Split-Test Experiments

– – –

  • Once you have a decent flow of incoming traffic/customers, you can start setting up “scientific” experiments with these people
    • A true split-test experiment: offer two different versions of a product, landing page, ad, etc. to customers as the same time and see which one performs best
  • Start with a clear hypothesis
    • Develop an experiment (set time frame)
    • Make sure to test only one variable
    • Test your predictions empirically
    • “If you cannot fail, you cannot learn.” – Eric Ries, The Lean Startup
  • For example, split-test a call to action button
    • Allow 50% of customers to see “Purchase” and 50% to see “Buy Now”– which one do you think will sell more product?
      • Test and see!
    • Just an example – there are endless opportunities to set up experiments
      • Verbiage used
      • Prices
      • Landing pages
      • Marketing material

– – –

Next Lesson: Create a Growth Hacking Machine

Access the Course Dashboard Here