Talk notes: Onlab [Growth] Hackers Conference 2013

Growth hacking is more, faster. More, as in more of what you already have. My notes from Ilya Lichtenstein‘s talk at Onlab [Growth] Hackers Conference 2013.

“Achieving hyper growth – How the world’s top marketing team scale campaigns fast”

About Ilya

  • 2006: SEO – committed a lot of resources but wasn’t so successful. SEO is a very competitive.
  • 2007: Facebook Ads – achieved big success via automated ad generation
  • 2010: Performance marketing and media buys – risky, no way of knowing what will work
  • 2011: Co-fonder of MixRank – see what ads are being tested in the market, what other companies are using, etc. Advertisers can decrease risk of the media buy process by getting an indication of what’s working before spending the money.

The growth mindset

  • Growth = leverage
  • Seek multiplicative effects
  • Do anything that gives you disproportionate outcomes for the smallest effort
  • Parato rule applies: 80% of the output will come from 20% of the input. Actually, marketers report that 93% of campaigns fail to deliver results
  • Minimize opportunity costs: Most campaigns will fail anyway, so think about what you’re giving up in order to do it
  • Ask yourself, does it scale?
  • Anyone can be a growth hacker: Sales people, marketers, etc. need to think about activities that are scalable

What is the right time to grow?

  • When you have initial product/market fit
  • When you understand key metrics
  • When the numbers make sense
  • When organic demand matches your roadmap. If people say they won’t buy your product because feature A is lacking, and it’s on your roadmap, that’s good timing. If it’s not, then it’s not the right time to bring in a lot of people
  • Many teams make the mistake of bringing too many people, too quickly. Their product is not mature enough to convert.
  • You cannot create demand. Solve a problem that’s already there.
  • One of the biggest skills of a growth hacker is to be able to walk away.
  • Growth hacking accelerates existing processes. It’s more, faster.

The growth process – “where are the people who will make us the most money?”

  1. Qualify your best users
    • More than engagement: It’s not always the ones that use the product the most, the most engaged might only use the free parts of your service. Someone who only comes in once a month but doesn’t cancel their subscription might be more valuable.
    • Look at revenue potential
    • Get first-party data (ex. Google Analytics) and  append with third-party data (ex. social network profiles, avg revenues based on postal codes)
    • Track cross-channel activity: have they Liked our page? how often do they read our e-mails? have they opened the newsletter from different IP addresses?
    • Cohort analysis: Segment based on traffic source, subscription length etc. The goal is to find patterns, and try to predict if they will signup/cancel. There’s not a lot of good software, just massive spreadsheets.
  2. Identify where they go online
    • Psychographics > demographics
    • Talk with your users: In person, surveys, e-mails
    • Users want to talk to you, engage with you, be heard. Leverage that!
    • Find adjacent markets
    • What are your users talking about? Get social profiles of your users, grab 50 most recent tweets, find the most common keywords
    • Where do they spend their time? Who is marketing to them now? What are the messages that is coming to them?
    • Graph search for power users – Facebook is underused! People who like us – what other pages do they like? Where do they work, what do they live, what other interests?
  3. Access distribution channels
    • Not just viral
    • Fail fast
    • It’s all about traffic quality. Correlated to the users that will make you the most money.
    • Optimize for what marketers call “Share of voice” – What share do you have of the voices that your users hear? 
    • Arbitrage low > high quality: How can you target the same people on a cheaper channel?
    • Leverage existing audiences
  4. Automate inefficient processes
    • What will help key metrics the most? For Mixrank, it’s revenue and revenue potential
    • Do NOT track everything! Too many data points cause paralysis by analysis
    • Remember opportunity costs
    • Not all solutions are technology solutions: ex. Outsourcing to Odesk is a good way to not depend on dev resources, which is a scare resource at a small company
    • Involve everyone on the team and foster a growth mindset: When everyone on your team is a growth hacker, that’s when you can scale. 
    • Control your growth rate