Growing Self-Serve Revenue at Rapchat (with CEO Seth Miller)

Do not index

Seth and Yohann talk about how Rapchat will achieve its ambitious revenue goals. Spoiler alert: user segments are a key part of their strategy.

Timestamps

00:00 Introduction

05:12 Rapchat's Path to Monetization

09:08 Using Segments For Better Monetization

14:38 What Segmentation Can Unlock

17:37 Correlating Segments with Revenue

23:33 Creating Segmented User Experiences

33:29 Rapchat's Key Revenue Metrics

Highlights from Seth's Appraoch to Self-Serve

How Rapchat began uncovering segments

Certain segments of Rapchat's user base were more profitable than others, as Seth's team discovered when they began selling subscriptions (as a first attempt at monetization after years of being a free product). A simple onboarding questionnaire helped them dig into user intent.

A game-changing thing was learning more about our users in the onboarding journey. Asking them "Why are you here? What do you hope to get outta the product? What's your level of music making?" Once we teased out that type of information, we could ask better questions about what proportion of users fell into the "have fun" segment versus the "aspiring to become the next music star" segment.

"Unsexy" work that matters to users

Their initial monetization efforts helped them realize the user experience mattered to conversions, renewals, and upgrades as much as any fancy feature.

We discovered that a lot of people would either turn off their subscription or leave because of a quality issue, or a bug, or a crash. And so we just spent a lot of gritty hours stabilizing [the product] and replacing our audio engine...stuff that's not sexy, that you wouldn't write a PR about.

Segments can change as users make progress...

An onboarding questionnaire might open the door to segmentation, but it can't replace a deep look at what goals are driving users. Without an understanding of what outcomes matter to different segments of users, the segments aren't a realistic reflection of user intent over time.

We roughly narrowed it down to three or four [segments], but it's tricky because people also move between segments with our product. A couple years ago, this artist named Jaleel made his first song ever on Rapchat, it was the first time he ever recorded over a beat or like recorded any music at all. And then three years later, someone sends me a podcast interview of him mentioning this, and he is now like one of the top up and coming artists, he's a music star now. And so that's crazy because like if you look his onboarding questions in a vacuum, [his answers] would be, "I'm just here to have fun with my friends and record my first song."

We've been trying to understand outcomes beyond just the surface level of "I'm a beginner" or something. Okay, this could be your first song, but then what?And it's hard to go that deep in onboarding, honestly. I think you're somewhat limited. So, [research] needs to be more ongoing than that to help us better understand what people want in the end, what success looks like.

Segmented experiences could be as simple as contextual feature gates

What do you do once you have segments? Use what you know about to tweak your monetization system. This doesn't have to mean overhauling the product, though.

We switched to managing all of our paywalls remotely through a company called Superwall. And that gives us the ability to design contextual paywalls and decide who gets what paywall, or whether we wanna run like special offer campaigns, or design a paywall around a specific value prop after someone uses a specific feature.

On the need for outcome-based metrics

Aggregate metrics (or metrics that report on the behavior of the entire customer base) aren't helpful when it comes to designing for outcomes that particular segments of users are pursuing. Nevertheless, outcomes align user value and business value by defining what exactly is driving the user-business relationship.

What does outcome-based reporting look like? 1) Figuring out which outcomes (and which segments) correlate with revenue and 2) Improving how many users make it to that outcome over time.

I've been working towards a new set of metrics that lines up more with outcome-based thinking. Because to me I really just want to get a better sense of the amount of people that want to accomplish something and see how many accomplish it. I want to have that be a real driving force in my analysis of how we're doing.

Outcomes open up value paths

One of Seth's problems might resonate with most other founders: "we've just built so much stuff." Segments are a way to turn an "undifferentiated mass of stuff" into distinct user journeys that convert better than a single one.

That's the thing: we've built so much stuff that [outcomes] give us a lens of how we can better package it into paths. So, okay, you might not use a particular set of features before you make a song...maybe you're like a level two music maker, and maybe this set of features probably isn't best for level two. We'd say, skip these features and make a quick song. And then we'd introduce all these advanced concepts when you're at level ten.

Growing Self-Serve Revenue at Rapchat (with CEO Seth Miller)