When we look at install cohorts across the Appolar customer base, one number jumps out reliably: the percentage of installs that produce an order within 72 hours is the single best predictor of that cohort's six-month retention. Cohorts that hit 25% first-order-in-72-hours retain meaningfully better than cohorts that hit 10%. The mechanism is not mysterious. The fast first-order group has installed for a reason; the cold install group is curious at best.
This piece is about how to engineer the 72-hour window. Not by adding more steps to onboarding — there are already too many steps in most apps — but by tightening three specific moves at install time.
Why 72 hours, specifically
The window is empirical, not theoretical. We tested a range — 24 hours, 48 hours, 72 hours, 7 days, 14 days — against six-month retention and the 72-hour mark consistently best predicted long-term behavior. Beyond 72 hours, the predictive power flattens. Within 72 hours, every additional hour matters.
The intuition: the first 72 hours is the window when the brand is still novel in the buyer's phone routine. They installed for a reason, they remember why, and the app still has the shine. After 72 hours, the brand becomes one of the dozens of icons on the home screen, competing for attention with every other thing. The first-order behavior captures the buyer while the conversion window is still open.
Move 1: a clear install-time offer
The first move is the install-time offer. A clear, simple value proposition the buyer sees in the first 30 seconds: "First order in the app, 15% off, automatically applied at checkout." Not a code to enter, not a tier to qualify for. A discount that lives in the cart from the moment they open the app.
The exact number matters less than the clarity. 10%, 15%, 20% all work depending on margin tolerance. What does not work: a tiered or conditional offer ("15% off orders over $100"), an offer that requires the buyer to find a code, an offer that competes with the brand's other discounts. Simple beats sophisticated.
The lift on install-to-first-order rate from a clean install-time offer is substantial. Brands that ran no offer hit 35–45% first-order conversion in 72 hours. Brands that ran a clean offer hit 55–70%. That is the largest single lever you can pull, and it costs you margin on orders you would otherwise lose entirely.
Move 2: push opt-in tied to perceived value
The second move: ask for push permission at a moment when the buyer has experienced value, not at app launch. The iOS default — asking immediately when the app opens for the first time — yields 40–55% opt-in rates. Asking after the first product save, the first cart add, or the first order completion yields 70–85% opt-in.
The framing matters. "Tell me when this item is back in stock" is a useful question with a clear answer. "Allow notifications" with no context is an interruption the buyer says no to. Tie the ask to the action they just took.
Push permission compounds in importance. Every buyer you fail to get on push is a buyer you cannot re-engage with the lowest-cost channel you own. Over six months, a 30-point difference in opt-in rate translates into a much larger difference in push-driven revenue. The investment in better opt-in timing pays back for the life of the install.
Move 3: the first-week reorder or saved-item hook
For buyers who place a first order within 72 hours, the next leverage point is the second order. The earlier that lands, the higher the lifetime retention. The mechanism: a push notification at day 5–7 with a clear hook — an item they viewed but did not buy, a complementary product to what they ordered, a "you saved this, here is what is in stock" reminder.
The hook should be specific to the buyer, not a broadcast. Buyers who got specific second-order hooks placed a second order within 30 days at 40–55% rates; buyers who got generic broadcasts placed a second order at 20–25% rates. Specificity is the lever.
“The first 72 hours is where the install graph proves itself. After that, you are no longer optimizing onboarding; you are optimizing reactivation, which is harder and more expensive.”— Cohort analysis notes, Appolar
What not to do in the first 72 hours
Three traps. Trap 1: too many onboarding screens. Every brand wants the buyer to see the full brand story, the routine builder, the size guide, the sustainability page, the founder letter. The buyer wants to see products. A long onboarding flow loses install-to-purchase conversion. One screen, maybe two, is enough.
Trap 2: gating the experience. Forcing account creation before the buyer can browse, requiring email signup before seeing the catalog, asking for preferences before showing products. Each gate is a tap that loses a percentage of the install cohort. Let them browse first, capture them after they have engaged.
Trap 3: overusing push in the first 72 hours. The instinct is to push a welcome message, a "did you forget?" reminder at day 1, a campaign at day 2, an offer at day 3. The result is push fatigue before the buyer has even bought. One welcome push, one targeted reminder if they have items in cart, and silence otherwise.
How to measure the 72-hour window
Track three metrics per install cohort, weekly: percentage with first order in 72 hours, percentage with second order in 30 days, percentage opt-in to push. These three numbers together capture most of what matters about the onboarding experience.
Improving any of them is concrete, not abstract. A 5-point lift in first-order-in-72-hours is worth more in six-month revenue than a 50-point lift in some downstream engagement metric, because the 72-hour cohort is the one that compounds. Spend your optimization energy here.
