Checkout Optimisation: Reducing Abandonment at Your Most Critical Point
Approximately 70 percent of shopping carts are abandoned. These are visitors who found products they wanted, added them to cart, and started checkout. They were as close to purchasing as visitors get. Then something stopped them.
.jpg)
The friction points that cause 70% of carts to be abandoned
Unexpected costs are the primary abandonment driver. Shipping charges, taxes, and fees revealed late in checkout feel like bait and switch. Visitors who expected one total and discover another often abandon rather than accept the discrepancy.
One of the best marketing decisions we made for one of our clients (U&U a ceramics kitchenware brand) was to state on the product page that all orders incurred a £2.99 shipping charge due to the fact that they use more packaging than most to protect their products (ceramics) whilst shipping on the product page, but in the checkout page we designed a message that said ‘as a first time customer we’ve waived our shipping fee’. This did two fundamental things. First, it reassured customers we cared about the product we made, and wanted to make sure it got to them safely. The second, made them feel like they were rewarded and we gave them something we didn’t have to by reducing the cost in the checkout.
Forced account creation is the second major driver for abandonment. Visitors who want to buy and encounter a registration wall face unwanted friction. Some comply. Many leave. The data captured through forced registration rarely compensates for the sales lost.
Complex or lengthy checkout processes create cumulative friction. Each additional step, each additional form field, each additional decision is an opportunity for doubt to emerge and motivation to dissipate. The visitor who was ready to buy at step one may reconsider by step five.
Guest checkout versus account creation: what the data shows
Guest checkout consistently outperforms forced registration. The friction of account creation costs more sales than the data captured is worth. This is especially true for first time customers who have not yet demonstrated loyalty.
The optimal approach offers guest checkout prominently while making account creation optional and beneficial. 'Create an account to track your order and earn rewards' converts better than mandatory registration gates.
Payment options and trust signals that move the needle
Payment preference varies by customer segment and purchase context. Missing a preferred payment method creates friction for customers who would have converted. Credit cards remain primary, but PayPal, Apple Pay, Google Pay, Klarna, and other options each capture customers who prefer them.
Trust signals matter most at checkout because risk feels highest when committing money. Security badges, payment logos, money back guarantees, and clear contact information reduce perceived risk. Their presence reassures. Their absence creates doubt.
Product Page Psychology: What Makes Visitors Add to Cart
Product pages are where purchase decisions are made. Visitors arrive with interest and leave with conviction or abandonment. The elements on this page determine which.
Effective product pages answer questions before they are asked. What exactly is this? What does it look like in use? Will it fit or suit me? Is it good quality? Can I trust this seller? What if I do not like it? Each unanswered question is friction that reduces conversion.
mages dominate first impressions. Multiple angles, lifestyle shots showing use, zoom capability for detail. Poor imagery undermines everything else. Strong imagery creates desire that copy reinforces.
Reviews and social proof address trust at the moment of decision. Seeing that others purchased and were satisfied reduces perceived risk. The absence of reviews creates uncertainty that some visitors will not overcome.
Category and Navigation: The Silent Conversion Killers
Site search is often the fastest path to purchase for visitors who know what they want. Poor search that fails to find relevant products drives these high intent visitors away.
Filtering enables visitors to narrow large catalogues to relevant options. Missing filters force visitors to scroll through irrelevant products. Poorly implemented filters that produce empty results frustrate and abandon.
Mobile navigation requires different patterns than desktop. Thumb reach zones determine what is accessible. Screen size limits what can be displayed. Touch targets require adequate spacing.
Many sites simply shrink desktop navigation onto mobile screens. The result is navigation that is technically functional but practically frustrating. Mobile specific navigation design acknowledges the different constraints and optimises for them.
More options can reduce conversion rather than increase it. The jam study famously showed that 24 jam varieties attracted more browsers but 6 varieties produced more purchases. Choice overload creates decision paralysis.
Ecommerce sites with large catalogues must manage this paradox. Curation, recommendation, and guided selling help visitors navigate choice without overwhelming them. 'Best sellers', 'recommended for you', and 'frequently bought together' reduce cognitive load while maintaining selection breadth.
.png)
Testing Strategy for Ecommerce: Where to Start
Not all tests are equal. Starting with high impact areas accelerates results. Starting with low impact areas wastes time and creates false conclusions about CRO effectiveness.
Checkout improvements typically deliver the highest impact because they affect visitors with demonstrated purchase intent. A 10 percent reduction in checkout abandonment directly increases revenue.
Product page improvements affect add to cart rates across the catalogue. Landing page improvements affect the efficiency of paid traffic. Homepage improvements affect navigation but rarely drive direct conversion.
Prioritise based on traffic volume, current drop off rates, and estimated improvement potential. Test where the combination of opportunity and likelihood of impact is highest. Build momentum with early wins before tackling harder problems.
.png)
.png)





.png)