Is starting a Web 2.0 company cheap?

May 6, 2006

Not to get beef with anyone, because I generally respect the crap out of all the perpetrators… but I have to respectfully ask some of you Web 2.0 entrepreneurs to reconsider whether it’s truly responsible of you to go around telling people how it’s so cheap to start Internet businesses now that folks should all just quit their jobs and follow their entrepreneurial dreams. It’s a very sexy line, but there are so many caveats buried in there that I get deeply uneasy whenever I hear it.

First off, let’s look at the whole “hardware and software are so much cheaper” theory. Maybe companies at the very beginning of Web 1.0 were prone to blowing tons of bucks on custom appservers and Oracle licenses and the like — but the truth is that by 1999 or so, a lot of Web companies (like the one I was working for at the time) had wised up to Open Source software. We used pretty much all no-cost software — C, gdb, PHP, Apache, Bugzilla, etc. — and it was just understood that no engineering budget would be going towards commercial software except where absolutely necessary. Yes we ran on Solaris boxes in production… but we only had like 24 of them. I can assure you that the Facebooks and MySpaces of the world have more than 24 servers, and those fatty dual-core dual-proc Opterons aren’t all that cheap either. So I can’t personally agree that hardware or software costs were a huge factor in why most Web 1.0 companies blew though all that money. When you come right down to it, staffing and possibly user acquisition are still the biggest expenses in this business.

Which brings me to my next point: even though it’s easier than it used to be, it’s still almost impossible to get a Web product out there without financial backing unless you’re an engineer, or are in a position to get engineers to work for you without cash. Now let’s look at the ways this can be true:

1. Your devs are students. Hey, it worked for Zuckerberg; but it’s not an option available to most people over 21.

2. Your devs have jobs and are working on your project in their spare time. As many an Open Source project lead can tell you, this is an inherently unstable situation for any length of time. One thing that happens frequently is that an unfunded project gets a certain amount of buzz and then the members start getting job offers from better-funded business that need similar skills… and how you gonna keep those devs down on the farm when they have a full-time well-paid job doing what they love?

3. Your devs are already rich from previous jobs, and can afford to work without cash remuneration. My question here would be whether these developer’s skills are really up to date. I know a gazillion enterprise Java types right now who are very interested in learning LAMP and AJAX and Ruby on Rails… but are they going to be your best bets for an innovative Web 2.0 business? With all due respect, I have not personally hired any. I recruited engineers with cutting-edge skills who could not have afforded to work without pay — and you know, even though I’m a webdev we had very little overlap in our skillsets, so it’s not like I could have produced a working alpha without the other guys.

Even if your product team can afford to work without pay for some time, there’s a limit for anyone. So the question is, how long does a Web 2.0 beta take to create? This is where I think people can sincerely but severely underestimate how long it will take to complete development on what I call “a complete thought”: getting your software project to a state where users can complete at least one core transaction smoothly, and therefore beta testing even starts to make sense. We all look with envy on the projects that were able to get a complete thought to the users after two or three months of intense effort but minimal cash — like Del.icio.us, Meebo, Tailrank, Meetro, Reddit, Upcoming, PBWiki — but a quick glimpse of TechCrunch’s Web 2.0 index will show you that these are sort of exceptions to the rule. For all that Web 2.0 is supposed to be about “just get something out there fast”, it turns out that a lot of what I would consider the most valuable and useful new applications — Jot, Ning, Skobee, Zimbra, Google Calendar, Dabble, etc. — took months if not years to get to the “complete thought” stage. And I am not giving myself a pass here either: I originally thought Renkoo would be a complete thought at about 6 months, but a usable and scalable beta turned out to take closer to 14 months. I wish it weren’t so, but that is the fact Jack.

This is not to suggest that people shouldn’t start the projects that make them happy. But I’ve had some bad experiences recently that lead me to believe that the “you too can start a Web 2.0 company for no money down and no money a month!” meme may have gotten out of hand. I’d advise anyone who is serious about starting a company to listen to other entrepreneurial types — like Caterina, or the folks at 106 Miles — before drinking too much Kool-Aid.

13 Responses to “Is starting a Web 2.0 company cheap?”

  1. George Schlossnagle Says:

    Solaris is free to use now on most commodity hardware. 🙂


  2. I think by “Solaris boxes”, Joyce means “Sun hardware”.

  3. Peter Yared Says:

    Joyce, I am _so_ disappointed to learn that you are running on Sun boxes. 🙂

    I think it is important to define what “Web 2.0” is. Most mashups (ie HousingMaps.com) don’t need to raise any money, especially if they are piping in other people’s servers directly into the browser. A viable company with a differentiated product clearly needs to have capital to grow.

    Enough said. Great post!

    Peter

  4. Troutgirl Says:

    Heh, you guys are too smart. Yes, in 1999 they were Sun boxes running Solaris. I guess in theory we were paying for the hardware and getting the OS for free? And Peter, I’m shocked to think you could believe I’d be using Sun hardware now! 😉 I run Linux everywhere: database, web server, pubsub server, and laptop.


  5. I think you nailed it about the cost reduction provided by open source — people who know what they’re doing were using open source in 1999. Usually when I hear the meme “startups are so much cheaper to do today because of open source” I find it’s usually coming from someone who’s taking an advocacy position on open source, not necessarily someone who’s actually done a startup.

    Bearing in mind that the big expense is indeed people, not hardware or software, my sense is that costs have gone down primarily because there are fewer time-wasting blind alleys for developers to go down these days. The good frameworks have bubbled to the top and the bad frameworks have met the merciless invisible hand of Darwin. All this is only tangentially related to the adoption of open source.


  6. I disagree with the writer of this blog. It’s obvious she has worked for companies that start out with large expenses. If I understood correctly, she has never developed a major project herself.

    On the other hand, I am an independant developer without programming skills and you very well can create applications at a low cost by utilizing freelance programmers like the ones found on Guru.com. You can check out my work at ezLAWSUITS.com; Mystockroom.com, Mindfreeze.com. I will not do any project unless there is true innovation and patent protection.

    Just be careful of who you get to code your work. Some tend to over exaggerate their skills. But when you find a good team. The cost could be as low as $14 hours for a programmer holding a masters degree in computer science. They tend to bid low because a lot of them are just doing it for (1) extra money from their regular job or (2) they are looking for interesting projects to take charge of so they can insert it into their resume and get better jobs.

    The key to success on the Internet now is “innovation”. Major Internet companies are buying up many little companies with an innovative idea. Take an old way and re-invent it with new and innovative functionality then patent it. it’s not that difficult. Start out with a provisional patent, which gives you one year to final the actual patent application. Keep in mind that in America, you automatically have one year patent protection when you invent something. America is based on “first to invent” while Europe is based on “first to file”. So if you want international protection, file right away and apply for international filing. Otherwise, have anyone looking at your idea sign a confidentiality contracts. The best is to file the provisional application because your programmer could steal your idea or tell others about it. Like in the lawsuit I filed against one programmer, the Judge stated in a recent order quote “the court finds that Mehmet has adequately demonstrated the merits of his claim that (programmer) breached the non-disclosure provision of the contract”. The idiot programmer started posting my patent pending project on the internet on employment sites to entice other employers to hire them, which included a direct competitor of mine and the fool even sent me a copy of my own invention in an email because he didn’t realize it was me he was responding to.

    Innovation will bring you much attention and money. Servers are not that expensive when your developing and tweaking your site. I’ve paid as low as $25 per month and currently pay only $40 a month for a dedicated server. As the business grows, so will your ability to pay for more, which only makes sense that expenses will grow.

    The downfall of the .com era, as they say it, was NOT due to the Internet business model being worthless, it was due to “ignorance”. A lot of young people everyone thought were brillant just because they understood code and people invested millions with them thinking that these smart programmers knew how to put together a successful business when they didn’t. I’ve met a lot of idiot programmers. Don’t take it for granted that a person is intelligent simply because they can lay down code. Laying down code is like someone who knows how to speak a foreign language. Some do it good, some do it bad. But it never tells you that they are intelligent. That is how you should treat programmers. That makes me think of one story about an idiot programmer. This fool created an attorney case management program and inserted code that would take the program down when a certain file number was added by the law firm. The dummy called the firm and asked what file number they were working on at the time. He did this more than once. Of couse the attorneys being more intelligent than the programmer caught on when the program went down when that file number was entered. They sued the programmer and got a judgment against him for $25,000, which was all the money they paid him to build the program and the court also hit him with punitive damages. Now that’s an idiot.

    Knowledge of code doesn’t make them any smarter; but because of ignorance amoungst many people, they think these programmers are so intelligent and many of the programmers play it off as if they are when they are not. That’s why the programmer mentioned above having delusions of grandeur thought he could outsmart a lawyer when lawyers tend to really be intelligent however arrogant.

    You must listen to what I’m telling you hear to avoid great pain and loss of money if you’ve never created a website before. You need to be very, very, very careful of these programmers because they will tend to go as far as lying to you to hook you and once your deep in it, they will milk you. I have sued two programmers and have a case against one pending in the Supreme Court of New York. So far I have been winning. Even the Judge acknowledged in a recent order that one of the programmers I’m suing quote: “acted improperly by representing to prospective clients that it had performed complete and satisfactory work on Mehmet’s computer program” unquote.

    So if you are thinking about starting your own website, be very, very, very careful on who you hire to code your site or even a graphic designer. My experience is that the majority of them are not good. Meaning, their work is not “original”. They will give you bad art work or copy someone elses work from the Internet. Be on guard when a graphic designer asks you to point out some websites that contain a similar design you want to create. But there are great designers. Very rare.

    As for the downfall of the .com era, you should rent or buy this video called “startup.com”. It is a perfect example why the .com era fell and shows you what NOT to do. These guys were so stupid that they actually let a major competitor of theirs into their offices and they were disclosing company secretes the him. When the competitor launched their similar program first, everyone in the office was saddened. They lost almost $50 Million of investors money. They spent millions on programmers to develop an application that didn’t work properly. This type of downfall was mimicked within other Internet companies. Young and foolish.

    My one advise to anyone who wants to start a business. Develop the application first. Test it on the market. If you get a good return then go out and spend the millions on office space, employees and equipement because your entire business will depend on that application.

    Remember, Google started out in a “garage” with two college students and eBay was created in a bedroom after work . Now those companies are Billion dollar companies. A good idea will always succeed with the right plan.

    … and thus, some reasons for disagreeing with the oringinal writer of this blog. Experience is extremely valuable and goes beyond book knowledge. If you don’t have it then work for someone who does until you do. You’ll be successful much faster with less pain.

  7. Charles Ying Says:

    Awesome insights, Joyce. FYI, FilmLoop was 8 months from concept to beta launch and I’ll tell ya, it was quite a ride.

    My bit of advice is not really about hiring product managers, or finding the best people, but something I picked up from John Carmack, who once said:

    “It’s not the one brilliant decision, it’s the 500 smart decisions that really make things good. It’s more a matter of being able to keep making smart decisions. Making one brilliant decision and a whole bunch of mediocre ones isn’t as good as making a whole bunch of generally smart decisions throughout the whole process. And there’s so many of them that have to be made.”

    I’ve experienced this first hand on every project I’ve ever worked on, and the one thing I try to watch out most for.

    Good stuff. Hope to meet you and the Renkoo folks sometime at 106miles.

  8. DML Says:

    You forgot something :

    Developpers can start their own company, this is where it’s so cheap and easy.

    Do you really think that you need 24 servers to start your own small business ? Ahaha you’re funny, you should read again what you wrote… Get down, open PHPEdit and start your dev.

  9. Chris Pierjkowski Says:

    Lets all look at what is beneath this post. Why would this author be so uprooted by others proclaiming that Web 2.0 companies are easy to start? As most of us know, Joyce is one of the leaders of Renkoo, a Web 2.0 company. Why would she oppose more Web 2.0 companies popping up? Simple because they would bring competition!

    Lets think back to high school economics. Supply and Demand. More Web 2.0 companies = more supply, while the buy-out demand stays the same. This means that the price people are willing to pay for Renkoo drops hard.

    Now why do I bring up a buy-out? Surely Renkoo can make money to sustain itself, they did get $3 million in funding after all, right? Wrong! Currently, there are 3 main exit strategies for the entrepreneur and their VC amigos: 1) Going Public 2) Creating a steady Revenue stream and sustaining themselves or 3) A Buyout.

    Renkoo will not create a steady revenue stream. They simply cant, as they have no (apparent) business model. Even if they could manage a revenue flow, their VC amigos would demand a large return on their $3million and simple ad revenue would not come close to cutting it. Renkoo will not go public, since the costs of going public are increasingly high (see Sarbaines-Oxley and Securities and Exchange Act of 1933). Finally the only option left is to be bought out. Furthermore, virtually every major web success story has been bought out. Think Upcoming, Delicious, and even Facebook (just recently have been offered to be bought out). The fact of the matter is that a Web 2.0 company cannot make that jump into Google-dom without the backing of a huge corporation.

    Although Joyces post brings up a lot of good points, I read it as a scare campaign. She writes it is a way to discourage competition, when competition is what drives the best products to be the best. Competition should be embraced not scared away by posts such as this one.

  10. Duc Says:

    I am a beta tester for renkoo right now. I still see like 2 features on the renkoo site. Does it really take 14 months to churn out a few scripts? What are all those engineers with “cutting edge skills” that you recruited doing???

  11. Troutgirl Says:

    Apparently they were not Sun boxes… they were x86 running Solaris, which we paid for.

  12. Mag Says:

    Maybe someone can start a web2.0 company depending on outsourcing to some other country with less cost skills.


  13. Web 2.0 is a lie. Ok maybe not- but seriously all business is business, the cost in 99- versus the cost now? Okay so hosting bandwidth etc is cheaper- but really the cost is the human cost. Your biggest cost is almost always going to be people.

    I have enough knots in my shoulders worrying about closing the sales and making salaries every month- this whole concept of web 2.0 startup is bull. its just business.

    -Kevin


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: