The Future of Retail eCommerce is Fixing the Last Mile

Where the next few years will take retail ecommerce, and why the doorstep is no longer the only finish line.

It’s perhaps a little cliché to say that retail ecommerce is changing. The more helpful question is which part? For the majority of the last decade the answer was discovery and payment, from how people found products to the way in which they paid for them. The area that got little attention was what happened after orders were placed.

So it’s no surprise that it’s now facing greater scrutiny, with the next few years likely to define its role.

TL;DR

• • Retail ecommerce has spent a decade fixing discovery and payment, leaving the last mile as the stage where the economics are hardest to manage.
• • The global out-of-home delivery market is on course to roughly double to $54 billion by 2030, with Europe leading and Germany in the middle of a build-out that will see its locker network triple inside the decade.
• • The economic case is already visible in the €800 million Germany loses every year to failed first-attempt deliveries.
• • HubBox research shows the checkout map itself does measurable work, with branded pins lifting immediate shopper comprehension from 12.5% to 80%.
• • Pickup adoption is increasingly shaped by what happens inside the checkout itself, with frictionless checkout paths removing the option for a large share of shoppers.
• • AI shopping agents are emerging as a new class of buyer that will favour delivery options built for reliability over ones designed for habit.
• • The doorstep is no longer the only place a parcel can land, a change happening faster than most retailers have noticed.

What the headline numbers tell us about the future of retail ecommerce

Global retail ecommerce is forecast to reach roughly $6.88 trillion in 2026, up around 7.2% on the year before. It will account for approximately 21% of all retail sales worldwide, which is quite the headline number. 

Online buying is also no longer a channel sitting alongside the physical store. For a growing share of categories it is the shop. But growth at such a scale puts new pressure on the parts of the journey that weren't designed for it, and it’s the last mile that feels the most pressure. 

The growing cost of home delivery

Home delivery represents the stage where the economics are hardest to manage. Industry estimates put last-mile costs at over 40% of total shipping spend, and the structural reasons are well understood. An empty house turns one delivery into two, with WISMO enquiries piling up in customer service in between. Each redelivery effectively doubles the cost of an already complex operation. Retailers have spent years optimising warehouses and checkout flows, but the doorstep itself remains the part of the journey carriers and retailers have the least control over.

Then there’s the cost that you don’t see on a logistics dashboard. In the United States, package theft has grown into a structural problem, with SafeWise’s 2025 report estimating that porch piracy cost American consumers around $15 billion over the year. That’s more than 104 million packages reported stolen and close to 250,000 a day. Factor in the various order remediation costs that retailers absorb, and the combined economic hit climbs past $37 billion. A system that loses a quarter of a million parcels a day to the doorstep isn’t one anyone would design on purpose.

That’s the backdrop against which out-of-home delivery has moved from a European curiosity to a serious answer in the States. Instead of leaving a parcel on a step, the shopper collects it from an alternative pickup location at a time that suits them. It’s a more reliable and cost-effective delivery method that also significantly reduces the risk of theft.

But it’s still a solution in the US. In Europe, however, it’s a reality. 

Out-of-home delivery is the infrastructure layer no one noticed being built

The growth is far from speculative. Mordor Intelligence values the global out-of-home delivery market at around $28.8 billion in 2025 and projects it will roughly double to $54 billion by 2030, a compound annual growth rate of 13.4%. 

Europe leads the way

Europe is the fastest-moving region and the largest today. Since 2023, the number of automated parcel machines across Europe has risen 29% to nearly 155,000 units, while staffed pickup-and-drop-off points have grown to over 349,000 locations.

In some markets, like Poland, the future has already arrived. Half of all parcels now move through out-of-home points, with lockers the dominant method. Meanwhile, Finland and Sweden are close behind. 

Germany plans to double its locker network to 30,000 units by 2030. It's a familiar pattern – what looks like emerging infrastructure in one country tends to be unavoidable a few years later in the next.

What’s striking, however, is just how quickly consumer preference follows availability rather than the other way round.

HubBox is the checkout software that lets retailers offer pickup points and lockers as a delivery option, and sits squarely in this shift. Our research helps explain why the model works once it’s in front of shoppers, and why the United States, despite its scale, is still earlier in the curve than Europe.

Germany is a paradox the next few years will resolve

Germany is one of the markets where the next phase of OOH will play out most visibly, and the dynamics there are worth understanding in their own right.

It sends more parcels than almost anywhere in Europe (over four billion in 2024), and is one of the highest per-capita volumes on the continent. It also has one of the most mature ecommerce ecosystems in the world. And yet, by the measure that’s most important for out-of-home delivery, it’s still catching up. There are around 15,500 DHL Packstations in operation, which works out at roughly 0.18 lockers per 1,000 people. Poland sits at a multiple of that, and the Nordics further still.

The discrepancy between parcel volume and locker density is what’s likely to close over the next five years. 

• DHL has committed to doubling its Packstation network to 30,000 units by 2030. 
• DPD and GLS have jointly committed to scaling from around 100 shared stations today to 3,000 by 2027. 
• Myflexbox is planning to grow from roughly 460 sites to at least 5,000 by 2030
• DHL's new DeinFach and OneStopBox networks are pushing carrier-neutral lockers into the same conversation. 

Taken together, the German out-of-home network is on course to roughly triple inside the decade.

The economic case for retailers is already there in plain numbers. Failed first-attempt deliveries in Germany cost the logistics industry an estimated €800 million a year. Every parcel that succeeds at a locker rather than a doorstep takes a slice out of that. Lockers cut last-mile CO₂ emissions by between 13% and 32% per parcel on most studies, and up to two-thirds in the densest networks. It's a meaningful figure not to be sniffed at in a market where emissions regulation is tightening faster than almost anywhere else.

What makes Germany a distinct case is the cultural fit, given that the country already runs on dense, well-trusted physical infrastructure. Shoppers are comfortable collecting items as part of their planned errands, often choosing the same locations repeatedly. The barrier in Germany was never going to be perception. It was density. But as density resolves, adoption follows almost mechanically.

Why the checkout interface is just as important as the network

It’s tempting to think that the amount of pickup points is all that matters. Yet, the research says the experience of choosing one is just as important. HubBox's user research found that 72% of shoppers begin on the map to orient themselves spatially before they pick anything, treating it as the way they understand a location rather than treat it as a decorative feature.

The quality of that map does measurable work, and only 12.5% of users grasped their options immediately when pins were generic. Switching to branded pins carrying the retailer's logo lifted immediate comprehension to 80%, close to a sevenfold improvement from a change most teams would consider cosmetic. 

Distance visibility had a near-universal engagement rate (95%) because a shopper deciding whether pickup is worth it is really asking one question. Is it close enough to bother? A good map answers that before they even have to ask.

Beyond helping shoppers choose a location, a well-built map also makes the trip itself feel worthwhile, which is the effect most often overlooked. Shoppers described using it to bundle a parcel collection into errands they were already running, turning a delivery into part of a journey. In a German market where customers already think in terms of efficient, combined trips, it’s a powerful mechanic. It’s the difference between a pickup point being tolerated and being preferred.

The speed of buyer journey is a problem that needs solving

Before getting to where ecommerce is heading, it's worth looking at where it already is. The shopper journey has been accelerating for years, and the stages retailers have optimised the most tend to be the ones that bypass delivery choice altogether.

Express payment options like Apple Pay, Google Pay and PayPal are the clearest example. They're now standard at checkout across most major retailers, and around 40% of customers use them, rising to 70% in some markets. They’re also one of the easiest places to accidentally remove pickup from the journey. One-click payment flows reset the delivery address back to the billing address, which defaults to home. 

Out-of-home options have to be reinserted after the fact, with most self-built OOH integrations not accounting for it. The result is a third of a retailer's customers never seeing pickup as an option, even when it exists elsewhere on the same checkout.

Native apps create a similar blind spot, as the app is now the primary shopping surface for most enterprise retailers. Over half of transactions for brands with apps come through them rather than mobile browsers. Beyond that, around 75% of online shoppers say they prefer apps to browser shopping. If a checkout solution only works on desktop or browser, it's invisible to more than half the customer base by default.

The third place pickup disappears is the logged-in journey. Returning customers expect the checkout to remember them, and ecommerce platforms reward that expectation by collapsing as many steps as possible,  including, often, the delivery selection itself. A logged-in shopper with a saved address can complete checkout without ever being shown a pickup option.

None of these are failures of the OOH idea, as such. But they are the consequences of checkout getting faster, and they're only solved by building pickup into the fast path as opposed to alongside it. Pickup adoption increases when the option lives inside the fast paths shoppers use. It's why HubBox's integrations are built to surface pickup inside express payments, native apps and logged-in journeys rather than around them.

What happens when AI agents arrive?

All of this is happening just as the front end of commerce begins to change shape again. Through 2025 and into 2026, the large AI platforms moved from helping people search to helping them buy. Google introduced agentic checkout across Search and Gemini with a buy-for-me capability live in selected retailers, OpenAI added shopping features to ChatGPT, and Adobe reported that AI-driven visits to retail sites grew by over 4,700% year on year. McKinsey has suggested agentic models could redirect $3 to $5 trillion in global retail spend by 2030. The forecasts are large enough to treat cautiously, but the direction isn't really in doubt.

Autonomous purchasing is still  at an early stage, and most of what works today centres on discovery and comparison over the agent completing the whole transaction unsupervised. Shoppers are happy enough to let an agent help them choose while remaining wary of letting it spend on their behalf. The connection often overlooked is the physical aspect. An agent can place an order, but it can't carry the parcel to a door. A new class of automated buyer is arriving, one that has every reason to prefer the delivery method built for reliability.

What the next few years will tell us

Put the pieces together and a picture forms, one where online sales keep growing and the pressure keeps loading onto a last mile that's already the hardest part of the economics to manage. Out-of-home networks keep expanding at double-digit rates, and in Germany the network is set to triple by the end of the decade. Where shopper resistance exists, it's driven by density and interface, and both are improving fast.

Treating pickup as a checkbox bolted onto checkout is easy. The harder job, historically, has been keeping it in front of shoppers as checkout gets faster. HubBox's integrations solve the issue by putting pickup inside the fast paths shoppers already use, in express payments, native apps and logged-in journeys. For a long time the doorstep was the natural end of the journey. It's an assumption that's coming undone.