New Scurri research reveals a gap between retailers’ AI ambitions and consumers’ expectations of transparency, control and human support
More than half of UK enterprise retailers are prepared to use artificial intelligence to manage ecommerce delivery strategy and carrier operations, but consumers are not yet ready to hand over the same degree of control.
According to new research from delivery management platform Scurri, which surveyed more than 1,000 UK consumers and 150 UK enterprise retailers, 58% of UK enterprise retailers are likely to use an automated AI system to manage delivery strategy and carrier operations. However, only 37% of shoppers are comfortable with AI making decisions that affect their deliveries.
A further 38% actively disagree that they would be comfortable with this, while 25% remain undecided, highlighting a potential trust gap as retailers accelerate their adoption of AI.
The research suggests that consumers are not rejecting AI itself. Six in ten would be likely to use an AI agent capable of automatically resolving a delivery problem. Instead, their willingness to use AI appears closely linked to whether they remain informed, retain control and can reach a person when necessary.
An overwhelming 88% of consumers believe they should always have the option to speak to a human, while 85% expect retailers to make it clear when they are interacting with AI. Consumers said they would feel more comfortable allowing an AI agent to act on their behalf if they could switch to a human at any point (34%), the AI always sought permission before acting (33%), they could immediately override its decision (29%) and they could see exactly what the AI had done (24%).
Acceptance also rises when AI is given a clearly defined and practical role. 64% of consumers would allow an AI agent to track a delivery, while 51% would be comfortable allowing it to resolve a late or failed delivery.
Rory O’Connor, founder and CEO of Scurri, said: “There is a dividing line between AI that removes friction and AI that removes control. Consumers are willing to let AI track parcels and resolve straightforward delivery problems because these are clearly defined tasks with an obvious benefit. Resistance grows when AI begins making decisions without permission, explanation or an accessible route to a human being. Retailers therefore need to be careful not to confuse their operational readiness with customer readiness.”
The disconnect is also evident in where each group sees AI adding value. Retailers’ leading applications are largely operational, while consumers prioritise visible improvements to their own experience, particularly real-time tracking, faster refunds, accurate delivery dates and earlier notification when something goes wrong. Retailers expect AI to make its greatest operational contribution by automating order allocation to final mile carriers, cited by 34%, predicting and preventing failed deliveries and managing technical integrations, both cited by 33%, and monitoring delivery problems and carrier performance (31%).
These applications could enable retailers to process large volumes of delivery data, identify emerging problems and select the most appropriate carrier or service more quickly than manual systems. But Scurri warns that the customer facing consequences of those decisions must remain visible and reversible. O’Connor said, “The most effective use of AI in delivery will combine automation behind the scenes with transparency and choice at the customer interface.
“AI might identify that a parcel is likely to miss its promised delivery window, select an alternative service or propose a new delivery option but the customer should know what is happening, understand why and retain the ability to approve, change or reject the proposed action. The danger is that retailers deploy AI to eliminate internal friction while unintentionally creating a new source of customer frustration.”
