Grow your startup faster with these 4 steps

Lean Analytics is a part of the Lean Startup school of thought. It deals with how a startup can use data to become better, faster. Here, Alistair Croll, the author Lean Analytics book, explains how startups can avoid mistakes early on and move faster using Lean Analytics. At 20 seconds, he clearly explains why startups must focus on improving only one thing at a time.

In Lean Analytics, startups generally keep iterating through the following cycle.

Lean Analytics helps startups use data to grow better, faster

How you can apply Lean Analytics to your startup

Basically, you are trying to pick one number which is important to your business and optimize it without changing anything else.

Step 1. Pick a number to improve

It may be number of visitors, conversion rate, time on site, etc. In other words, it should be a metric (number) which is critical to your business growth. If you are not sure which metrics matter, look for companies with similar business models. Find out which metrics they track. But pick only one metric at a time. So that you can optimize it. Also, define the value of this metric beyond which your efforts are considered a success.

Step 2. Take an educated guess

Take an educated guess as to what might be the reason for your metric not having a higher value.

If you have no data, you can run a survey, refer to best practices or study how market peers or companies with similar business models are doing it. If they are doing something really well, try it out and see if it works for you. E.g, landing pages, marketing campaign, customer on boarding, etc

If you have data, then find out what is different about the people who do what you want. You will need to segment them based on their characteristics like background, location, position in company, what type of company do they work for, age group, gender, time of activity, etc. Also, find out what is common among the people who do what you want. Are they college students? Do they all come from same city? Do they have similar interests ?

Step 3. Create an experiment to test your guess

You need to ask 3 questions – Whose behavior are you trying change? What do you want them to do? Why should they do it? You are trying to trigger a bunch of people to take action. You think this will improve the value of your metric. General experiments to get results are Cohort Analysis, A/B testing and Multivariate testing. If these terms are too nerdy for you, simply change something (e.g UI, landing page, etc ) for the target audience and see how they respond. Record it and see if it is more than expected.

Step 4. Take action based on result

If the measurement cleared the expected value, it was a success. That’s great, move to the next metric and repeat the cycle. If it failed, it means you are going in the wrong direction. Go back to step 2 and form a new hypothesis(educated guess). If the measurement didn’t clear the expected value but is close, then it succeeded only to a small extent. It means you are headed in the right direction but your actions were not effective enough. Go back to step 3 to create a new experiment to nail it this time.

For more detailed explanation about this approach (with case studies), please read this wonderful article by Avinash Kaushik. Read it at leisure, it’s worth it.

What do you think? Do share your thoughts as comments.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>