Could you grow your users at 5% for 15 months? Or at 7% for 12 months? Or at 10% for 9 months? That’s how long it would take to double your user base at those growth rates.

Though there is some science to growth developed by the technology-fluent marketer, aka growth hacker, growth remains a confounding art because there isn’t one feature - like implementing FB Connect or promoting Tweets or having a big red button on a home page. Rather, it is the sum product of a hundred little things.

Growth is iteratively engendered; growth is tested, discovered and explored. Growth means multivariate testing combinations of product, marketing copy and design. Growth is using behavioral psychology to cajole users into doing what we ask. Worst of all, growth is a playbook whose plays change with time.

On Friday, there is a Growth Hackers Conference in San Francisco with some great speakers presenting their experiences from Facebook, LinkedIn, Square and Google.

I suspect the unifying theme throughout the conference will be leveraging multivariate testing software to develop continuous small short term wins into long term compounding growth.

At least, that’s how we grew at Google - testing different ad targeting models. Building a library of Google.com UI experiments. If there’s one constant for growth, one secret, it’s scalable experimentation and testing frameworks.

Up and to the right, up and to the right, up and to the right is the mantra we all chant. Startups have to grow users, engagement and revenue to survive. Consistent small wins drive the long term growth we all seek.