by • August 30, 2010 • Best of, Lean startup meta, Listening, Tools & frameworksComments (1)152

The Risk Validation Pyramid

The short version: You should validate business risks in the quickest way possible. The faster you can resolve the current mystery, the more likely you are to see the next one before dying.

"At least he'll always have his TechCrunch coverage."

The long version: Startups are a pile of risks. Some will kill you now and others lurk like spiders beneath the toilet seat. If you claim your business has no risks, you’d better be telling me that on your waterproof phone while leaping off a diving board into a hot tub on your private jet.

Your company’s ultimate risk is “Can I sell this to lots of people for more than it costs me?” The lean startup movement is basically an attempt to make that one big question more granular: first ensure your future customers actually have the problem you think they have, then find out if their budget allows them to spend the money you want, and then you see if a few crazy outliers will pay you.

Every risk needs to be validated. However, not every risk needs to be validated equally. After all, the only way you can truly prove your beliefs is to take someone’s money, and that takes time. The crucial innovation of Customer Development is the idea that before asking someone to give you money, you first ask them for their opinion. It turns out people love doing this, because it’s the only time in life that you get to bitch about what annoys you to someone who’s so excited to hear it that he writes down every word and buys you coffee.

May I offer you a gift card in exchange for this amazing conversation?

Interviewing someone is easier than building and selling a product. That’s a step in the right direction. Komisar & Mullins have added the idea of an analog, which is even easier — you jumpstart your thinking by stealing the lessons learned from comparable companies. Zynga and IMVU have popularized the idea of fake buttons and some people like to test demand through landing pages & search ads for non-existent products. That gives us 4 tiers of validation. The fifth is to simply watch people. Observation is the easiest (and least reliable) and is helpful when your competitors have a forum full of pissed off customers or when there’s an active mailing list or conference circuit for your industry.

The 5 levels of risk validation

  1. Compare to an analog
  2. Observe what your customers already do
  3. Interview customers or experts
  4. Fake it with a dummy product, dead-end button, or “demo”
  5. Build the actual product

Now, do bear in mind that people lie. They lie to themselves, they lie to their lawfully wedded partners, and they will lie to you. They don’t necessarily intend to; they may just want to make you happy. Regardless, they’re unreliable. But that’s okay because it’s way faster than fully proving everything and infinitely safer than considering nothing.

One of the places the 4 Steps has really aged is in its chronological scope. Ten years ago, software took a long time to build, which gave you a lot of time to hang out with customers and sanity-check your beliefs. These days, great startups commonly go from idea to launch in 3 months. For most web apps, MVP launch time is the hard ceiling for any sort of upfront risk validation or customer development.

The validation pyramid exists to give you a range of available speeds and certainties. You’re looking for the first good-enough evidence of validity, where “good-enough” is based on how dangerous a false positive would be. For example, accepting a loose analog and ending up with the wrong value proposition in your sales deck is only going to set you back a couple meetings. On the other hand, you’d want to spend a lot of time running experiments and talking to the involved parties before launching a major channel partnership (or even a big feature) because the move could be irreversible.

Every startup needs to be fast and you can never be completely certain about your decisions. Still that doesn’t mean you can skip them. You need the appropriate balance between speed and certainty that will allow you to quickly answer the questions that matter most. So what’s about to kill you next?

PS: Did you know the startup toolkit now has a risk dashboard (you need to be logged in) for keeping track of these thrilling dangers? It’s quite new and I’d love to hear whether or not it’s useful to you.

Comments are at Hacker News

Related Posts

One Response to The Risk Validation Pyramid

  1. [...] You should start with the riskiest assumptions that you can test and try to make them fail. Here is why you should start at the bottom of the risk validation pyramid. [...]