Thursday, December 13, 2007

Improving Your Internet Application Success

Many VCs and Entrepreneurs struggle with the problem of predicting when to invest money and time into a new idea and when to pass. Some just guess, others invest in people and passion. I decided to try and see if we can analyze and find some success predictors.

A method often used by investors include a mental simulation that assumes the investor is the target audience (even if it is a product for kids) which will usually end up with one conclusion “I would not use it, why would anyone else”. When in doubt, they will say let’s wait and see if it gets some traction. This is a good test but if the product gets traction without marketing budget, you may not need the investment, or the valuation would be different.

Entrepreneurs often get excited an in love with an idea and forget the basics such as the business model. As a frequent visitor to the NY Tech Meetup forum I can tell you that the forum organizer banned the Business Model question simply because, many entrepreneurs had no idea what it is.

To make that process a bit simpler I have decided to do a summary of my non-scientific observations of some internet product features that can be used as predictors for success.
As a reference I chose to look generally on some successes such as ICQ, Skype, MySpace / Facebook, Google, Twitter, Wikipedia and try to ask the question “what traits do they all have in common?”

I came up with the following list of “success” factors (traits or features) that seem to repeat themselves on the more popular applications and websites. I have also added some of my own experience to the mixture.

Let's start with those:

  • Viral or Social Characteristics
  • Simple & Reliable
  • No installation
  • User Participation (UGC)
  • “Usefulness
  • Business Model
  • Marketing and Budget
  • Investment Requirement

Viral or Social Characteristics

If a product is more useful to you when your friends have it, it means that it poses the viral trait. For instance what can you do with ICQ or Skype by yourself? Maybe talk to strangers. Yet, when your friends have it, you are really thriving.

If we look at my MySpace, the need to show people your new psychedelic page background drives you to ask others to join in.

If you have designed your product with a viral approach, good, if not, try to ride on top of an existing Social Network that has strong communication tools, so that it would very simple for a user to tell a friend, see what I found.

LinkedIn for instance has Poor communication tools, they adopted the “Don’t e-mail me, I will call you” approach whereas Facebook did the opposite with features such as Poke, Status, and Simple internal mail.

The use of OpenSocial API (led by Google) will enable new application to get closer to the Holy Grail of interoperability and Viral Distribution with the mantra of “write once run many” (on many Social Networks that is).

Simple & Reliable

Not much to say here, the product should Simply Work (Like Skype) vs. for instance MSN Messenger or other SIP phones that at least initially suffered from all kind of NAT (networking and fire wall) issues. I would say that new products should be designed with security in mind and work behind corporate firewalls. It seems that people now work from home, and “shop” from the office.

The GUI must be as simple as possible, like twitter; it should work on every OS, and on any browser. I know it is not always simple to do, yet it is a good design consideration, that can be achieved over time.

No installation

Every installation that a user is required to do is a friction point, even Skype in my opinion may have been more popular if it did not require setup for the first second.

My recommendation is to use Flash when possible (if a web page browser based technology is not sufficient). There are new similar or Flash like next-generation environment but it is better to stick to those already deployed with end users. The most popular applications such as G-mail, Facebook, MySpace, don’t require installation.

Sometimes the audience you need for Business Development is different than the one really using your product. Skype Developer Program manager said in a developer session in NY “Make sure your application can run on a Mac” so more people in skype will know about you.

User Participation (UGC)

It makes sense, that applications, which let users share private or public information, get more attention. First the users who placed the information keep getting back to update, or share more. Second it provides for an incentive to tell others about it.

In a way it feeds our basic curiosity to see what other people expose and our need to be exposed. In other words the basic human Voyeurism and Exhibitionism. Good examples for that are Q&A sites, Facebook and MySpace.

It turns out that UGC it is also good for SEO (Search Engine Optimization), which dramatically contributes to the economics of the site.

SEO In general is something that needs to start from day one, from the point when you register the domain name, or buy one that has traffic, and on, on every page that comes up on the website.


Usefulness or Utility is hard to predict, what useful for me may not be useful for others and vise versa. I would never think that emoticons will be useful in any way, yet almost two Million users of Zlango think otherwise. When Skype came out people might say why I need another Messenger/Yahoo/AIM/ICQ/… well apparently me2 is not always useless and if it always works with better audio quality, it is useful.

A product is also useful if fills a personal social need like, promoting global knowledge, by writing articles in Wikipedia or a need for personal promotion by writing a
personally related factual page in Wikipedia.

Fun is another form of utility if people can derive fun of a product it is useful for that purpose.

So how do you predict if a product or a webpage is useful, I guess by asking as many people as you can, or finding a specific need that the product covers. One way to try and predict is by using the
Kawasaki map of Uniqueness vs. Value.

Business Model

There are many business models that one can come up with, yet the most successful are Ad Revenue which is based on traffic or Eyeballs, Recurring Subscription fees as done with Linked-In, and WOW (World of Warcraft) and so on. The last is selling non tangible replicable products such as Music, Videos and things like Software Download. A type product that makes you money while you sleep.

Commission based models that are driven by paid clicks (and not by organic traffic) are getting hard to support with the climbing prices of popular keywords. In some popularlucrative domains (sex for instance) a click can cost more than $10, imagine how much money has to be made per transaction to get the net revenue balance the cost of the click divided by the conversion rate. The party receiving the gross revenue is the one that need to support the ad, not the party receiving the commission, and this is a hard situation to create.


Never leave marketing to pure luck. The more successful companies (leaving out those that got lucky or had other traits working for them) started with nice marketing budget. They started with their SEO activities, such as creating content, creating back links, grass root buzz creating activities, Ads, Videos, TV commercials, radio commercials, free “minutes”, and more.
A good young marketing person is an asset worth having on board. Pure luck is simply to hard to manufacture.

Having said that a startup need to make sure it allocates sufficient funding for marketing purpose.

Investment Requirements

A product or a web site is not built in one day, it takes time and time needs funding. As an entrepreneur I would recommend that the first Angel or VC round will take in account R&D and Marketing. From an entrepreneur point of view if you finish the product and you can not market it to create traction; you will not be able to raise the next round. Investor place no value on Internet technology those days, as sophisticated as it might be.

An investor on the other hand would argue that it is better to raise money for marketing after the product is proven, or has gone few improvement phases. This obviously seems to minimize the risk, and drives the company to make improvements trying to make the product get traction without marketing.

I say, raise the money up-front. Spend on marketing only after you have done some improvement phases. Don’t get caught with a product everyone says is good, but has no traction to prove them right.

I will conclude by stating that not every application or website encompasses all the ingredients from day one, the more it has from the get go the better its chances are to succeed.
You must keep adding the ingredients and funding in order to move from the initial success to the ultimate exit opportunity. If you are stuck without one or the other, start writing Blogs :).

Alon Cohen