A General Approach To Web Startups

It used to be the case that if you wanted to start a business, you’d spend a lot of time writing and researching a detailed business plan, in-depth financials, and doing all sorts of analyses of risks and possible scenarios. I naively tried starting an education business like this in 2005 and wasted a whole lot of time in the process.

It’s self-evident that the higher your cost of failing, the less times you can afford to fail. It used to be the case that starting a business was a costly enterprise – you’d have to build a factory, invest in machinery, and embark on a costly sales process. There was a lot of upfront capital expenditure required before the first product was purchased. In that scenario, if you failed it was unlikely that you’d have the capital to restart and give it a second try. Hence, minimizing your chance of failure was key.

Today, startup costs are minimal and hence the cost of failure is way down. This means that the value of the time required to write and research extensive business plans is higher than the cost of executing on a loosely formed idea and iterating on it a few times as you learn what isn’t working.

This build-learn-repeat approach is old news for web startups, but one that is a key starting point for any new web business. Steve Blank’s “Four Steps to the Epiphany” is an oft-quoted treatise for this approach (note, Steve is an investor in my current company) and a great starting point.

A Word on User Acquisition

I vacillated between titling this post “user acquisition” or “customer acquisition” – the latter has some business model implications so I decided to go with the former, but alas in these economic times we’ll need to revisit the latter very soon.

So, you have a cool website/app now how do you get people to actually use it? If we boil it down, there are three main ways in which users find new websites:

1. They search for it
2. They find out about it through a friend
3. They find out about it in the media

All three channels can be either free or paid channels, they can also be online or offline channels. This lends itself to a neat little matrix. I made the size of the bubbles to be an estimate of how “big” each channel is. I think that the best measure of “big” is attention and impact, so if anyone has any input on how big the bubbles should be (and the data driving it), I’m looking to make this more accurate.

We can do a deep dive on each of these user acquisition channels as well as the tools that are relevant for each. That said, the first two questions every company has to answer for itself are (1) Which channels are most appropriate for my business? and (2) How do I minimize the acquisition cost for my channels?

This also raises another question about return on investment (ROI) and how to make sure that you keep that positive based on what you know (or estimate) the lifetime value (LTV) of a user to be. Sounds simple, but to make good decisions you need to ensure you can segment LTV by channel (and optimally to an even more granular level), which requires robust metrics and instrumentation. Further, you may have little to no data on LTV, which makes it ever more difficult to produce adequate estimates.