Scurri report 2026
Retailers are ready to give AI control. Shoppers demand the final say.
Retailers are moving AI into the operational heart of eCommerce, with 58% ready to use automated AI to manage delivery strategy and carrier operations. Consumers are more cautious: only 37% are comfortable with AI making decisions that affect their deliveries.
How to read this report
Two studies. One delivery gap.
This report puts the same year to both sides of delivery: the retailers investing in AI, and the consumers living with the results. The findings are most useful read together, because the distance between retailer readiness and consumer conditions is where investment will either pay off or stall.
The most effective use of AI in delivery will combine automation behind the scenes with transparency and choice at the customer interface.
RORory O’ConnorFounder & CEO, Scurri
Key findings
Six numbers that should shape your 2026 delivery roadmap.
The headline shifts from both studies. Select a card to go to the detail.
Retailers are ready to hand delivery to automated AI. Only 37% of shoppers are comfortable letting it decide.
The AI gap 10.6%The average delivery saving retailers expect from AI, but 65% can't cost the exceptions it's meant to fix.
Cost 50→33%Trust in AI with personal data fell as usage rose: the one measure moving the wrong way.
What's new 31%Of shoppers think AI wouldn't improve their delivery at all, the single most common answer.
Priorities mismatch 88/85%Want a human option and clear disclosure. Rising adoption hasn't loosened the guardrails.
The ceiling 60%Would use AI to fix a delivery problem, but only ~16% expect to let it buy. Assist, don't decide.
ROISince 2025: what a year did
Shoppers use AI far more than they did a year ago. They trust it less.
Over the past twelve months adoption climbed sharply and trust fell. The conditions shoppers attach to AI stayed almost exactly where they were.
UK shoppers, 2025 to 2026
Net agreement / likelihood, directional- Want AI disclosed94% → 85%
- Used AI when shopping38% → 65%
- Resolve a delivery problem45% → 60%
- Open to an AI agent49% → 49%
- Trust AI with personal data50% → 33%
Read the crossing
The same shoppers now reach for AI more often and believe in it less. Greater familiarity has brought closer scrutiny, not greater confidence. Three shifts define the year.
Adoption crossed the halfway line
Using AI to shop went from a third of shoppers to two-thirds in a single year. It is now mainstream behaviour.
The conditions didn't move
Rising use has not made shoppers more permissive. They expect disclosure, a human fallback and control just as firmly as they did in 2025.
Trust went the wrong way
Confidence in how AI handles personal data fell by a third. More exposure has made shoppers more wary, not less.
The AI gap: the heart of the report
Retailers and consumers both see a role for AI in delivery, but they want it to solve different problems.
Retailers are ready to automate the delivery operation faster than shoppers are ready to hand over decisions. The retailers that close this gap will start where the benefit is visible, measurable and easy for the customer to control.
Takeaway
Before you scope an AI project, define both the operational problem it solves and the customer benefit it creates. AI that cuts cost without improving the experience confirms what many shoppers already suspect: that it is there to serve the retailer, not them.
Same parcel, two agendas.
Retailers are concentrating AI on operational tasks such as carrier allocation, failed-delivery prevention and integrations. Consumers are more interested in visible service improvements such as tracking, faster refunds and accurate delivery information.
Back end · where retailers point AI
- Checkout
- Order allocation
- Carrier ops
Front end · where shoppers want AI
- Live tracking
- Doorstep
- Returns
The six gaps
Same technology. Different priorities.
The same pattern repeats across AI focus, exceptions, trust, autonomy, delivery priority and Peak: the retailer view and the shopper view are close enough to connect, but different enough to create a gap.
What retailers think
Point AI at back-end automation: 34% would automate carrier allocation and 33% would predict and prevent failed deliveries.What shoppers want
Want visible reassurance at the front end: 29% want real-time tracking and 20% faster refunds. Only 12% choose failed-delivery prevention.What retailers think
Target the operational exceptions: 48% say AI could resolve WISMO queries and 46% incorrect parcel-size rates.What shoppers want
Want proactive communication: 42% want to be told before a parcel is late, and 40% want a revised date.What retailers think
Lean towards efficiency: 20% list AI for efficiency as a priority for the year ahead, against 14% for AI that improves customer experience.What shoppers want
Suspect the cost motive: 64% worry retailers use AI mainly to cut costs, and 88% want a human option before they extend trust.What retailers think
Ready to hand over the operation: 58% are likely to let automated AI run delivery strategy and carrier operations.What shoppers want
Hold back on decisions: only 37% are comfortable with AI making decisions that affect their delivery.What retailers think
Rank delivery mid-table: only 14% name an enhanced delivery experience as a priority for the next 12 months.What shoppers want
Treat delivery as a deal-breaker: 73% would abandon a purchase if suitable delivery options were not available.What retailers think
Brace for stock pressure: 23% name managing inventory across channels as their biggest challenge for the next 12 months.What shoppers want
Expect AI to help when it is busiest: 51% expect retailers to use AI to improve service during periods such as Black Friday.The tension
Efficiency gains matter to the business. Visible control and communication are what shape trust.
The opportunity
Put at least one AI project where the shopper can actually feel the benefit, and make sure it holds up at Peak volumes.
Priorities mismatch: where AI meets the customer
Retailers optimise the operation. Shoppers look for a better delivery experience.
The sharpest divide sits at the customer edge. Retailers are prioritising the machinery behind delivery, while shoppers value the moments they can see and feel.
of shoppers think AI would not improve their delivery at all.
is retailers' most common live AI use today: returns.
Where each side would put AI to work first
% choosing, multi-selectWhat the bars say
All four of the retailers' top choices sit inside the operation. All four of the shoppers' sit at the front of the journey, where they can see them. The two lists barely overlap. Preventing failed deliveries is retailers' second-biggest bet. Only 12% of shoppers rank it near the top of theirs.
The permission ceiling
Shoppers hand AI the problem, not the wallet.
Comfort stays high when AI helps with a defined task, and falls once AI starts making decisions for the customer or spending their money.
The guardrails shoppers set for AI
% agreeingWhere the line falls
Shoppers readily hand AI the work of tracking, comparing and chasing a late parcel. Once AI recommends, chooses or buys, comfort drops below half. The line falls where a decision starts to cost the customer something. Investment on the right side of that line meets little resistance.
Give shoppers these four things, and they'll let AI make the call
% who'd accept AI decisionsof shoppers worry retailers are using AI mainly to reduce costs. Trust will follow only when shoppers can see the benefit in the service they receive.
Cost, permission and ROI
Prove the saving. Earn the permission. Show the return.
With the permission line drawn, the business case comes next. AI creates a stronger case when three things line up: the problem has a measurable cost, the customer still feels in control, and the benefit shows up in revenue as well as efficiency.
Average saving UK enterprise retailers expect from AI-powered carrier management.
have a defined cost for their principal delivery exceptions.
cannot yet put a number on the problems AI is meant to fix.
are investing to tackle all four major exceptions next year.
The four major delivery exceptions
WISMO: “where is my order?”
Incorrect parcel-size rates
International compliance & tax
Incorrect shipping data
The business case
The saving is an expectation. The baseline is the proof. Cost the exceptions first, then use AI where the operational saving and the customer benefit can be seen together.
ROI: proving the investment
The return on AI includes the sale you keep.
The fastest payback comes from pointing AI at what shoppers already reward: visible tracking, honest communication about delays, fast refunds and a human within reach.
What shoppers reward: where the ROI actually comes from
% who value itCost is the top delivery factor
Would accept slower for cheaper at Peak
Free delivery, most valued option
Delivery cost influences the purchase
Would accept a late Peak order if warned first
A human to talk to while waiting
Certainty matters more than speed. Only 44% would pay more for faster delivery, and same-day matters to just 13%. An accurate, guaranteed delivery date is worth more than raw pace.
Peak 2026: the stress test
The business case gets harder when volumes spike.
Black Friday, Cyber Monday and Christmas compress every gap in this report into a few weeks. Volumes rise, promises tighten and a missed delivery date costs more than a single order.
expect to spend more online this Peak than in 2025; 9% expect to spend less
want standard 3–5 day delivery to be free during Peak
would avoid the retailer again if a Peak order arrived late
Takeaway
Use the efficiency AI releases to fund the service improvements shoppers pay for. The return then shows up in loyalty and conversion, which a cost saving on its own will never move. Prove it before Peak: an accurate, proactive warning is what turns a late parcel into an accepted one.
Trust, control and transparency
AI adoption can rise without giving up the human option.
Last year’s report placed transparency and control at the centre of the AI experience. This year’s findings confirm it. Shoppers are using AI more, yet they still want to know when it is acting and to keep a route to a person. Adoption and the human option can grow together.
Where AI is going next
From helping shoppers to acting for them, but only as fast as they allow.
The direction of travel is clear. The pace will be set by shoppers, and by how much permission they grant at each step, rather than by the technology.
would use an AI shopping agent that searches, compares and recommends
are comfortable letting AI buy automatically within a budget they set
Retailers: priorities for the next 12 months
% selecting, up to fiveShoppers: how they expect to shop online a year from now
% of UK shoppersRead the two together
Retailers are building the data, automation and technology foundations for the year ahead. Shoppers are moving more slowly: a third of them use AI to search, compare or advise, but only ~16% expect AI agents to complete purchases within a year. The technology will be ready before the permission is.
The future
Consumers are mastering AI fast. Their expectations arrive before the brand does.
The 2025 report framed the future around seamless, transparent and personalised post-purchase experiences. The 2026 research adds a sharper constraint: autonomy has to be earned through visible value, permission and control.
Takeaway
Invest in assisted AI now and earn the right to autonomy later. Rushing the decision step spends trust that is hard to win back.
Returns: the new battleground
Retailers want returns cheaper. Shoppers want them free and fast.
Returns are the top operational priority retailers are targeting this year, measured against a consumer standard that has not moved. The five gaps play out here in miniature.
Returns: retailer priorities vs what shoppers want
% choosing / valuingAI is the bridge
Home collection still matters to 22%. Here AI can close the gap rather than widen it: nearly half of shoppers would let an AI agent arrange a return, and 50% would let one request a refund. Pointed at the shopper, the automation retailers need and the service shoppers expect work in the same direction.
Returns, redesigned
Make the return easier, not more autonomous.
The 2025 report showed how AI can support returns and post-purchase service. The 2026 findings sharpen the opportunity: automate the admin, speed up refunds and keep the customer in control of the decision.
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.
The retailer shift
Why retailers say they're adopting AI has changed.
A year ago the driver was internal. Now it is the customer, and the growing complexity of the journeys shoppers take.
2025: then
Efficiency first
Business efficiency was the runaway reason to adopt AI, while navigating complex customer journeys sat near the bottom of the list.
2026: now
The customer first
“Navigating increasingly complex customer journeys” is now the top driver, ahead of efficiency. The case for AI has turned outward, towards the shopper.
Why retailers are adopting AI in 2026
% naming each as a top-three reasonCustomer-facing reasons Internal reasons. Two customer-facing reasons now sit at or near the top of the list, level with or ahead of business efficiency. That is the shift: AI is being justified by what the customer experiences, not only by what the operation saves.
The playbook
Six moves for the next twelve months. Earn the right to let AI decide.
The research points to a practical sequence: establish the baseline, target clear operational problems, and give customers transparency and control. Start with visible value, build the controls, then widen the decision boundary.
- Cost the exceptions first
Measure WISMO, mis-rated parcels, compliance and bad data before claiming the saving.
- Improve what customers can see
Prioritise tracking, faster refunds and more accurate delivery dates alongside internal efficiency.
- Make communication a core AI use case
Identify delivery problems early and communicate before the customer has to ask.
- Keep transparency and a human option
Disclose AI, explain what it is doing and keep the route to a person open.
- Don't over-automate the decision
Use AI as an assistant first, with approval built in where the customer is affected.
- Measure the saving and the benefit
Prove the operational return while making the customer-facing improvement visible.
A practical readiness check
Six questions for your next planning session.
Tick each question you can answer with confidence today. The ones left unticked show where AI should earn its place first.
0 of 6 answered with confidence.
The last word
Put AI where it creates operational value, and where customers can see the benefit.
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 need to be careful not to confuse their operational readiness with customer readiness.
Takeaway
The retailers that pull ahead in 2026 will be those that deploy AI where shoppers can see the difference, then use that trust to take on more. Point AI at the customer first, and the cost case becomes far easier to make.
Early access
Be the first to read our next report.
Sign up for early access to Scurri's reports, research and newsletter on delivery and AI.
By signing up you agree to receive emails from Scurri. You can unsubscribe at any time.
The future of AI in eCommerce delivery.
2026 Report