Abandoned baskets in ecommerce: How AI helps recover lost sales
There is hardly another term that has shaped digital commerce in recent years as much as artificial intelligence (AI). While tools such as ChatGPT are on everyone’s lips, the question arises: how extensively is AI really used in ecommerce?
Status quo: lots of potential, but some reservation about usage in day-to-day
A recent study by uptain shows that around 71 % of online retailers surveyed have already experimented with AI tools. However, only around 12 % of shops use artificial intelligence on a daily basis. In most cases, it is only used for specific applications, such as creating text for products or campaign content in marketing.
Most popular areas of AI use: Content before marketing
The most important areas of use in 2025 can be clearly identified:
- Content creation (30 %): automatic generation of product texts, category descriptions and blog articles.
- Marketing (18 %): e.g. for ad texts, email campaigns or social media posts.
- Data analysis & forecasts (14 %): e.g. to determine sales potential or seasonal trends.
- Customer service (13 %): e.g. chatbots or AI-supported FAQ systems.
- Product recommendations (8 %): personalisation based on user data.
Abandoned basket rates remain high: a key problem in online retail
Despite technological advances, one key challenge remains: abandoned baskets. In the second half of 2024, the average abandoned basket rate was 72 % – meaning that almost three quarters of all shopping baskets do not result in a purchase. Mobile users (74 %) are particularly affected, as are customers in price-sensitive or sectors such as erotica, travel or insurance, which require extensive consultation, as well as late evening or night-time sessions when support availability is low and decisions are made more quickly.
A particularly critical point is that the recent average shopping basket value for abandoned purchases is at €58.46 – a sum that can quickly add up over the course of a month. What’s more, the decision to abandon a purchase is being made increasingly quickly. The average session duration before abandonment has fallen to 4 minutes 31 seconds (down 8% on the previous year) – an indication that users are becoming less and less patient and leave quickly if their expectations are not met.
The most common reasons for abandoned purchases
Not every abandoned purchase can be explained by a single reason – usually several factors come into play at the same time. The most common stumbling blocks from the user’s point of view are:
- Unexpected shipping costs: Additional fees that only become visible at checkout lead to a loss of trust and abandoning the basket.
- Missing or unsuitable payment methods: If preferred options such as PayPal, invoice or Klarna are missing, many users will abandon their purchase.
- Complicated checkout: Unnecessarily long processes, e.g. due to many mandatory fields, are conversion killers.
- Uncertainty about data protection, returns or delivery times: Especially with first-time purchases, a lack of transparency causes shoppers to become sceptical.
Two typical dropout personas
- The price-conscious ones: younger, rational, looking for vouchers or hidden additional costs.
- The uncertain ones: older, emotionally driven, need reassurance regarding data protection, returns and payment methods.
AI-supported prevention: From observation to response
Modern AI solutions such as uptain can analyse user behaviour in real time and recognise typical abandonment patterns – such as prolonged hesitation at checkout, scrolling behaviour or mouse movements towards closing tabs. The response is context-sensitive:
Persona profile: The price-conscious shopper
Nina (20) is a strategic online shopper who likes to compare before she buys. She knows what she wants – and at what price. Her behaviour is rational: she researches in advance, compares carefully and pays attention to hidden additional costs. Vouchers play a central role in her purchasing decisions. If there are any surprises at checkout – such as unexpected shipping costs or missing discounts – she abandons the purchase.
When the AI recognises typical abandonment behaviour, it reacts in real time with a measure that is precisely tailored to the situation at hand. For example:
- through personalised exit-intent pop-ups that offer relevant assistance or a suitable incentive
- or through personalised trigger emails with a direct link to the abandoned shopping basket.
These measures are not generic, but context-specific – and thus achieve measurably higher conversion rates.
Practical example
Simon from Dresden has been running an online fashion shop with a partner for three years. Despite his shop’s professional look with high-quality design, strong product photos and over 41,000 visitors per month, he faces a critical problem: many of his customers abandon their purchases at the last moment – mostly because of unexpected shipping costs that only become visible at the end of the checkout process.
His social media campaigns appeal primarily to spontaneous, price-sensitive buyers – but it is precisely these buyers who are particularly sensitive to price barriers. The result: around 3,100 abandoned purchases per month with an average basket value of €35, which corresponds to a potential loss of revenue of over €108,000.
Through targeted activation via a voucher pop-up at checkout, Simon can now directly engage potential abandoners. The discount emotionally offsets the shipping costs – and convinces many to make a purchase. As a result, the shop recovers abandoned shopping baskets worth over €7,700 per month.
7 best practices to prevent abandoned baskets
- Simplified checkout process
Avoid unnecessary mandatory fields and in the checkout settings, only mark fields as “required” that are really necessary. Keep your custom texts in the checkout as short as possible. - Transparent shipping costs
Communicate additional costs early on and use shipping cost calculators or flat-rate models. - Popular payment methods
The more options, the better. At least PayPal, invoice and credit card should be available. - Trust signals (badges, reviews)
Use certifications, customer experiences, and product reviews visibly and strategically. - Personalized triggers (pop-ups, emails)
Use context-sensitive prompts when abandonment is imminent—such as discounts, service information, or reminders.
Conclusion: AI is not a sure-fire success – but it is a powerful lever
The widespread adoption of AI in ecommerce is still in its infancy. Currently, there is often a lack of expertise, resources or suitable tools. However, those who link data-based decisions to situational user behaviour can already achieve measurable results today: lower abandonment rates, higher conversion rates, increased customer satisfaction.
AI is not a cure-all. But when used correctly, it can become a precise tool for solving specific challenges in digital commerce.
Guest author: Harald Neuner / uptain.
Harald Neuner is co-founder of “uptain”, the leading software solution for recovering abandoned shopping baskets in the DACH region. He is particularly keen to provide small and medium-sized online shops with technologies that were previously mainly available to the big players in ecommerce. With “uptain”, he has succeeded in doing just that.
Sources
AI in E-Commerce: Online Stores overlook valuable potential
Focus Cart Abandonment: Semi-annual report 2024
Harald Neuner (uptain)
Harald Neuner ist Co-Founder von “uptain”, der führenden Software-Lösung für die Rückgewinnung von Warenkorbabbrechern im DACH-Raum. Ein besonderes Anliegen ist es ihm, kleinen und mittleren Online-Shops Technologien zur Verfügung zu stellen, über die bisher vorwiegend die Großen im E-Commerce verfügten. Mit “uptain” ist ihm genau das möglich geworden.
This post is also available in: German
