Compounding Improvements
Exploring how tiny changes in a product lead to substantial increases in conversion
The Leaky Funnel
One lesson I’ve had to relearn repeatedly over the years is that, often times, hard things can be made easier by simply reframing how you approach them. This past week, in reflecting on the consumer tech products I’ve built, I was able to put into words (and numbers) one such reframing approach that has always benefited my work. To explain it, I need to take a step back…
Product development is, at its most simplified form, an attempt to answer this question: How do I get the highest number of users to take a specific action?
Once you dig into the actual math, it’s easy to get overwhelmed. That’s because reaching that desired outcome often requires a user to take multiple actions and hit multiple milestones.
This results in a phenomenon I’ve referred to as the Law of Funnels.
law of funnels [lȯ əv fə-nᵊlz] n. — The more steps a person has to go through to do something, the less likely they are to do it.
It’s called a funnel because, at every milestone step, some percentage of users will opt out of performing an action. Earlier in the journey, most users will drop off. But even later, some number will as well. The resulting shape of the conversion graph looks like a funnel:
When in doubt, it’s always better to reduce the number of steps a user needs to go through. The challenge with funnels though is that, at a certain point, they can’t be reduced any more. Consider, for instance, a made-up social platform called Z where users can post content. For the purposes of this article, let’s come up with a hypothetical funnel that’s been reduced to its simplest form:
- User becomes aware of Z
- User visits Z
- User starts creating an account
- User finishes creating an account
- User desires to post content
- User actually posts content
And let’s assume a (somewhat unrealistic) fairly steep drop-off at every one of these steps of 90%. In other words, only 10% of users who reach a given milestone in our funnel will proceed to the next one.
Z suffers from what’s referred to as a leaky funnel problem.
The Slightly Less Leaky Funnel
What fascinates me about funnels isn’t just the fact that they single-handedly determine your conversion rate (they do) or that they represent some of the hardest problems to solve in product development (they do that too). It’s that very minor improvements at each one of those steps leads to a substantial improvement to total conversion at the end.
Let’s look at how many people (starting with the total addressable market at Step 0) actually reach Step 6 of posting content on Z:
Of all of the users who could theoretically post, only 0.0001% (one out of every one million) actually does.
Now watch what happens if, at every step, you convert just a little bit more. Instead of 10%, we’re going to have each step convert 11.2%:
An increase in conversion at each step from 10% to just over 11% doubles your total conversion. You’ve 2x-ed your output. You’ve cut your posting-customer-acquisition-cost in half. All with a little nudge.
The more cynical or mathematically inclined might rightfully challenge me here. You might say, “Increasing from 10% to 11.2% conversion is misleading! Because it’s a percentage point change, that 1.2% difference is actually a 12% increase in conversion. That’s a huge change. Nir, don’t be silly.” The argument rightly brings up that 11.2/10 = 1.12, which translates to a whopping 12% improvement to the product.
It’s here that I’ve discovered the secret reframing strategy.
Do you actually need a 12% improvement to your product? Improvement shouldn’t be calculated on the basis of those who are using your product the right way, but on those who aren’t yet. Rather than aiming to convert more users, let’s think about preventing users from churning. That gives us a much bigger denominator.
In other words, let’s reframe the problem and ask how much effort it would take to reduce the 90% that don’t convert down to just below 89%. We’re ultimately talking about the same numbers. But that inversion makes it clear that we only need just over a 1% improvement to this step in the funnel to raise our effective conversion by 12%.
This subtle reframing entirely changes how we can approach product improvements and the feasibility of converting more users. Rather than frantically telling yourself, “I have to find a way to increase conversion at each step by 12%!” try asking yourself, “Of every 100 people reaching this step, 90 of them are choosing not to. How can I convince just one or two of them to change their minds?” The unachievable just became very achievable. And the impact of that after only a few steps in a funnel leads to big improvements.
(Side note: If you’re enjoying this article and aren’t yet a subscriber to my free weekly newsletter, please consider subscribing to Z-Axis here!)
The Power of Compounding
This example showcases the power of the phenomenon of compounding numbers. Typically, compounding effects manifest in the form of interest earned over time.
So let’s do the same thing here. Let’s look at the impact that doubled total conversion would have on the overall health of a business over time.
First, twice as many converted users means twice as many evangelists for your product promoting your product via whatever growth mechanic you may have built in.
Second, twice as many converted users means twice as much revenue to reinvest in development.
And both of the above similarly benefit from compounding effects. Your growth flywheel spins faster and brings more users into the top of the funnel the more evangelists you have. Your reinvestment in development is further opportunity to raise the conversion of each funnel step.
And that better conversion leads to better growth leads to more revenue leads to more improvements leads to better conversion…
In other words, the effects of funnel improvements don’t only compound. Their compounding compounds.
Albert Einstein apocryphally said, “The strongest force in the universe is Compound Interest.” In reality, he probably didn’t actually say that. And in fact, the strongest force in the universe is the strong nuclear force. But the sentiment is fitting nonetheless.
What’s Next?
If you’re interested in reading more about topics related to business, tech, startups, etc., check out my newsletter, Z-Axis. Here are some articles I’ve written recently that you might enjoy:
- How AI Works: An entirely non-technical explanation of how AI works, using analogies of food and meal planning.
- Doing Things That Don’t Scale: The best advice I know for someone trying to get a new idea off the ground.
- The Faux First Mover Advantage: Being first isn’t as good as people thinking you were first…
- You’re At a Local Minimum: Using concepts of machine learning and math to understand why people get comfortable in their jobs.
- The Triangular Friendship Experiment: A new way to make new friends!
