The following is a guest post by Frank Mashraqi a scalability expert. Meet Frank at Startonomics, and hear him speak during his session: Scalability for Startups: How to Grow Up without Blowing Up.

Startups either completely ignore scalability, a non functional requirement, or put too much emphasis on it, mostly depending on whether the founders have technical backgrounds or not. Unfortunately, neither approach is right.

Non-functional requirements are specific criterion commonly used to evaluate the operation of a system.
Web 2.0 has changed the rules of the game. It gave us, among other things, a venue to be social. With social web came social graphs and open networks. Today it is hard to find a social network or a popular application that doesn’t offer access to its platform one way or another. What this means for startups is that there are more opportunities and more data available today than ever before.

Chris Saad, chairperson of DataPortability project refers to Web 3.0 as the Personal Web and I agree with Chris. His vision is of a Web where instead of people adopting applications, Applications adopt people; a world where people are free to take their social graph, media and everything else from one site to another.

The opportunities of Web 2.0 and upcoming Web 3.0 mean more openness, more virality, more data and more interactions around which startups will have to be build. Scalability will end up becoming more important and a pre-requisite to functionality instead of being considered unrelated to functionality.

For data and media intensive startups, ignoring scalability is simply not an option. Unscalable sites risk adoption by frustrating users. Same is true for startups that are looking to create presence on an existing massive platform such as Facebook or MySpace, not having the right foundation can be very costly as these platforms have the potential to send an influx of users and data.

A company that was literally turning away its users after reaching a pre-defined high watermark because it simply couldn’t scale to accept more users once approached me to help. That must have hurt!
I’d like to share some of what I hear from startups as they address scalability challenges.

“Let’s just go with a working prototype for now. We can change it when we grow big.”

What’s happening here is that although this startup is decreasing their time to market, they are taking on a technical debt, the cost of which is unknown. For a data intensive startup this is a recipe for disaster. I know of several startups where growth actually became a painful experience. Some of these startups successfully went through the colonoscopy like experience of strengthening their foundation during high growth period at a mind-blowing cost. Others were not so fortunate and ended up closing their doors.

“Potential partners really like our product. The problem is they aren’t willing to partner with us because of fears that our infrastructure cannot handle integrating with them.”

Believe it or not, I hear this story more often than I’d like to. Startups who are faced with this situation are very reluctant to take any steps I recommend. Most of the time these startups have spent most of their raised money getting to this particular point and some big partnerships must be made before they can raise the next round to address the foundational issues that are causing them not to be able to scale. It kind of becomes like a chicken before the egg problem.

“Our VC/advisor told us that we’re worrying about scalability too early. He/She doesn’t want us to blow our scarce resources on preparing for success.”

If you were to ask me what I think of this advice, my answer would be, “it depends.” For some startups, this advice is perfectly right. For others, it’s not.

The word “scalability” means different things to different people.
If your startup is data or media intensive, for instance, you are going to need scalability. It will be a much better investment of your resources to build the foundation for it as early as possible. Pennies now compared to ten dollars later, I often tell people.

Personally, I have been approached numerous times by startups that had experienced growth to the point that they just couldn’t afford to strengthen their foundation anymore. Cost was usually not the factor, rather the primary reason would always be to the tune of “we just can’t afford downtime now” or “we can’t maintain two development trees” or “is there any way until we can raise the next round or make an exit.”

“Premature optimization is the root of all evil.”

I completely believe in this quote that’s often thrown at me by developers. Nothing makes my blood boil more than this quote being used out of context. The problem here is that even among technical folks confusion exists regarding how optimization and scalability are two vastly different things. It’s like comparing the foundation of a house with how well the paint has been applied on the walls. In the early stages, build the right foundation so your startup can be scalable but don’t sweat on making an algorithm run x percent faster.

“We’re planning to host on a cloud. That should solve our issues.”

One of my favorite questions to ask at cloud computing events is that “Can your cloud fix Twitter?”
The answer is that it probably can’t unless once again, your foundation is built right and is scalable. A badly designed architecture simply won’t scale.

“We plan to throw hardware at the problem.”

Throwing hardware is a short run win, expensive and it will only take you so far. Eventually, you’ll have to address the core issues that are making your startup not scale.

Find answers to your most pressing scalability questions at my Startup Scalability Strategies session at Startonomics.

At the session, you’ll be able to learn about:
• Keeping score: What to measure?
• Discerning the difference: What to focus on?
• Making your choice: What do you want?
o Horizontal or Vertical? Up or Out?
o Synchronous or Asynchronous?
• Choosing your architecture: How to partition?
• Serving requests: How to deliver?
• Walking the line: How to balance time to market with the need to scale?
• Building your team: Who to hire and what skills to invest in?
• Going down the wrong path: How to blow up?
• Thinking ahead: What about the future?
• Reaching new heights: What to do now?
• Offloading scalability: Is it for me?

After the session, make sure to grab me for any questions that you may have. Hope to see you there!

Heard enough yet? Join us at Startonomics