Français

Day 13: Analytics, Exploits and Beer

May 1, 2013 at 12:32 pm

Alistair Croll popped in to Notman to lead an insightful workshop with the companies. Alistair co-authored the recently released Lean Analytics – one of the best reads out there for startups. Seriously. Everybody needs to read this book.

lean-cover

Most startups don’t know what they’ll be when they grow up. The majority need to figure it out as they go, pivoting and iterating based on what they learn along the way. Without metrics and analytics, it’s not possible to know which way to go.

We say we’re living in an information economy. However, an economy is defined by what is scarce, and information is certainly not scarce in today’s world. We’d be more accurate to call this an attention economy. The biggest challenge for a startup is thus to create a product or service that commands attention.

A smart approach to figure out what you should be building is to build a product in order to figure out the product you need to build. Anticipate your users’ needs and go from there. Would a focus group have led Apple to create the iPad if they had been asked what they wanted?

A great example of using metrics and analytics to experiment is how Airbnb came to offer photography services. They came up with a hypothesis – that better quality photos would increase bookings – and they tested by sending photographers to the spaces listed and measuring the impact on bookings of these spaces. We don’t know what incited this hypothesis, Alistair ventures that it could have been exploratory — they saw a correlation in their data between photos taken with good cameras and rentals — or maybe it was a qualitative hypothesis where someone at Airbnb thought that nice pictures are more attractive and perhaps nice photos would boost bookings. Whatever inspired the hypothesis, the key is that Airbnb picked a metric to measure that would change the way they behaved: if better quality photos did in fact lead to increased bookings, then they would start offering the services of professional photographers. The take-home message is that if a metric won’t change the way you behave, then it is a bad metric.

Alistair advices startups to focus your analyses on the stuff you don’t know you don’t know. This means exploring, asking questions, observing. A team at Procter and Gamble wanted to disrupt the soap industry, a big feat given how stable this industry had been for a very long time. So the researchers brought in a group of seniors, spilled stuff on the floor and stared the seniors down until they felt compelled to clean (the story, as told by Alistair). The team observed that instead of getting out a mop to clean, the seniors were using paper towels. This led to the “ah-ha moment” where the team realized they didn’t need a better soap but rather a new gadget, and the Swiffer was born. Swiffer is a 500-million-dollar-a-year product.

Deviating from the topics of Lean Analytics, Alistair gave the companies an unexpected piece of advice: be subversive. This advice popped up many times throughout his talk, which left no doubts about how important this message is. Alistair believes that to become successful, a startup needs to have a subversive attitude – think evil – ask yourself: “how would I do this if I were a hacker?”. Find the hack that helps you change the future. This is the nature of trying to disrupt a market. Only idiots play by the rules.

There are many things your startup can be doing (that aren’t illegal) to get an edge. Airbnb sent people who posted accommodations on Craigslist a message inviting them to post on Airbnb. The guy who created DuPropio drove around and called every house that had a black and orange “for sale” sign, and offered to list them on his homepage for free.

If you have a two-sided marketplace, get the side that makes money involved first. Buy them in if you need to — this strategy can cost less than marketing. To start, Uber paid every limo driver in San Francisco to sit around – and be available when a Uber user called. Create an artificial marketplace, and if you can create demand, you’re golden.

By the way, if you can’t make it work when you’re paying for the supply, then your business model won’t work. On a similar note, there’s nothing wrong with buying users if each user is going to bring on more users.

The harsh truth is that anyone can code; instead of focusing on building, your startup has to find the one idea that’s going to crush the competition. Alistair advices that until you find that killer idea, come up with 5 hacks everyday.

Oh… Apparently brainstorming works well when beer is involved!