Train Thoughts

Web based businesses have the ability to be incredibly efficient.  For ad supported businesses or new types of services, there are two phases.  Phase one is getting the adoption and distribution.  Phase two is monetizing the distribution by getting the business model fit to the product.

During phase one, I think it’s important to be as lean and as efficient as possible.  At this stage, you are spending money, but not making any.  Web based businesses lend themselves nicely to achieving large distribution with few people and low cost (assuming you don’t have massive infrastructure needs like a search engine, storage company, or someone that streams a ton).  Do this as leanly as you can.  First, the longer you can extend, the more scale you can achieve and more product iterations can be completed.  Second, you’ll need time and money to figure out the business model once you have scale.

If you spend too quickly, which for many web based businesses is highly correlated to hiring too much, you’re likely to find yourself in a very difficult position.  Too many engineers before you know exactly what to build, and too many sales or business development people before you know how you’ll monetize are the grim reaper of businesses.

My recommendation is to build incrementally, hire as few people as possible.  But. Once you see the scale kick in, move to figuring out the business model. Hire one person.  Get the proof points and figure out how to scale.  Then.  Once the distribution is kicking and the business model is working.  Scale it as quickly as possible.  But until then.  Be as lean as you can.

Here are two quick examples.  Let’s say your are building a vertical content site for women’s health issues.  The plan is to build an audience and then monetize it with brand ads.  Phase one is to build the audience to a size where you can attract a brand advertiser.  Say that’s 2 million uniques per month.  To get to this level of uniques, it’s going to take some time.  Watch at your cash, analyze your traffic growth and make sure you are within one quarter of having 2 million uniques before hiring a sales person.  Then.  Hire one sales person.  Get money coming in the door before hiring account managers, ad ops etc.  Only hire the support people once the revenue is coming in the door.

Example two is say you have a new service like Twitter.  Maybe it will be add supported, maybe subscription, maybe there’s a paid search model in there, maybe sponsored tweets.  Who knows exactly what the model will be.  But.  You do know that you have a media darling.  People love it.  And.  You’re scaling your audience rapidly.  What you don’t want to do is go hire a large sales organization when you are not sure how the ultimate business model will fit your product or service.  Hire one person that is responsible for designing very small scale monetization tests.  Run tests.  Measure the results.  Find out what can scale and then build from there.  If you hire big too soon, you are most likely to spend a lot of money and not have much to show for it.