YC application - partner catch 22
I have a potential partner for my business that has all the right qualities, minus the technical skills so I'm faced with a dilemma. The potential partner is committed to the idea and the partnership, has extensive business know-how, is a go-getter who is diving in and working on the idea as hard as I am...even though I have not made a commitment to her yet on a partnership. She is a good friend and we have very complimentary personalities. We resolve issues easily and productively. So why the hesitation? When I read advice from PG and JL, it seems to always indicate that there has to be a technical co-founder. My question is, how technical do they need to be? I am a User Experience professional with years of front-end coding experience. I have worked with developers and have been editing code for years. I'm a pro at DHTML, CSS and pretty good at JavaScript. I understand ASP and XML too. Plus I'm an expert with the Adobe Suite, and in the past have learned ActionScript and Director's old Lingo to complete projects. I have run over 50 usability studies and my focus has been prototyping based on end-user evaluation. I have and always will learn what I need to learn to accomplish goals. Am I technical enough? How is a "technical partner" defined? What's more important...a highly technical partner or a highly compatible and motivated one?
8 comments
Typicall tech co-founder in a early startup person is the one who can code your whole application without any outside help. He/she should be able to go through the code and fix any bugs if it is needed. This person should have good knowledge in all aspects of technologies. For ex: if you are using a website built on .NET Framework with SQL Server backend, generally then this person should be familiar with databases, writing store procs, building web services, writing C# code and lot of other stuff.
When evaluating a tech co-founder you should ask with 2 of your guys can you complete your application end to end and go public without any outside help. If the answer is yes then he/she can be your tech cofounder.
Since in the early days without any funding, money is tight you should be able to complete the prototype/product with almost no resources.
When evaluating a tech co-founder you should ask with 2 of your guys can you complete your application end to end and go public without any outside help. If the answer is yes then he/she can be your tech cofounder.
Since in the early days without any funding, money is tight you should be able to complete the prototype/product with almost no resources.
This. The main thing to look for in a co-founder is someone who can do everything you can't.
You CAN have multiple co-founders obviously, but that raises more questions than anything.
You CAN pay a developer as well, but are they going to care as much as a founder? Probably not.
To me, it doesn't sound like you have a Catch 22, you just have picked the wrong co-founder. That said, it sounds like, from the position you're in, you might not be able to get away with less than 3 founders. If you're front-end, and she's marketing/business, you obviously lack a backend developer.
You either need to find someone else who fits that need, or start learning it yourself. If you already know frontend, Node might present the lowest barrier to entry. The only real alternative is to find someone with business, marketing AND backend dev experience, which is likely going to be a very hard find.
You CAN have multiple co-founders obviously, but that raises more questions than anything.
You CAN pay a developer as well, but are they going to care as much as a founder? Probably not.
To me, it doesn't sound like you have a Catch 22, you just have picked the wrong co-founder. That said, it sounds like, from the position you're in, you might not be able to get away with less than 3 founders. If you're front-end, and she's marketing/business, you obviously lack a backend developer.
You either need to find someone else who fits that need, or start learning it yourself. If you already know frontend, Node might present the lowest barrier to entry. The only real alternative is to find someone with business, marketing AND backend dev experience, which is likely going to be a very hard find.
Well, can back that down a little!
For C#, there are some advantages in using just Visual Basic .NET instead: I'm doing a startup and developing a Web site intended to be a serious business and am using Microsoft, .NET, etc. C#? Never touch the stuff! Instead I just use Visual Basic .NET. The keys are the Microsoft 'common language runtime' and .NET collection of classes, and VB has plenty of access to these although C# has more in some advanced respects. VB has easier syntax, that is, is not 'idiosyncratic' the way C# is borrowing from C/C++. So, VB is easier to learn and easier to read on the page. So, VB is easier for me, the founder and also for anyone I hire.
SQL Server stored procedures? Never touch the stuff! I like the idea of relational database via T-SQL. That's fine. For my work, T-SQL via essentially 'command lines' and/or ADO.NET is just fine, plenty powerful.
'Command lines'? Yes: Use the little program SQLCMD.EXE that comes with SQL Server. So, if want some DB operation, just type the T-SQL commands into a simple text file and give it to SQLCMD.EXE. E.g., can do queries to get data for 'reports' of wide variety. SQLCMD.EXE is not part of the UI but can be just fine for a lot of the rest want from DB.
I don't even like the idea of a stored procedure because it 'distributes' the source code of the software logic in a clumsy way. ADO.NET and T-SQL are so powerful that there's s a good chance won't actually need stored procedures. If someday do need a stored procedure, then maybe download one or have a friend write one for you. Or learn a little about stored procedures and then write one.
To get a good start on the programming, get the relatively well written
Jim Buyens, 'Web Database Development, Step by Step: .NET Edition', ISBN 0-7356-1637-X, Microsoft Press, Redmond, Washington, 2002.
Then get copies of the standard Microsoft Press books on VB, ASP.NET, ADO.NET, and SQL Server. Then learn how to use the several thousand Web pages of .NET documentation at MSDN. Here's what I do: For some topic, say, class System.Uri, I start with what I have and/or a Google search. When I find a Web page I like, then I have my favorite Web browser save it. I save it in a file system directory I have on my development computer. So far I have four such directories, one for each of Windows, VB, ASP.NET, and SQL Server. In total I have about 3000 such Web pages. For the file names used for saving a Web page, I have a little macro that assigns those just as MSDNxxx.HTM where the xxx is the next unused sequence number. In each directory I have a table of contents file that acts like an 'index' and collection of 'abstracts': So for each file MSDNxxx.HTM, I have in the TOC the full tree name ending with MSDNxxx.HTM, the URL of the page at MSDN, the title of the page, the first paragraph or so of the page, and maybe some additional notes.
Eventually you will see how MSDN is organized: In .NET there are 'namespaces'. Typically for each there is a nice few Web pages of overview discussion. Those are good to read. Those overview pages have links to more details, often with more overviews, some tutorials, and some pages "How to: ...", usually quite practical with code samples. Then there are links for each .NET class. In the page for the class there are links to pages for the class constructors, methods, and properties. For some of the classes, e.g., STRING, it can take maybe a few dozen MSDN pages to document all of it.
For typing, I just use my favorite text editor. So I use it to maintain and search the TOC files and also to enter source code. No, I won't touch Visual Studio. The text editor I use is KEdit, and it is my most important tool. So far I have found that KEdit is just fine for developing .NET software. For running the VB compiler, I use either a command line script or use the way ASP.NET works in development (which is excellent). For a scripting language, I just use the old ObjectRexx from IBM, available for download for free. ObjectRexx is both elegant and serious.
I've written software, especially scientific and engineering, for decades in old Basic, assembler, Algol, Fortran, PL/I, C, etc. There read a little documentation and write nearly all of own 'logic'. Developing a Web site with .NET is quite different: In simple terms, .NET has a class for darned near everything so far found of significant general usefulness in commercial software development, especially for Web apps. So, .NET is enormous, a major event in computing.
So, for developing a Web app, heavily what you do, at least the first time, is read a lot and type a little. Your code is mostly just really simple 'glue' code between calls to .NET methods.
So in my code, to me, one of the crucial issues is just what the heck does the .NET class, method, or property do? For the answer, just ahead of each significantly obsure usage of .NET, I in put in comments with the tree names of the relevant files MSDNxxx.HTM. Then one keystroke in my editor will display the Web page of the MSDNxxx.HTM file. To me, a big part of the really meaningful asset of my code is just the code that works together with the links to the MSDN documentation. It's that documentation, readily available right there in my code, that lets me 'desk check' and understand what the code does.
So far, it seems to me that what Microsoft has done with .NET, 'managed code', MSDN, etc. is just terrific stuff, a real step forward in computing.
My experience is that you will have questions. Then you can often get answers on fora and blogs via a Google search. In particular, MS runs several fora where MS experts hang out and provide sometimes some quite good answers. I've gotten maybe a dozen really good answers on some knotty questions. You need to accumulate a collection of such helpful fora.
For more, it's possible to pay MS and get more. Once you have serious revenue, to save time, likely you should just pick up a phone and call an appropriate MS expert and have them just walk you through some 'task' as you take notes. Then the next time, just use your notes!
I believe that you will find that your most difficult issues are not in the programming but in 'system management and administration', e.g., for installation, configuration, backup, recovery, security, performance, reliability, of Windows Server, SQL Server, other MS servers, and the networks you use. For such work, once my business has significant revenue, I intend just to pickup the phone and call MS, Cisco, HP, etc.
You can do it. What you didn't get in a BS or MS in computer science, someone who did can give you a few lectures and get you nicely caught up on what you might need occasionally. But it's not easy getting good with several books, each 500-1000 pages, and a few thousand Web pages at MSDN.
For C#, there are some advantages in using just Visual Basic .NET instead: I'm doing a startup and developing a Web site intended to be a serious business and am using Microsoft, .NET, etc. C#? Never touch the stuff! Instead I just use Visual Basic .NET. The keys are the Microsoft 'common language runtime' and .NET collection of classes, and VB has plenty of access to these although C# has more in some advanced respects. VB has easier syntax, that is, is not 'idiosyncratic' the way C# is borrowing from C/C++. So, VB is easier to learn and easier to read on the page. So, VB is easier for me, the founder and also for anyone I hire.
SQL Server stored procedures? Never touch the stuff! I like the idea of relational database via T-SQL. That's fine. For my work, T-SQL via essentially 'command lines' and/or ADO.NET is just fine, plenty powerful.
'Command lines'? Yes: Use the little program SQLCMD.EXE that comes with SQL Server. So, if want some DB operation, just type the T-SQL commands into a simple text file and give it to SQLCMD.EXE. E.g., can do queries to get data for 'reports' of wide variety. SQLCMD.EXE is not part of the UI but can be just fine for a lot of the rest want from DB.
I don't even like the idea of a stored procedure because it 'distributes' the source code of the software logic in a clumsy way. ADO.NET and T-SQL are so powerful that there's s a good chance won't actually need stored procedures. If someday do need a stored procedure, then maybe download one or have a friend write one for you. Or learn a little about stored procedures and then write one.
To get a good start on the programming, get the relatively well written
Jim Buyens, 'Web Database Development, Step by Step: .NET Edition', ISBN 0-7356-1637-X, Microsoft Press, Redmond, Washington, 2002.
Then get copies of the standard Microsoft Press books on VB, ASP.NET, ADO.NET, and SQL Server. Then learn how to use the several thousand Web pages of .NET documentation at MSDN. Here's what I do: For some topic, say, class System.Uri, I start with what I have and/or a Google search. When I find a Web page I like, then I have my favorite Web browser save it. I save it in a file system directory I have on my development computer. So far I have four such directories, one for each of Windows, VB, ASP.NET, and SQL Server. In total I have about 3000 such Web pages. For the file names used for saving a Web page, I have a little macro that assigns those just as MSDNxxx.HTM where the xxx is the next unused sequence number. In each directory I have a table of contents file that acts like an 'index' and collection of 'abstracts': So for each file MSDNxxx.HTM, I have in the TOC the full tree name ending with MSDNxxx.HTM, the URL of the page at MSDN, the title of the page, the first paragraph or so of the page, and maybe some additional notes.
Eventually you will see how MSDN is organized: In .NET there are 'namespaces'. Typically for each there is a nice few Web pages of overview discussion. Those are good to read. Those overview pages have links to more details, often with more overviews, some tutorials, and some pages "How to: ...", usually quite practical with code samples. Then there are links for each .NET class. In the page for the class there are links to pages for the class constructors, methods, and properties. For some of the classes, e.g., STRING, it can take maybe a few dozen MSDN pages to document all of it.
For typing, I just use my favorite text editor. So I use it to maintain and search the TOC files and also to enter source code. No, I won't touch Visual Studio. The text editor I use is KEdit, and it is my most important tool. So far I have found that KEdit is just fine for developing .NET software. For running the VB compiler, I use either a command line script or use the way ASP.NET works in development (which is excellent). For a scripting language, I just use the old ObjectRexx from IBM, available for download for free. ObjectRexx is both elegant and serious.
I've written software, especially scientific and engineering, for decades in old Basic, assembler, Algol, Fortran, PL/I, C, etc. There read a little documentation and write nearly all of own 'logic'. Developing a Web site with .NET is quite different: In simple terms, .NET has a class for darned near everything so far found of significant general usefulness in commercial software development, especially for Web apps. So, .NET is enormous, a major event in computing.
So, for developing a Web app, heavily what you do, at least the first time, is read a lot and type a little. Your code is mostly just really simple 'glue' code between calls to .NET methods.
So in my code, to me, one of the crucial issues is just what the heck does the .NET class, method, or property do? For the answer, just ahead of each significantly obsure usage of .NET, I in put in comments with the tree names of the relevant files MSDNxxx.HTM. Then one keystroke in my editor will display the Web page of the MSDNxxx.HTM file. To me, a big part of the really meaningful asset of my code is just the code that works together with the links to the MSDN documentation. It's that documentation, readily available right there in my code, that lets me 'desk check' and understand what the code does.
So far, it seems to me that what Microsoft has done with .NET, 'managed code', MSDN, etc. is just terrific stuff, a real step forward in computing.
My experience is that you will have questions. Then you can often get answers on fora and blogs via a Google search. In particular, MS runs several fora where MS experts hang out and provide sometimes some quite good answers. I've gotten maybe a dozen really good answers on some knotty questions. You need to accumulate a collection of such helpful fora.
For more, it's possible to pay MS and get more. Once you have serious revenue, to save time, likely you should just pick up a phone and call an appropriate MS expert and have them just walk you through some 'task' as you take notes. Then the next time, just use your notes!
I believe that you will find that your most difficult issues are not in the programming but in 'system management and administration', e.g., for installation, configuration, backup, recovery, security, performance, reliability, of Windows Server, SQL Server, other MS servers, and the networks you use. For such work, once my business has significant revenue, I intend just to pickup the phone and call MS, Cisco, HP, etc.
You can do it. What you didn't get in a BS or MS in computer science, someone who did can give you a few lectures and get you nicely caught up on what you might need occasionally. But it's not easy getting good with several books, each 500-1000 pages, and a few thousand Web pages at MSDN.
To answer your question: "What's more important...a highly technical partner or a highly compatible and motivated one?" Hands down motivation over skill in my opinion and experience. I've seen one go the other way, or finding a way to get help, but I've barely ever seen the reverse. Also, one quote of Richard Bronson’s I find myself repeating often "Hire for attitude, train for skill."
I think your team might benefit from a 3rd, back-end/server side cofounder. Then between the three of you should be unstoppable with all bases covered. If you cannot find that then you're doomed with or without your non technical cofounder, unless you learn enough to be dangerous, build enough to be able to afford an employee who is a back-end person, who most likely will end up trashing all your code and building it from scratch. I'd sure hope you'd welcome that.
You never said what your dilemma was.
I think your team might benefit from a 3rd, back-end/server side cofounder. Then between the three of you should be unstoppable with all bases covered. If you cannot find that then you're doomed with or without your non technical cofounder, unless you learn enough to be dangerous, build enough to be able to afford an employee who is a back-end person, who most likely will end up trashing all your code and building it from scratch. I'd sure hope you'd welcome that.
You never said what your dilemma was.
Oh, and in my opinion, if you want to be a successful business person, then don't apply the general YC rules to your business. Those rules are generalizations, and partners often make exceptions. Do what's best for your business, and they will see it too. The best way to think of your app is to think honestly and truthfully about your business. If there was no YC and its set of guidelines, would you have considered tossing this person out? How would you have built your app? Think of YC as a co-founder in your business instead of thinking of them as a college you're applying to. They will be impressed by the same things a potential cofounder might be, every company has its unique set of challenges based on which you will be gauged. I am not related to YC in any way, but cannot imagine them looking for the same six things of everyone that walks in.
My point is, give your own intelligence credit, and YC partners might recognize it. If they don’t now then there is always the next cycle and you can try again.
My point is, give your own intelligence credit, and YC partners might recognize it. If they don’t now then there is always the next cycle and you can try again.
I am very glad to have posted this question. The feedback coming in from different perspectives is helpful, and yours is especially refreshing. Thanks
The dilemma is deciding whether to take on a business specialist partner (with all the right qualities) and hire out the development, or keep looking for the perfect match in a technical co-founder. I have many hacker friends who are ready to pitch in and I trust them. We've worked together for many years. But no one that is in a position to leave their day job. I do have a few leads through my network and am still scoping things out. I wanted to get insight from this forum though and glad I did.
I think you're looking at the problem all wrong. The question is not whether she is the right co-founder to present to PG, JL, etc., for the sake of applying to YC. Instead, you should be asking yourself if she is the right person to be building a company with. A startup is essentially a shotgun marriage where the company is the baby. Can you see he being the right person to bring a company into this world? If the answer is "yes" then start working with her. Whatever you do, don't base your decision on what YC is or is not looking for. They look for teams that have qualities that suggest success and that are aligned with their investment thesis. If you have a team that doesn't fit into their investment thesis, but is the right team, then that's all that matters.
I didn't mean to imply that I am picking a partner to satisfy YC....that never crossed my mind! However, I highly value the experience and advice coming from YC partners, therefore, the fact that they place such high importance on having a technical co-founder made me question whether I needed to add a 3rd partner if the two of us don't cover enough "technical" ground. That being said, I now have a new question. I have officially made the partnership with her and I'm so excited about it! I just don't think I could possibly have a better partner. She is as motivated and hard-working as I am and we're up until 1AM every night working on our company. The problem is that I would like to add her to my YC application and I can't access to it anymore. What's the best way to get the message to YC?
There are a lot of issues that will come up through out the process that require "general" technical know-how. My cofounder and I are both technical, so we learn the biz side of things as we go and the tech just happens easily. But your tech cofounder should have a good understanding of back and front end coding. Pick a language and run with it. Also, it's not just about building the web app but also deploying. So being comfortable behind a unix terminal and utils goes a long way toward making your life easier.
Can the two of you build your vision without (trivial) outside help?
Thanks, I appreciate those answers. It definitely makes sense to be an all-in-one team. I am leaning towards learning the additional skills I need. Another option is to complete most of it and hire a developer to complete the rest. What's your perspective on getting the product completed in DHTML and then hiring a developer (a developer friend) to build the .NET Framework with SQL Server backend?
I am not saying that you have to build the site in .NET Framework. You can pick any technology which you like. I specified it just as an example. Before you hire a developer and pay him, you should write down couple of points such as
1) What does the web application do
2) What are the core min features that you need to satisfy first few users (don't get into the habbit of feature creep)
3) What is your intended audience
4) What is the core problem you are trying to solve.
Also as a prototype do you really need the Database and other stuff. It should not matter what database or language or framework you use but you should be able to answer the above questions.
Don't listen to the people who say one language is greater than other language. It shouldn't matter. You should pick one which is easier for you and others who join you later.
Also as a prototype do you really need the Database and other stuff. It should not matter what database or language or framework you use but you should be able to answer the above questions.
Don't listen to the people who say one language is greater than other language. It shouldn't matter. You should pick one which is easier for you and others who join you later.
the technical cofounder has to have knowledge about server side code, database, etc, so at least one of you is totally in picture of what's going on.
also can cooperate with new employees, subcontractors on a technical level, and certainly can act instantly in a given situation...and probably won't complain about working round the clock :)
with such team, i would feel much safer as an investor too.
also can cooperate with new employees, subcontractors on a technical level, and certainly can act instantly in a given situation...and probably won't complain about working round the clock :)
with such team, i would feel much safer as an investor too.
You're fine.
Google Birchbox.
Google Birchbox.