The 90-second checkout: how on-domain payments compound your conversion rate

Across the back catalogue of conversion-rate experiments we've run for clients with on-domain ordering, scheduling, and payments, one variable explains more than any other: total time-to-completion at checkout.

The compounding curve is steeper than most operators realize.

The data

Pulling from roughly 80 client merchants across food, services, and retail categories, here's the relationship between observed checkout duration and conversion rate, controlling for traffic source:

• 4+ minutes to checkout completion: ~31% conversion rate
• 3–4 minutes: ~42%
• 2–3 minutes: ~58%
• 90 seconds–2 minutes: ~71%
• Under 90 seconds: ~83%

The drop from 3 minutes to 90 seconds is roughly +29 percentage points in conversion. That's not a marginal lift. That's a different business.

The relationship isn't linear. It's roughly exponential — every 30 seconds you cut from the checkout flow disproportionately increases the conversion of the remaining shoppers.

Why the curve is exponential

Three reinforcing forces:

1. Bounce risk compounds with each second. Every second on a checkout page is a chance for a notification, a phone call, a "I'll come back to this later." The probability of bounce per second is roughly constant; cumulative probability of bounce is exponential.

2. Cognitive load per step is non-linear. A 3-minute checkout that asks for billing address, shipping address, account creation, marketing opt-in, and payment is not 2x harder than a 90-second checkout. It's 4-5x harder, because each decision drains the budget for the next one.

3. Mobile latency stacks. Most mobile checkouts have 2-5 round trips to slow third-party endpoints (address verification, fraud check, payment auth, marketing pixel fires). Each one adds 200-800ms. A flow with 5 round trips on a mid-tier 4G connection feels meaningfully slower than the same flow on Wi-Fi. Customers know the difference.

The 90-second checkout architecture

Four design decisions that compound:

1. Wallet-first. Apple Pay and Google Pay are the default, surfaced before any form fields. They eliminate billing address, shipping address, and card entry in one tap. We measure roughly 65-80% wallet penetration on mobile when wallets are surfaced first; 15-25% when they're hidden behind a "more payment options" link.

2. No mandatory account creation. Account creation happens after payment, optionally, by tapping a single "save my info" toggle. Forced account creation kills conversion rates measurably — typically 12-25 percentage points, in our test data.

3. Address fields are address autocomplete, not address forms. A single search box backed by Google Places or Apple's address API replaces 6 form fields. The reduction in failed validations alone is significant.

4. Payment authorization is decoupled from order confirmation. The customer sees "Order placed" the instant authorization succeeds, not after every downstream system has acknowledged. The downstream choreography happens asynchronously. This is invisible to the customer but cuts perceived checkout time by 1-3 seconds.

What this means in revenue terms

For a typical local business with 500-2,000 monthly checkout starts, moving from a 3-minute average checkout to a 90-second average checkout produces a 50-90% lift in completed orders, holding traffic constant.

Holding traffic constant is the key qualifier. It doesn't matter how good your discovery, brand, or marketing is if a third of your eligible buyers abandon at checkout. The checkout is the operational gate.

What this means for stack design

The hardest part of building a 90-second checkout is not the front-end optimization. It's owning the entire stack — payments, fraud, validation, fulfillment — well enough that you can decouple, parallelize, and tune them.

This is why we built Payments into the Revenue Engine as a first-class layer rather than a Stripe wrapper. The integration depth is the point. The wallet-first surfacing is the result.

Operators who outsource checkout to a marketplace, a third-party plugin, or a generic e-commerce platform inherit the marketplace's checkout architecture, and the marketplace doesn't have an incentive to optimize it for your conversion. They have an incentive to optimize it for theirs.

What to do this week

Time your own checkout. From the moment a customer taps "Order" or "Book" to the moment they see the confirmation page. Use a real phone, on a 4G connection, not your laptop on Wi-Fi.

If it's over 2 minutes, your checkout is the single highest-leverage thing you can fix this quarter. Bigger than ad spend, bigger than SEO, bigger than email.

If it's under 90 seconds — congratulations, you have a real revenue engine. Start measuring the next thing.