A Look at Impactful Growth Cases
When we release a carefully crafted feature in the product, I excitedly check the data the moment it goes live (even before the rollout hits 100%). First, I look at the "usage rate" to see how many people are actually using the feature. Then, I zoom in a bit closer to look at the "conversion rate per funnel" leading up to that usage.
It's rarely easy to achieve a satisfying usage rate right at the first launch. While it takes plenty of thought during the product spec phase on how to drive high usage, the grading on the "good usability" test is ultimately up to the users anyway. So rather than dragging out the deliberation process, it's better to just launch and let the data give you the seeds for improvement. You map the user's journey, find which step in the product's funnel has the highest drop-off, figure out why, define the problem, and fix it. This is what we generally call funnel optimization. In short, funnel optimization is a method for boosting a specific conversion rate within the product.
As someone who picked up knowledge about product growth by throwing myself into it rather than studying books, I don't really know what the "textbook" funnel optimization strategies are. But looking back at my various growth experiences, I think the approach largely boils down to two main directions.
1) Increasing entry rate: The direction of boosting the rate at which users reach the feature's screen (funnel inflow).
2) Increasing post-entry conversion rate: The direction of boosting the conversion rate within the feature itself (in-funnel conversion rate, completion rate).
The two influence each other to some extent. If the "entry rate" is high, the "post-entry conversion rate" might dip because users who aren't even interested in the feature might be mixed into the traffic. Conversely, a low "entry rate" means a handful of users took the time to deliberately find and enter the feature, which can result in a high "post-entry conversion rate."
Most features have a low "entry rate" right after launch. This is because users likely haven't noticed them, and even if they have, the features are still very unfamiliar. Because of this, it's better to interpret the "post-entry conversion rate" a bit more conservatively early on. The metric might be inflated since it's driven by a highly motivated minority of users who are eager to try out new things.
Entry Rate vs. Post-Entry Conversion Rate,
Which Should You Tackle First?
If you need to check the stability of a feature or want to solidify it through user feedback, it's better to tackle the post-entry conversion rate first. You can always reach out to a broader audience after satisfying that core group of early adopters. Separately, if your initial data shows that the leaks in the later stages of the funnel are excessively large, you obviously need to fix the post-entry conversion rate first.
In all other cases, driving up the entry rate is absolutely crucial. What good is preparing a delicious meal if there are no guests around to enjoy it? The entry rate is important when you think about the very reason the feature exists, but it's also vital from an impact perspective. Its influence on the feature's overall usage rate is significantly larger. If increasing the post-entry conversion rate can give you a 1.2x to 1.5x growth boost, increasing the entry rate can easily multiply growth by 10x.
Increasing Entry Rate
When releasing a new feature, you have to decide where to place it within the product. At this point, rather than prioritizing the feature's usage rate, you usually base the decision on what won't disrupt the overall information architecture and hierarchy of the product. This is partly to prevent cognitive overload as features pile up, and partly to help users navigate easily so that the overall usage of all features remains healthy.
Moving a specific feature's location to boost its entry rate is one method, but this should be approached carefully as the side effects, like cognitive overload, can be severe. (Though here is a case where we achieved a 10x improvement without the cognitive dissonance — Multiplying Free Trials by 10x by Repositioning a Feature). Aside from relocating it, another strategy to boost the entry rate is the "entry strategy." This involves creating a shortcut that lands users directly into the feature. The entry strategy can also be divided into two types:
A. Home Entry: A blunt entry that maximizes exposure.
B. Nudge Entry: A highly targeted entry with lower exposure.
I plan to cover the details of each entry strategy in follow-up posts.
Increasing Post-Entry Conversion Rate
From here on out, what matters most is the specific problem defined for each situation. It's tough to give a universal answer to the question, "Why did they drop off at this step?" Reasons like "I didn't understand it," "The steps were too long," or "I didn't feel I needed it" are about as universal as it gets. Therefore, rather than focusing on the specific problems we solved, I think it would be better to share some of the interesting (?) ways we improved the funnel.
A. Reordering the Funnel: Boosting the conversion rate simply by changing the order.
B. Increasing the Number of Funnel Steps: Boosting the conversion rate by "adding" a step that provides more persuasion.
C. Flipping the Funnel Structure: Placing the final conversion CTA at the very first step of the funnel.
D. Eliminating Funnel Leaks: Not just improving the drop-off rate, but removing the problematic step entirely, even if it means changing the product spec.
Likewise, I plan to cover these cases in separate posts.
I didn't start out with a deductive approach, breaking down strategies under the umbrella of "funnel optimization." But looking back inductively, it all organized itself quite neatly like the above. Come to think of it, subscription growth strategies can also be grouped into directions that boost the purchase screen entry rate and those that boost the post-entry conversion rate. Our squad is now expanding our domain coverage beyond just subscriptions to features that touch the entire product. Through that experience, our funnel optimization strategies will only get more robust and diverse.
Stay tuned!