B Lead Discovery" or "Website Scraping" for something.

I recently have been playing around with scraping websites for different types of leads, particularly B2B, and so this suited me perfectly.

I then started applying to some of these with not just the specifics of what I have done, how I would tackle their specific task, but I would even send them sample results for similar leads to what they were asking for. So say someone was looking for wedding planners from each state (an actual job posting) where they would need the $CompanyName, $Website, $Email, $PhoneNumber, $Address. I replied telling them I have experience doing exactly this....in fact, I recently did this exact thing for accountants, so I replied explaining what I have done and how I can help them and I sent them a CSV file with a list of sample accountants, along with a picture of my script producing those results.

In one case, I crawled the specific website they wanted crawled and showed them pictures of the script doing that and then I gave them a suggestion based on what they were looking for and what I found. There was a disconnect between what they wanted, and what could be technically scraped from the website (they wanted email addresses for all users on MySpace to be exact). So I informed them that unless MySpace has an API that gives out this information, and unless you are looking for email addresses that people post within comments on the music throughout the site, this is a waste of time and I provided proof from my script.

Suffice-it-to-say, I did a lot of work on each proposal. I did about 7 - 10 of these specific proposals for scrapers, and about 15 - 20 other specific but not as specific proposals. I also didn't change the price they asked. So if they said their budget was $10, I replied with all of the above with a $10 budget. This is crazy, I know...but I did it just to experiment.

The results? Not even 1 reply. Not even 1. You can see screenshots here [1].

Yes, my portfolio on Upwork could be weak (although I doubt it because I think it looks pretty robust), and my profile could be a deterrent (because the language I use is a mismatch to what these clients are looking for) and my rates could be high relative to the rest of the marketplace, but the real issue is just an overall non-response from ANY of the 20+ proposals I submitted over the period of a week.

Something feels fundamentally broken with that, especially when considering my experience with the other-side of this experience.

I believe that there is some middle ground between the "elitist" Toptal and "broken" UpWork. So, I would like to try an experiment.

Do you have any high value ($30K+ -- note this is a floor, just to weed out inappropriate clients) development projects that you would like done? Either generic projects where no tech stack is specified or Ruby and Rails jobs for starters. I won't specify the types of projects, but something where you would prefer a "high-quality" developer help you see it to fruition rather than the cheapest developer you can find. Perhaps you have tried other developer services/gig boards and are unhappy with the process.

Do you want a product manager to help drive the entire process for you, from beginning to end?

If this sounds interesting to you, please send me an email to: marc+hnexperiment@mymvpblueprint.com.

If I can find a pattern for how to find these types of projects consistently, I would love to work with other developers to fill these needs. Until then though, let the experimentation begin!

[1] http://imgur.com/a/MjHYk

I tried to post a refactor request, as a C++ developer I can't help but noticing C++ is not in the dropdown list :)

It is there...just as "cpp". The "+" would mess with the form & parser input :|

I was thinking from my own experience, I have worked for big companies that have tons of legacy code that were written in subpar standard, sometimes developers were asked to refactor a piece of code written in a technology they are not familiar with. Have a refactor tool like this can serve as a mini contract to outside developers (or QA in some perspective).

Hrmm...this is a very interesting use-case that I never quite thought about.

Perhaps even encouraging companies to increase their employee productivity by using this marketplace to refactor old code and have them focus on more valuable tasks.

Very interesting idea.

Your idea about the group is interesting. Not sure if that exact implementation would work, but I do get where you are going.

I love these thoughts.

Thanks again!

This is interesting, how many languages do you support?

Well..it depends on what you mean by 'support'. There are two modes. You can either make your code runnable or not. In terms of the code being runnable, 32 languages. In terms of pretty syntax highlighting, but just static code....much more than that. If you try to add a question or suggestion, you will see a dropdown list that contains all the languages. I believe it is roughly 90+ languages for static, syntax highlighting.

This can be very useful for people learning new language, for example I'm learning angularjs and javascript, I write a lot of code with them that work but I'm pretty sure they can be improved a lot by a angular/js expert.

Hrmm...this is a very interesting perspective. I've never quite thought about it from that angle, but you are right.

How about a feature to open a bounty and pay real money to people who refactor the code, and only the refactoror and initiator can see the code?

So I have thought about implementing the first part, and I probably am going to facilitate this.

The latter part is interesting, I guess the obvious question would be.....why would we want to prevent people from seeing the code? Because the code belongs to a proprietary project or for other privacy reasons? I am not being facetious, just genuinely trying to dig into this issue fully because I have gotten similar feedback about making it private.

So assuming we decide to allow a private feature, I guess the obvious question is...what's private? It would have to be only the refactors, right? Otherwise, no one will know what to refactor and whether or not they want to bother trying.

Taken further still, so let's assume that we can settle on a reason and mechanism to keep it private and allow the refactoror to see it, I would think that it would be in your best interest to have as many people viewing both your original post and the refactored suggestions so you can be sure the refactor that was produced is indeed the best possible refactor for your desired solution.

I can empathize with the need for privacy and keeping some code proprietary, but I guess the real question is, is it a safe assumption to make that the person that does the first refactor does the best job? Or even the person willing to jump through the hoops to see your refactor, does the best job? I would figure the only way to know is by making everything open and open to feedback from the community.

That way, if I submit something to you, someone else can build on my suggestion and we both get credit. You get the best refactor possible, and we both get credit and likely both learn.

That's my initial thinking on this stuff, but I am open to hearing other perspectives for sure.

Also may I ask which editor do you use in the new question form?

Summernote - http://summernote.org/ - it works beautifully with Bootstrap and Simple Form (if you are using Rails).

By the way, I would love a suggestion with finding content to beef it up a bit.

I know GH is a huge source of OSS code, but how do I (easily & quickly) find code I can use that could be a good candidate to be refactored?

Ideally, it would be good if the code didn't have too many dependencies or it would be easily understandable so I can explain what the goal is for the refactor.

I tried just searching GH for the phrase "please refactor me", but that turned up a lot of old code and crap...basically.

I would love to be able to go through some of the large OSS projects (Rails, jQuery, any other large ones) and quickly find code that needs to be refactored and preferably have an explanation.

Anyone have any suggestions for how I might be able to do that without actually fully understanding the code snippet I am referencing?

Edit: Hat Tip to pkroll for also confirming that this should be a viable strategy to get content in his comment on this thread.

Great idea. The truth of the matter though is, I don't really want to make it too restrictive.

I would love to see what people use it for and then we can kinda pivot from there.

The general purpose is, do you have some code that you think could do with some loving, post it and let's see if we can get the community to refactor it.

If enough people start using it for other things, then I would definitely shift in that direction if that's where the usage is taking me.

Yeh the domain name 'refactorme.com' was taken, so I had to look for an alternative.

My wife suggested the 'plz' and while it wasn't perfect it was much better than the others I thought about...so in the spirit of shipping something, I ran with it :)

Thanks! :)

Thanks much :)

Can run the code... but, not the tests, which would make in-place refactors considerably easier.

That's definitely something that I would love to implement in the future - if I can find a way to make it viable. This was just v1 to see if there is even a path that makes sense.

Clicking Gear for settings doesn't go back to edit window when you're done and you click the Gear again, which is how I first tried to get back to the editor.

Yeh, I am using the API for Glot.io -- which provides no way to customize that iFrame. So I am stuck with that functionality for now.

No place to set "license" specifically for each piece (and maybe a default in your account settings) making it impossible to take the code and use it without opening Pandora's box.

Ohh...this is interesting. Never considered the 'license' bit. Is that really an issue though? You can't add a license on SO, and that doesn't seem to stop people from pasting code snippets.

So little there, and all six days ago... you could grab examples from open source projects and put them in, if you could do that while obeying the license on the code (which comes back to at least showing the license, and maybe also linking to the URL for that piece of code or project).

I agree fully, and it seems like it should be pretty easy to do from OSS projects, eh? The issue I ran into was identifying what snippets of code really needed to be refactored and why. That's the counter-intuitive bit, in order to be able to explain why something needs refactoring you have to fully understand what it does.

I can most certainly do that for some Ruby code (where most of my experience lies), but once we get into other languages I am kind of outside of my depth (to be frank).

So another alternative is to simply search GH Issues for the term "please refactor me". I tried that too, and most of the immediate stuff I found are usually too old or too basic.

That being said, if you have any suggestions on specific bits of code from OSS projects you have used that you have seen that could be a good candidate please feel free to share them.

I am definitely going to be looking myself too. It's just not as 'slam dunk' as it appears at first thought.

As this is ultimately "one piece of code with several examples of another piece of code to replace it" despite it being for refactoring, you can imagine it'll be used for "this was a bug, this is the fixed version." Might be useful to provide specific support for that somewhere down the line.

This is very interesting and I would love to hear more about what you are thinking. Do you mean like someone would post a snippet of code that has a bug and ask for the community to fix it? Then the site would track the various iterations and show the final version? If so, that does sound intriguing and is inline with some other ideas I have.

If not, I would love to hear what you meant or just some more thoughts on potential future direction.

That being said, thanks for even taking the time to give me such a thoughtful response. I really appreciate it.

For the longest while I have wanted a site that just focused on refactoring, so I finally just built one and shipped it. A bit light on content right now, but you should get the idea.

So did I. He had waaaaay too many things. I have an app running on Heroku that handles his peak traffic, daily, for just $16. The only difference between mine and his is the SSL add-on...but that's $20.

He loaded up too many things.

1 web dyno, the free Heroku Dev DB, SendGrid free starter plan, Google Analytics, and the SSL would run him about $27 ($20 for SSL and $7 for 1 paid Web Dyno so the app doesn't go down every 16 hours).

That's all he needs. $250/mo is OVERKILL that he doesn't need.

Pretty sure this is not a Heroku issue.

P.S. As others have pointed out, this is a simple checklist app. He doesn't even need the SSL add-on until he verifies this is something people will pay for. It's not that serious bro. People can login with their personal info over a normal HTTP connection for now.