Here’s a summary:

 

Your objective is to hit or exceed your weekly growth rate goals by breaking it up into in weekly sprints

1. Start with your Top Level KPI. Ex: Sales (if you’re an ecommerce business) and break it down into its drivers.

Sales = # of customers x Average order value (AOV)

(while you can play with AOV by changing pricing, # of items added to cart etc. for this example, we’ll focus on # of customers)

Sales = (# of visitors x Overall Conversion Rate) x AOV

(and so on and so forth)

The left side if your Top Level KPI, and the right side are your drivers or levers

Some levers are easier to move than others. Your next step is to pick the levers that has the most “play” or can be moved the most distance with least effort, i.e the lowest hanging fruits.

   2. Pick levers to work on right now:

For each of the levers note down the baseline (the current value), min possible value and max possible value (ask other founders/mentors/experts for the current values they’re seeing for these levers). Ex: if your ecommerce conversion rate is 1%, industry average is ~3% and the industry leaders are around 11-13%. So there’s scope for you to improve here. Pick 2-4 levers you want to work on this week.

(note, as a particular lever gets closer to its current max, it gets harder to move, i.e, it’s no longer the lowest hanging fruit. When that happens, repeat this process to find your next lowest hanging fruit)

   3. Brainstorm experiments to move these levers. i.e what can you do to improve your conversion rate from 1% to 11%?. Ex: if you assume the conversion rate is because of lack of trust, what experiments can you run to prove or disprove this assumption? Add testimonials? Money-back guarantee? Dump all experiments in a backlog and prioritize

   4. You probably have hundreds of “ideas” you want to test, but which ones are the best use of your time?  Prioritize the experiments in your backlog by scoring each experiment one 1-10 for their Impact & Effort. You want to run experiments that have the highest impact for the least effort.(we talk more about how to pick big experiments over small ones in Testing 101 (link)).

  5. Re-examine each experiment and the assumption you’re trying to prove or disprove with the experiment and come up with a Minimum Viable Test - what’s the least effort you can put in to quickly determine whether you’re directionally correct. Less effort = more experiments. Since most of your experiments will fail, you will need the momentum to keep you from being depressed.

  6. Run the experiments (more on this in testing 101)

  7. When it reaches statistical significance, for every experiment, analyze how much the experiment moved your lever by and why. What worked, what didn’t work, why/why not? Was your assumption proved or disproved? What were your lessons learnt about customer/product/campaign etc.?

  8. Update your KPI-dashboard with how much each lever moved (because of an experiment you ran)

  9. Weekly Growth Meeting to zoom out, share lessons learnt, plan for next set of experiments to run the following week

 

A similar process is shown below