A shopper selects a product, starts checkout and then encounters a new fee, an account requirement or a failed payment. More advertising sends more people to the same obstacle. Review the path from basket to confirmed payment before buying additional traffic.
At a glance: use your store’s data to choose checkout changes. Check delivery costs, account creation, forms and payments, then measure paid orders and contribution after fees, discounts and returns. Austria, Germany and Switzerland need separate payment assessments.
Why do shoppers abandon carts?
Baymard’s compilation reports 70.22% average abandonment across 50 studies, with the list last updated on 22 September 2025. It is a broad reference, not a DACH norm or a target for an individual store. Browsing, price comparison and postponed purchases also create abandoned carts. Baymard’s compilation.
Separate adding to cart from starting checkout. Otherwise a payment failure and a shopper comparing products become the same event in your report.
| Possible barrier | What to check |
|---|---|
| Unexpected costs | Shipping, fees and the full amount before the final screen. |
| Mandatory account | Visible guest purchase, login and password recovery. |
| Delivery too late | A realistic delivery window before confirmation. |
| Lack of trust | Contacts, returns, payment operation and clear terms. |
| Difficult form | Unnecessary fields, autofill and useful error messages. |
These checks reflect established checkout usability research. Their prevalence in your shop requires your own evidence. Combine funnel data with payment failures, support questions and user observation. Session recordings require appropriate data handling and masking; sensitive payment fields must not become research material.
Which payment methods should you assess in DACH?
Availability depends on merchant country, buyer, currency, product category and provider. A logo in a presentation does not establish an available integration.
| Method | What to verify |
|---|---|
| Cards | Supported markets, authentication, failures, fees and refunds. |
| PayPal | Availability for your business and buyers, checkout flow and contract. |
| Klarna | Available payment options, markets and category restrictions. |
| EPS | Suitable Austrian customers with a supported bank account. |
| TWINT | Support for Swiss buyers through your chosen integration. |
| Apple Pay / Google Pay | Actual availability on the customer's device and browser. |
| SOFORT | Legacy status: Stripe ended the standalone method on 31 March 2025. |
Stripe documents EPS as an Austrian bank payment method and Klarna options as dependent on market and payment conditions. Klarna also serves supported Swiss transactions. Check the current configuration rather than assuming all options apply.
TWINT supports online shops through providers and integrations. Assess it for Swiss customers on actual demand and terms. For a legacy Stripe SOFORT setup, use the discontinuation notice and available alternatives.
Do not add payment-preference percentages to calculate unique customer coverage: the same person may use several methods. Select additions by completed payments, customer requests and incremental contribution after costs.
When is guest checkout useful?
A one-off fashion, gift or seasonal purchase should not unnecessarily become an account setup task. On mobile, switching to email for verification can also interrupt the journey.
Make guest checkout visible and request the information needed for payment and delivery. Accounts can serve subscriptions, service history and loyalty benefits, but a high order value alone does not require registration. Test the needs of your buyers rather than applying a universal €100 threshold.
An optional account after order confirmation is a useful alternative. Explain its benefit without silently subscribing the buyer to marketing. Also test returning customers: autofill, login and password recovery should preserve the basket. Guest checkout will not repair an address field that rejects valid postcodes.
One page or three stages?
The screen count alone does not determine usability. Buyers need to understand the remaining steps, go back and correct information without losing it.
Three stages — delivery, payment, order review — can organise multiple countries, collection options and different B2B billing and delivery addresses.
One page can work for a short flow, everyday purchases and express payment. A long page with concealed sections can still be confusing. Test keyboard use, autofill, back navigation and failed payments, especially on mobile.
Choose a meaningful comparison event, such as confirmed paid orders among people starting checkout. Review cancellations and repeat payment attempts separately. More clicks on Continue do not establish a better checkout.
Two illustrative tests
Fashion store: compare a more visible guest option or clearer delivery cost with the current path. A popup or new payment provider should not obscure the order total or become another mandatory step.
Electronics store: evaluate address entry, order review and express payment for buyers who need delivery and warranty details. Compare mobile completion with the existing flow. These are test designs, not reported client results.
Calculate revenue with the correct period
Suppose monthly eligible baskets remain at 5,000 and paid order value at €85. A hypothetical fall in abandonment from 74% to 58% means conversion rises from 26% to 42%: 800 additional orders and €68,000 revenue per month, or €816,000 over twelve identical months. The calculation assumes the same denominator and basket conditions; it is neither a client case nor a forecast.
Revenue is not ROI. Deduct product costs, payment fees, discounts, fulfilment, returns and implementation costs. Use a suitable comparison to establish what the change caused. Our A/B-testing guide explains the measurement approach.
Which trust information belongs near payment?
Show the full price, delivery window, return conditions and a way to contact the shop. A Trusted Shops or TÜV badge requires a current right to use it. Payment logos should match methods buyers can actually select.
HTTPS is necessary; a badge saying “100% secure” does not establish risk-free purchasing. A money-back promise must match the actual terms. A privacy notice must reflect all processing purposes; “only for your order” is inaccurate if separate marketing also occurs. See our privacy guide.
Unknown badges and long logo rows can take space from practical answers. On mobile, keep the price and delivery conditions easy to find.
How should you test cart recovery?
First confirm that the order is unpaid. Delayed payment confirmation must not trigger reminders to someone who has already bought.
An entered email address does not automatically authorise marketing. In Austria, direct-marketing email generally needs prior consent, with a limited existing-customer exception subject to the statutory conditions. A partially entered basket is not sufficient on its own. Check the buyer’s applicable rules. TKG 2021, section 174. Personalised Google advertising also requires the disclosures and consent covered by its EU user consent policy.
Where contact is permitted, an example test sequence is:
- Around one hour: a simple product reminder without a discount.
- Around 24 hours: delivery information, returns or a genuine review.
- Around 72 hours: a limited incentive if its economics work, then stop the sequence.
The timing is a hypothesis. Do not invent viewer counts or use a resetting urgency timer. Email and dynamic retargeting may claim the same sale; do not add their attributed orders as incremental revenue. Evaluate discounts against margin and the possibility that repeat buyers learn to wait for a coupon. Our cost guide helps frame those decisions.
Three checks for this week
Day one: use the store’s test mode on phone and desktop. Record fees, fields, errors and the point where conditions become visible. There is no universal maximum of four clicks.
Day two: review failed payments and support questions. Select one likely obstacle, then evaluate provider fees, integration effort and how to restore the previous setup if needed. Availability does not imply a one-to-two-day implementation.
Day three: prepare a limited change and the success criterion, such as guest checkout visibility measured by paid completion. Our CRO service describes the diagnostic process; a valid A/B test depends on traffic and design, not a fixed four-week promise.
Frequently Asked Questions
What cart abandonment rate is normal?
There is no universal target for a store. A broad research average does not establish your category or traffic mix. Check basket, checkout and confirmed-payment measurement, then compare periods and segments with consistent definitions.
Is Klarna worth the fee?
Use your current provider contract and assess additional paid orders, order value, refunds and contribution after costs. A higher average order value alone does not prove profitability, and there is no universal fee for every integration.
How do I calculate a checkout improvement?
For illustration, 5,000 eligible monthly baskets, conversion from 26% to 30% and €85 per paid order give 200 extra orders and €17,000 monthly revenue before costs. It is a calculation, not a client result or forecast. Establish causality with a suitable comparison and assess contribution after product costs, fees, discounts and returns.
Should I use an exit popup?
Test it only if it leaves fields, terms and the payment button usable. Check closing, mobile behaviour and repeat displays. Fixing delivery information may be more useful; a coupon can discount an order that would have happened anyway.
How often should I review checkout?
Set a regular review, such as quarterly, and repeat it after platform, payment or delivery changes. Monitor payment failures and sudden completion drops more frequently rather than waiting for the next review.
If shoppers reach the basket but the losses are unexplained, discuss your checkout with Canem Errant. Describe the store, buyer markets and failed step so we can agree the diagnostic scope.
