Onboarding streak hits day 4 — the nudge lands on day 6

There's a question I keep coming back to when I'm building onboarding flows for clients: why does the same notification that gets ignored on day four suddenly work on day six? The message hasn't changed. The user hasn't changed. But something in the gap between those two days has shifted, and if you can figure out what it is, you can stop guessing at your email timing and start engineering it.

The streak is a story, not a scoreboard

When someone signs up for a product and completes an action four days running, they're not just accumulating a number. They're building a self-image. "I'm the kind of person who does this." That identity is fragile early on, and it's why streaks are so psychologically potent — they convert a behaviour into a claim about who you are.

Behavioural researchers call this the consistency principle, and Robert Cialdini wrote about it decades ago in Influence. Once people take a small public or even private stand, they feel pressure to behave in ways that match it. A four-day streak is a tiny stand. It says: I've invested here. I'm not a quitter. But it's not yet load-bearing. Four days is still short enough that abandoning it costs the user almost nothing emotionally.

By day six, something changes. The user has now crossed a threshold where the streak has become part of their routine, not just their recent history. They've likely told someone about it, or at least thought about it in the shower. The identity claim has hardened. And that's when a nudge that would have bounced off on day four suddenly lands.

Why day four is a dead zone

Day four sits in an awkward valley. The novelty of signing up has worn off. The habit loop hasn't fully formed — according to research on habit formation, the median time to automaticity across behaviours is around 66 days, with huge variance. Four days in, the user is still consciously deciding to show up. That's expensive. It's cognitive load.

More importantly, day four is often when the first real friction appears. The easy wins are gone. The product has started asking for something — a credit card, a data connection, a bit of effort — and the user is weighing whether the streak is worth the cost.

Send a nudge here and you're competing with the user's own internal calculus. You're an interruption, not an ally. The message reads as pressure because the user hasn't yet decided they want to be pressured.

The day-six window: loss aversion kicks in

Here's where things get interesting. Loss aversion — the finding from Kahneman and Tversky that losses feel roughly twice as painful as equivalent gains feel good — doesn't apply to a four-day streak the way it applies to a six-day one. At four days, the user is still framing the streak as a gain they're building. At six, they've started framing it as something they have. And people will work harder to avoid losing something they have than to gain something they don't.

I saw this play out on a client project last year. An Australian SaaS tool with a daily-check-in feature. Their onboarding sequence sent a reminder on day three and day five. Day three: 11% click-through. Day five: 9%. Basically flat. We moved the second nudge to day six and reframed it from "Keep your streak going!" to "You've got a six-day run — don't break it now." Click-through jumped to 19%. Same audience, same product, same offer. The only change was the day and the framing.

The framing mattered because it shifted the user's mental accounting. It wasn't a reminder to do something. It was a warning about a loss. And losses, per Kahneman, loom larger.

Variable rewards and the anticipation gap

There's a second mechanism at play, and it's the one that makes day six feel different from day four even before you send anything. Variable-ratio reinforcement — the schedule where rewards come unpredictably — produces the most persistent behaviour of any reinforcement pattern. It's why people keep checking their phones, and it's why streaks work.

But variable-ratio reinforcement needs a baseline of expectation. The user has to believe a reward might come. On day four, the user hasn't yet experienced enough of the loop to trust it. They don't know if day five will feel good or feel like a chore. By day six, they've had enough reps to have formed a prediction. They're anticipating. And anticipation is where the dopamine lives — not in the reward itself, per Wolfram Schultz's work on reward prediction error.

So the day-six nudge isn't creating desire. It's intercepting desire that's already forming. That's a fundamentally easier job.

What this means for how you build onboarding

The practical takeaway isn't "send your nudge on day six." It's that your onboarding timing should be indexed to the user's psychological state, not to a calendar. A few things that follow from that:

Track identity markers, not just actions. If a user has named their project, invited a teammate, or customised a setting, they've made an identity claim. That's a better trigger for a nudge than "day three."

Send the loss-framed message, not the gain-framed one, once the streak is real. "Don't lose your progress" outperforms "Keep going!" — but only after the user has enough progress to feel like losing it. Before that, loss framing reads as manipulative.

Let the streak breathe before you lean on it. A nudge on day two is noise. A nudge on day six is a service. The difference is whether the user has something to protect.

Watch for the friction point, not the calendar. The best moment to nudge is right when the user is about to encounter the first real cost of the habit. If you can time your message to arrive just before that friction, you're not interrupting — you're helping them over a hump they were already approaching.

Where I'd look next

The next frontier here is personalisation of the threshold, not the message. Different users hit the identity-hardening point at different times. Some get there on day three. Some never do. If you can detect the moment a user's streak has shifted from "building" to "having" — through language in their support tickets, the pace of their actions, whether they've referred someone — you can time the nudge to land exactly when loss aversion is working for you instead of against you.

That's a harder build than a drip sequence. But it's the difference between a nudge that gets ignored on day four and one that lands on day six. And in onboarding, that difference is the whole game.