How Creed Doubled ARPU To Scale Ads Profitably With Helium
5min Read
Creed is a fast-growing religious companion app. Here's how they nearly doubled their revenue per user in just six weeks — and used that to scale paid acquisition profitably.
Dmitri Mirakyan
Founder, Creed
Context
Creed's breakout moment
Creed is a religious companion app that burst onto the scene fast. The kind of early growth where every week the numbers look a little better and the opportunity feels a little more urgent.
Dmitri and the team had built something that genuinely resonated — an app that helped people deepen their faith and strengthen their connection to their religion. Users were showing up, engaging, and coming back. The product was working.
Now they needed to scale. And they needed to do it profitably.
The challenge with scaling ads
The math behind scaling paid acquisition is simple: your revenue per user needs to exceed your cost to acquire them. If it doesn't, you're paying to lose money.
Creed's team knew the fastest way to close that gap wasn't to squeeze more out of their ad spend — it was to increase the revenue generated by every user who came through the door. That meant paywall optimization, and it needed to happen now, while the momentum was there.
"The first place that you start with, the place where, in a matter of a couple of months, you should be able to get the dollars per download up by 50%, is going to be at the paywall testing stage."
Why nothing else on the market worked
Before Helium, Dmitri had looked at everything available. RevenueCat lets you test price — but that's about it. Generic A/B testing platforms don't integrate with payment rails. And building it in-house would have taken years.
"There's not a single thing currently on the market that looks like a relatively seamless integration and gives me the data and the tools in one place to effectively run payment tests."
The core problem was that no tool showed the full picture. The moment you introduce trials, discounts, and abandoned cart flows, those experiences start competing with each other. A test can look like a winner on one paywall while quietly cannibalizing conversions elsewhere. Without seeing net impact on overall ARPU, every "winning" experiment was partially a guess.
17 experiments in six weeks
Creed integrated Helium in about a week. From there, the pace was relentless.
With roughly a third of one engineer's time, the team ran 17 experiments over six weeks — testing discounts, abandoned cart paywalls, pricing variations, and onboarding flows. All without shipping a single app store update. For context, most subscription apps run a handful of paywall experiments per year.
"With a third of an engineer over six weeks, we ran 17 experiments. And probably three or four of those drove double-digit improvements."
Each experiment informed the next, creating a compounding cycle where the team got smarter with every test. But what really set Helium apart wasn't the speed — it was the analytics. Helium showed ARPU for users both inside and outside of each experiment, so Dmitri could tell whether a test was genuinely lifting revenue per user or just shuffling conversions between touchpoints.
"Scrolling to the right and seeing the ARPU numbers for people outside of the paywall — that was awesome. That changed everything for us."
Since starting with Helium, Creed's ARPU has nearly doubled
Within a couple of months, Creed's ARPU was up roughly 85%. Nearly doubled.
With almost double the revenue per user, Creed could bid more aggressively on paid channels, reach more people, and grow faster — all while staying profitable. More revenue per user meant more budget for ads. More budget for ads meant more users discovering Creed. And more users meant more people strengthening their connection to their faith — which is why Dmitri built the app in the first place.
"If I had started doing this on my own, I probably would have gotten to where I am with Helium two or three years down the line, at like 100x bigger scale. […] It's a story that's easy to tell because there's no exaggeration needed from my part."
