Running Successful Growth Experiments (w/ Andres Glusman, DoWhatWorks)

Author's avatar Metrics & Chill Podcast UPDATED Feb 20, 2024 PUBLISHED Apr 5, 2023 4 minutes read

Table of contents

    Peter Caputa

    To see what Databox can do for you, including how it helps you track and visualize your performance data in real-time, check out our home page. Click here.

    Andres Glusman (CEO, DoWhatWorks) shares a simple framework to help you run more successful growth experiments, and save time and money.

    How They Move The Needle

    Andres Glusman loves running experiments. It’s one of the things he did best to help drove growth at Meetup. But he noted a major problem most of us face: that most growth experiments fail. How many? He says about 80-90%. That means that about 1 out of every 5 growth experiments actually moves the needle in a significant way.

    He believes the main reason for this is that we tend to take a biased approach in our assumptions. We think our target customer is like us: they would like or want the same things we would. Rather than validate our hypothesis and check our bias at the door, we move forward assuming one of our experiments will drive significant growth. And as a result, it lands ineffectively.

    Andres believes the solution to this is to implement a sort of “truth-seeking” framework. The purpose of the framework is to quickly get early signals on what’s going to work, so you can more confidently invest your time and money there, in order to ultimately do things that have the biggest impact on your customers. And as a result, you’ll innovate faster, driving more growth, and wasting less time and money along the way.

    Better yet, the framework is really simple. Here’s how it works…

    First, identify the thing you want to work on. 

    It might be growing x amount of revenue or users next year, launching a new product, or moving into a new category. Next, articulate your assumption of how you’ll get there. You might call this your plan, or your hypothesis. It’s what you’re currently assuming is the best way to achieve your stated plan or goal. For example, “we’ll use these 5 marketing channels”, or “we’ll launch this new product, which our customers will love.” 

    Now, it’s time to validate that hypothesis. To do this, Andres recommends asking yourself one key question: “what has to break in my favor, in order to succeed?” In other words, what has to be true, in order for this plan or hypothesis to prove successful?

    Here’s an example:

    Stated goal: gain 3,000 users to your SaaS product next year.

    Your plan ( = assumption of the best way to go about this): run Google Ads, LinkedIn Ads, double down on SEO, and improve your homepage conversion rate.

    “What has to break in my favor, in order to succeed?”

    LinkedIn and Google Ads have to yield an average cost per click (CPC) of $x, and an average conversion rate of y%. Homepage conversion rate would have to go from x% to about y%. And our SEO-driven articles would have to continue converting at x%, and be able to take us from y sessions/mo, to z sessions/mo.

    Now that you’ve identified what would have to be true in order for your hypothesis to be true, you can start to run experiments to find the truth about your market, customers, and situation. 

    Your goal is to try and get to a midpoint, that shows promise of trending the right direction. This midpoint only needs to be about 30-40% of the way toward where you end up, because when you decide to invest fully in this plan, you’ll be able to further optimize it to get the rest of the results you want. 

    In our example, you might spend 6 weeks testing every single channel. And at the end of that, you’ll find that some assumptions were way off, while others were better then expected or right on the mark. For example, you might find you can’t get anywhere near the assumed CPC to make Google Adwords effective, but the rest of your plan shows promise. You find that with some optimization, you’ll be able to hit the numbers you need to drive the signups you want.

    Imagine you’re at sea in a heavy fog looking for a ship. You fire a bullet from your raft, into the darkness, and listen for a “ping” that indicates you’re headed toward your target. Once you get the early validation you need and learn the truth about the market conditions or what your customers want, you can invest more confidently and quickly toward your goal.

    Author's avatar
    Article by
    Jeremiah Rizzo

    More from this author

    Get practical strategies that drive consistent growth

    Read some