Howdy guys!

So, I completely scrapped my first onboarding flow for Little Moments and rebuilt it around one idea: don't explain the app. Make people use it and then leave.

That decision changed everything about how the first two minutes feel.

Let me walk you through what happened.

My first onboarding flow was this
Splash screen (3 slides) → Sign Up → Why did you join? → A personalized Answer → 6 marketing slides showing off 6 different features → A 2-part upgrade page → Enable Notifications → Today Page

I mean, that was tiring to even read and type.

I loved making it.

But after 3 people told me they were confused in quick succession, I realized I’d made several selfish decisions. And for a consumer app, onboarding is literally the most important thing you have. It’s like speed dating…2 minutes or less to make someone interested in coming back. It could be even less than that.

It was beautiful though.
I put in work making pretty screens and nice graphics in Figma.
And I was precious and proud of it.

I'd built all these features and I wanted everyone to know about every single one. That was selfish. Onboarding isn't a product show-and-tell to get a gold star for UI—it's the user's path to realizing why they're here and why you’re worth their time.

Onboarding without action is a TED talk.

What I had made was too long and had multiple attempted hooks—so none of them actually hooked anything and landed. It was self-indulgent and not focused on one thing I needed a new user to do in order to stay…capture their first moment.

After a little sulk, I had to decouple my preciousness and realize I was putting pride in the wrong thing. And made a note to self: you can't be precious in vibe coding when making things is this easy.

So, on Saturday morning I took it out back behind the barn. And there it stayed.

And now I really am proud of it and the feedback has been a complete 180.

The v2 onboarding is led entirely by Ellie — the elephant companion inside Little Moments. I love elephants…they have the best memory of all animals 🙂

Instead of slides explaining features, she walks you through actually using the app. Chat-first. Everything feels like a progressively guided conversation, not a tutorial.

Splash → Sign up → Why are you here? → Personalized letter → Ellie Activation → App

A splash screen focused on one hook.

Google, Apple, or email sign up.

Then the interesting part starts.

I kept this from v1.

Before you touch any feature, I ask one question. Six statements, pick what lands. For example

Based on what you pick, the app (pre configured of course) knows your main reason for being here. Pick "days are blurring together" and it knows you care about holding onto time. Pick "I want a habit that sticks" and it knows consistency matters to you.

Then the next screen speaks directly to that reason. Tell the app you feel like days are blurring together and you get: "You've noticed it — days blurring together, weeks disappearing before you've had a chance to hold onto them. You're not losing your memory. You're just not catching the moments before they slip away. That's what Little Moments changes."

Say you want a journaling habit that actually sticks? "You've tried journaling before. Blank pages, guilt when you miss a day, apps that ask too much. We get it. Little Moments is different."

The idea here is I want the user to immediately feel like the app gets them. And that feeling — before they've used a single feature — is what makes them lean in.

I also get good data on why people are joining which helps me prioritize different things to build and ways to pitch Little Moments.

After the personalized letter, Ellie takes over.

She introduces herself — quick, warm, casual — and walks you into your first word prompt—the main activation event to save a moment. She gives you a word. You hit Start. Two-minute timer. Speak or type whatever the word triggers. A memory, association, story…a few sentences or a ramble.

You're not reading about the feature. You're doing the feature. In the first minute of using the app and you've already captured a moment.

Then Ellie saves it, shows you how it looks in your timeline, and you can feel the thing working. That's the activation event — not signing up, not completing a profile. Saving your first moment.

After that, there's an optional photo prompt — the app pulls a photo from your camera roll and you talk about it. I had to do some work to make sure it was a real photo, good quality, and not a shitty screenshot. If it’s a bad pic, that’s a bad experience.

The goal here is to activate the second big prompt pattern—photos.

My favorite prompt I’m actually kind of addicted to already.

You get to experience it, you don't read about it. The better I get my photo retrieval logic (p.s we don’t store any user pics it’s all on device) the better the aha moment of this will be.

Then notifications, directly in chat, with a clear value prop as to why you should enable them.

Up until literally this morning, I was trying to get power users to activate and try the third type of daily prompt…a question. I’ll show my data on this below, but here’s the bottom line and the lesson.

Activation was very successful so far, users who activated a word almost always activated a photo…capturing two moments. A great finding.

Then they clicked to initiate the third prompt type: a question asking “What was a little moment from the past few days you’d share at the dinner table?”.

Before choosing to do the third type, users could skip it and go to the Home screen. But clearly the steps before were working so well, people wanted to try the third prompt type to see it.

But once they started…huge drop off. And I realized after going through it a few times myself, that

a) this screen had no exit path…you literally couldn’t opt out and the only way to get to the app was by completing it…and most importantly…

b) I didn’t know what to say.

This was way to much friction for onboarding. Activation 1 and Activation 2 required NO THINKING. Simple. Just talk. But Activation 3 was pressure.

And there’s good friction and bad friction. This was bad friction I caught early.

So, I removed it. And instead, I am closing onboarding off with a final message from Ellie.

She sends a few quick messages reemphasizing what you’ve just done. The goal is to close out the experience without any “What now?”. At this point, not only is onboarding done, but I don’t need the user to do anything else besides come back tomorrow.

I am trying something though before completion that I’ve never seen in any other app—so, we shall see if I am smart or I am silly 🙂

See below, but Ellie says this

This might not be the right question, but the purpose is that I want to know before I let the user into the wild, and while I have their attention, if they feel good about what comes next and how Little Moments can help them.

I’m already in the Chat-first UI, so why not leverage that to remove doubt and really make sure this new user feels good and gets it.

Besides tracking their answer (which can help me improve onboarding and is an interesting metric), it’s more so about using that answer to either send the user into the app, or take them down an optional discovery path.

See below…I have some extra questions the user can ask Ellie during onboarding. They can skip to the app, or ask as many as they want to.

This I think is good friction, and only surfaces more info to those who need and ask for it.

Then into the app.

Done.

Now in reflection, in theory, there are actually MORE steps in this flow than v1 onboarding — but they don’t feel like steps. They are actual app usage actions. I did consider another option…what about just Splash → Sign Up → App

Let people play with the app and try it out.

But I remembered this Lenny case study from a company that nearly killed their business by going freemium and removing onboarding friction. Their biggest lesson? "The goal of onboarding is not for people to complete onboarding. It's for people to get their first moments of value from your product." Removing friction is actually detached from that goal. Sometimes the right friction — the kind that makes you do the thing — is exactly what keeps people around.

That's the whole bet I'm making with Ellie Activation.

By the time you land on the home screen, you've already completed at least one daily moment. Maybe even two.

You're activated. There's nothing left to figure out. Come back tomorrow.

Two small details I'm proud of.

If we don't have your name from Apple or Google sign-in, Ellie sends you a message on the home screen: "Hey, I don't know your name yet." It's not a settings field. It's a conversation.

Same with notifications — if you skipped that step during onboarding, Ellie nudges you later in chat. Not a system prompt. Not a modal. A message from a character you've already met.

Small things. But they make the whole experience feel like you're talking to someone, not configuring an app.

15 real users. Tiny. A few friends, and some folks here who messaged me down to try it out before release . But it's cool to see actual numbers from a direct cohort. Here’s what mine look like so far…

Speaking of tracking all this though …

Install → Sign up → Started activation: 100%

Every single person who installed the app signed up and started the activation flow. No drop-off at the top. Whatever I'm doing to get people from splash to activation, it's working. Sign up is smooth. Don't touch it.

Word activation (core activation): ~80%

12 of 15 saved their first word. Three dropped during the word step — that's my first real friction point. Need to watch the session replays to understand why. Bug? Confusion? Just got interrupted? With 15 users, every person still counts.

Super activation (word + photo): ~75% of word-activated users

9 of the 12 who saved a word also did the optional photo prompt. Given it's optional, that's a strong signal — engaged users are leaning in.

Question prompt (3rd optional step I killed this morning): almost nobody

After the photo, there was that optional question prompt. Almost no one does it during onboarding. This was a signal I couldn’t ignore.

The question I need to keep asking as more people join goes way beyond onboarding: which prompt type — a random word, a photo from your camera roll, or a question — has the best capture rate day to day? Which ones get people sharing fastest? Do users who lean into one type over the others behave differently — share more, upgrade more, stick around longer? I need more data. But this is the crux of what I need to figure out. Not just immediate activation, but how onboarding feeds into the daily habit, and which habits are best to build for stickiness and monetization.

Reached the today screen in first session: ~20%

Only 3 of 15 people made it to the today screen during their signup session. 12 people activated but only 3 landed where I want them to land. I found out where the other 9 were going after activation…they couldn’t get past the question prompt after opting in to doing it, and nobody was doing it.

I hope this is fixed now.

The sample is small obviously. But even with 15 people, the shape of the funnel tells me something: top of funnel is healthy, activation is the right bet, and there are specific things to fix.

I might be wrong about the question prompt as a daily rotating type altogether. The data from onboarding says almost nobody engages with it there. Maybe it works better as a power user tool for general moment capturing — when you just want to log something and need a starting point — rather than part of the daily rotation. The daily habit might be best suited to words and photos. I need more users to tell me.

But the core bet — activate during onboarding, show don't tell, chat-first through Ellie — that feels right. And 80% activating suggests we’re on the right track.

Here’s what Zac said to me…

My goal with onboarding is simple: get the user to Capture a moment, and get the purpose of capturing moments. If they see it's free, easy, and the value compounds — a memory Capsule that grows, monthly Chapters we make for you, Threads we connect across your moments, an archive you can share with people or look back on — then my onboarding job is done.

V1 I was proud to show off design skills and nice UI and app shots. But again — selfish. This is not a pony show. It's a mile of the app built for the user to get in, get value, and get out. So they feel good about coming back tomorrow.

If you want to find out if I achieved that — hit me up and I'll send a link. Shoutout to the folks who already jumped in.

I'll report back soon folks.
— Jaryd

(view my stack of non-negotiable building tools here)

———————————————————————————