The older I get, the more out of step I become with current trends – and, quite frankly, the less I care about whether I’m in step or not. When I need to buy something, for example, my inclination is almost always to drive to a local store. Increasingly however, the typical consumer is finding ways to shop electronically without ever leaving their home (or office). I do my share of online shopping because the convenience and value can be undeniable. But it isn’t my default the way it is for many people.
 |
| Delivery Vehicle in a No Parking Zone |
The problem is that cities have been built with retail space as a substantial and prominent element of the urban fabric. As e-commerce reshapes the way we shop, it is concurrently reshaping the form of cities. Shopping districts that used to be urban focal points are struggling, suburban malls that once were dynamos of sales tax generation are leaning on businesses that offer experiences instead of products, and many strip centers that used to be the bread and butter of urban shopping are finding vacancies hard to fill. Some retail centers, of course, are doing just fine or even thriving, but there is a general malaise around brick-and-mortar retailing that should be worrisome for city leaders. It might not be quite as dire as the continuing struggles of the office market, but it isn’t far behind and the secondary ripple effects may be more far-reaching.
My goal with this article is to examine the current state of e-commerce, the trends that are shaping its future, and the impact that all of this is having on cities. As with many aspects of our society, change is happening so rapidly that it is straining our ability to adapt. The retail industry has shifted dramatically in the past 20 years but our cities are populated with retail buildings designed to last 50 years or more and are served by public infrastructure designed to last 100 years or more. Cities need a strategy for addressing that mismatch.
The Three Dimensions of E-Commerce
The practice of online shopping has become so ingrained in our lives that it is easy to forget how recently this technology began. The general concept and early pilot implementations have been around since the 1970s, but it didn’t seriously take off until internet access was commonplace, electronic payment methods had been ironed out, and cell phones became ubiquitous. All of that wasn’t really in place until the late 1990s or early 2000s, and even in 2010 e-commerce accounted for just over 4 percent of total retail sales. But in the past 15 years, e-commerce has exploded.
Volume. Aside from recessions, retail sales volume continues to rise steadily over time – up roughly 150 percent over the past 20 years. E-commerce volume, however, has ballooned during that same time frame – up approximately 1,500 percent and now into the trillions of dollars. The rapid increase in online sales can be partially explained by the fact that 20 years ago it was just picking up steam and the volume was relatively low, but if you look at e-commerce sales as a percentage of total sales the trend line is steeply upward.
It is common for a new technology to have a steep growth curve initially and then gradually plateau, but e-commerce shows no signs of slowing down. The proportion of total sales spiked sharply during the early stages of the COVID pandemic and then dropped slightly afterwards as returning to stores felt safe again, but since 2022 the ratio has resumed its upward path. E-commerce currently represents roughly 16 percent of total sales and seems likely to reach 25 percent of total sales within the next 8 to 10 years. The latest indicator? Amazon recently passed Walmart as the nation’s largest retailer.
Breadth. It now seems hard to believe, but when Amazon was founded a little over 30 years ago it was focused almost exclusively on selling books. It turned out that selling books was an extremely logical entry point into the nascent field of e-commerce. Avoiding expensive retail locations provided a price advantage, selling from a warehouse meant that many more titles could be in stock than any bookstore could carry, books were an unambiguous commodity (no one needed to test drive or try on a book), and the ability to read reviews from other customers presaged the addictive rise of social media. It didn’t take long, however, for the company to diversify into music, electronics and other consumer goods. In just six years books were less than half of the company’s revenue and eight years later they weren’t even the largest product category.
Amazon (and its integrated roster of third-party vendors) now sell seemingly every product under the sun. There have always been skeptics, of course, that believed that certain products would always be sold in person and be immune to the impact of e-commerce. That list, however, is getting smaller and smaller. For example, groceries (particularly fresh produce and meat) were thought to be safe from e-commerce but that has changed dramatically. It is estimated that roughly 60 percent of all US households are either current users of e-commerce for groceries or are at least capable of doing so. Walmart – the nation’s largest grocer – has made major inroads into online grocery shopping in recent years. Sign up for a Walmart account and they will give you the option of having your groceries delivered within a couple of hours. For those willing to share the code to their smart-lock or garage door keypad, Walmart offers the slightly creepy option (in my opinion) of coming into your home and putting the groceries away for you.
Buying and selling cars is another product category that seemed immune from e-commerce, but Carvana broke that barrier in 2013. Now there are an assortment of businesses that sell both new and used cars online (including Amazon as of 2024). Virtually every retailer of nearly every type and even modest size now has an online presence in order to survive. Many people still enjoy the experience of in-person shopping, but it is rarely necessary to shop in person if you don’t want to do so.
Speed. The Achilles heel of e-commerce has always been the time it takes for your purchase to be delivered to your door. Sure, in-person shopping takes time as well, particularly if you have to go to multiple stores to find what you want, but it still beats waiting several days for your package to arrive. E-commerce businesses are obviously aware of this issue and have invested billions to streamline the delivery process. Average delivery time across the industry improved from approximately six and a half days in 2020 to just over 4 days by mid-2023. [1] Shipping time is even shorter for the businesses that do the biggest volume of sales – Amazon is now down to an average of less than two days. Delivery speed is a crucial perk for those willing to sign up for membership programs. For example, roughly 75 percent of Amazon shoppers are members of Amazon Prime. U.S. based Prime members received over 8 billion items in 2025 the same or next day, a 30 percent increase over the prior year. [2]
While faster is always better in general terms, things have become a little more nuanced in recent years. In a 2024 survey, the importance of delivery speed was actually slightly below delivery cost and delivery reliability. Ninety percent of consumers are willing to wait two or three days (at least some of the time) if it lets them avoid shipping costs. [1]
The Future of Retailing
You might be wondering why this section isn’t titled “the future of e-commerce.” The reason is that brick-and-mortar retail and e-commerce retail are becoming more and more entwined. Not only are brick-and-mortar businesses rolling out sophisticated online shopping sites, but e-commerce stalwarts such as Amazon, Chewy, Warby Parker, and Wayfair are opening physical stores in selected locations. Amazon is building several 30,000 square-foot “department stores” and may even experiment with much larger “supercenters” similar to Walmart.
The official name for this trend is “omnichannel retailing” and it has been underway for the past 5 or 10 years. The term basically means that every retail business is scrambling to connect with their customers using every conceivable communication channel possible. Don’t want to come into our store? Visit our website! Prefer learning about new products through social media influencers? Then checkout our presence in Facebook Marketplace, Instagram Shops or embedded product tags in Reels and Stories! Even TikTok has an estimated global e-commerce impact of more than $30 billion. Once a haven for small retailers, TikTok Shop now attracts major brands that are reporting huge year-over-year sales increases. [3]
Predictably, the latest “channel” is artificial intelligence. Instead of visiting multiple websites, comparing production selection and price, and then placing an order, simply describe what you need to your AI tool of choice and let it find the best product and the best price. Within a few seconds, AI will have recommendations clearly explained and organized for your review. According to a recent survey, nearly 60 percent of online consumers have used artificial intelligence to help with shopping decisions. [4] AI has become a trusted friend to many people – a friend, it should be mentioned, that has instant access to almost unlimited information. Nearly half of users trust AI more than their human friends to help make fashion decisions.
I decided to test AI as a shopping channel when my old laptop shot craps the other day. I told Google Gemini what I was looking for and asked it to find the best deal. Almost instantly it identified what it thought were the best options and included links that took me directly to the respective websites with the product pre-selected and ready for check-out. I used one of the recommendations for my purchase without searching to see if I could get a better deal. I’m not sure I trust it enough to be a regular user, but the experience was eye-opening. Speaking of trust, some people are reportedly empowering AI agents with payment information so that AI can not only find the best deal, but go ahead with finalizing purchase and delivery details. I am definitely not there yet and may never be, but younger generations are unlikely to have the same reservations.
The rise of artificial intelligence is likely to hurt physical retail stores in two additional ways. First, AI tools are likely to focus shopping decisions on value or fit which will weaken the impact of brand recognition – something that traditional stores tend to rely on heavily. The exception may be at the very high end of the spectrum where value and function are less important than the social statement that is made by wearing a particular brand.
Second, AI tools are going to enable what is often referred to as hyper-personalization. Your favorite AI-bot will know so much about you that its recommendations will almost seem like an extension of your own mind. Virtual shopping software will assist in this process by letting you “see” how clothes, eyeglasses or makeup will look on you. Physical stores will try to replicate this ability with AI-powered kiosks or display panels that supplement sales associates, but for many people the online experience will be too convenient to ignore.
Impact on Cities
To my surprise, the more I looked into e-commerce the more I realized that the impact on cities was not uniformly bad. Online shopping is forcing some uncomfortable transitions, but there are some positives as well which means that the overall impact is a mixed bag.
The rise of warehousing. From a construction standpoint, the big e-commerce winner has been warehousing space. For each one percent rise in the share of retail purchases going online, there is a corresponding increase in warehouse demand of 50 to 70 million square feet. [5] E-commerce fulfillment centers have been popping up all across the country in recent years and continue to be built at a rapid pace.
This isn’t, however, the type of warehouse space that most people are familiar with. A traditional warehouse houses bulk quantities of goods shipped from one business to another. A fulfillment center may receive goods in bulk, but they are quickly broken down into individual units that can quickly be shipped to individuals. Although fulfillment centers come in a variety of sizes to suit a variety of purposes, they typically are very large (500,000 to 1,000,000 square feet), high ceilinged and very automated. Many centers basically operate 24 hours per day, seven days per week.
From a local government standpoint, e-commerce warehousing has both pros and cons. On the positive side, these facilities pay a moderate amount of property tax but demand relatively little in terms of public services (e.g. parks, schools or police services). They often prefer to be located near railroad tracks or interstate highways – sites that generally aren’t great for residential neighborhoods – and they are relatively benign from an environmental standpoint.
On the down side, fulfillment centers are out-of-scale aesthetic disasters that do nothing for the character or vibrancy of the city, and produce little sales tax or secondary economic benefits. No one says “hey, let’s go hang out at the industrial park and watch the UPS trucks come and go.” Fulfillment centers might be a useful part of modern life but they are not what makes a city great.
Somewhat more troubling is a recent trend spurred by the demand for same-day or two-hour delivery service which is leading to mini-warehouses stocked with frequently requested items. These micro-fulfillment centers are often just 5,000 to 15,000 square feet in size and are typically located in the heart of urban centers. They are great for meeting the needs of shoppers desperate for a bottle of Advil, a package of Huggies, or a bag of Purina Dog Chow, but they are a dead zone in terms of pedestrian activity. They typically occupy street level space that might otherwise have been a shop or restaurant, but now is a warehouse with blacked out windows. Fortunately, they tend to gravitate toward Class-B or -C commercial space that might otherwise be vacant for an extended time, but the facilities generate delivery and pick-up traffic that the buildings are ill suited for and which may disrupt other nearby businesses.
Shifting retail space. You would think that e-commerce would be ringing the death knell for retail development but it isn’t quite that simple. Keep in mind that retailers are constantly chasing disposable income and disposable income keeps moving, thus sustaining new retail development. Retail shops particularly covet affluent young families with children, new houses, and unfulfilled ambitions because they want/need to buy lots of stuff on a regular basis. In the past, affluent families congregated in sprawling suburban subdivisions accompanied by new parks, new schools and new shopping centers. Demographic trends (i.e. fewer kids), high interest rates, long commutes and housing affordability issues have put a crimp in that pipeline, but it isn’t empty yet.
Plus, many people still prefer in-person shopping in physical stores (at least part of the time). After all, eighty percent of purchases are not made online. Even though that number is dwindling, it means there is still demand for brick-and-mortar storefronts. Hence, new retail centers are being built, but not as many as a decade or two ago. The shifting retail demand pattern inevitably leaves some shopping centers or retail districts with declining sales, however, and the loss of business to e-commerce platforms simply makes things worse. Particularly hard hit have been retail categories such as electronics, books, apparel and home furnishings.
The perceived k-shaped economy – where the top 20 percent of households thrive and the remainder struggle – has helped luxury brands and value retailers at the expense of mid-range stores. Some chains have either closed many of their locations or gone out of business entirely. The remainder are trying to find some hybrid middle ground that captures the “omnichannel” strategy discussed earlier. Walmart, for example, is pushing hard into e-commerce and is using many of its 4,000+ stores as mini fulfillment centers. Because of its wide geographic coverage, Walmart can deliver to nearly 95 percent of U.S. households in 3 hours or less. Customers now expedite roughly a third of store-fulfilled orders so that they arrive in 1 to 3 hours. [6]
The final shift in brick-and-mortar retailing is an increased emphasis on making the experience of in-person shopping fun. Many centers that were struggling are now jam packed with restaurants, bars, entertainment uses (e.g. arcades, putt-putt golf, etc.) or retail services (e.g. day spas, nail salons, etc.). This shift was partly a desperation move to fill vacancies left by traditional retail stores that drastically cut back locations, but it has morphed into a sustained strategy for keeping rent income flowing and foot traffic high. Fun is a valuable differentiator compared with online shopping. This approach doesn’t work everywhere, however, which means that some older shopping areas have high vacancy rates and no viable turnaround plan.
Delivery traffic. If the previous two sections showed the impact of e-commerce as less destructive than expected, then surely the thousands of delivery vehicles roaming every major city will make up for it by being unabashedly bad. After all, it is a rare day when at least three or four UPS, FedEx or Amazon trucks don’t go down my quiet residential street. But as it turns out, there are pluses and minuses here too and there is actually a surprising lack of definitive studies on the impact of delivery vehicles. I suspect this is because the true answer is “it depends.”
So here is my two-part hypothesis. First, in low-density residential neighborhoods the presence of delivery vehicles may be very visible but the overall impact is probably positive. Keep in mind that the delivery route of a single Amazon truck replaces dozens of shopping trips taken by individual household residents. And in contrast to the relatively random approach to shopping that I and many others take, every part of an e-commerce delivery has been ruthlessly optimized. All of this doesn’t make a great deal of difference because most residential streets are well below capacity in terms of the vehicle trips they carry, but I am reasonably confident that e-commerce ends up reducing the total number of trips in low-density areas.
In addition, most of the large e-commerce businesses and delivery services are shifting to electric vehicles much faster than the general public. Thus, total pollution levels are lower now or at least soon will be.
On the other hand, in dense, urban areas the impact probably flips to mostly negative. There may still be fewer total trips, but other factors come into play that make e-commerce more problematic. To begin with, many household shopping trips in heavily urbanized areas are likely to be completed using walking, biking or transit as the primary transportation mode. Where a car is used for shopping, the trip typically ends in a parking garage or lot where it doesn’t obstruct traffic. While there are some e-commerce deliveries by cargo bike, the vast majority use delivery vehicles and that is problematic for two reasons: (1) they increase total vehicle miles driven on streets that are near capacity, and (2) they have drivers who routinely double park during deliveries which obstructs the flow of traffic.
In a low-density setting a UPS delivery truck can almost always pull over to the curb so that traffic can easily go around and the delivery time frame is measured in seconds. In a high density setting, drivers can almost never find a loading zone that takes them out of traffic and the time it takes to find the correct recipient or package drop-off location in a 30-story building can easily be several minutes. Thus, traffic - which is already bad - is almost always made worse by the e-commerce delivery process.
Cities, of course, can address the problems caused by double-parked delivery vehicles by issuing tickets. Most cities do just that, issuing tickets for double parking by the thousands. Unfortunately, it has not had the desired effect. Drivers have been instructed by their employers to double park anyway so that package delivery times can be kept as low as possible. Paying tickets has become simply a cost of doing business.
A recent study in San Francisco found that the city issued nearly 15,000 parking tickets in 2023 to the top delivery services. [7] No city, of course, wants to deal with the administrative processing costs for 15,000 tickets if it can be avoided, so most have some form of consolidated payment option which allows bulk violators to pay a reduced fine amount in exchange for not contesting any of the tickets. In the end, it is something of a win-win for both the city and the delivery companies but it underscores the fact that the e-commerce delivery system is incompatible with the urban form of most large cities.
The Bottom Line
E-commerce is changing the face of retail which, in turn, presents challenges for the urban form of cities, but it is not a complete disaster. It is simply one more change agent that cities need to adjust for as future zoning and infrastructure decisions are made. There is no magic wand, but I have a few suggestions for minimizing the pain.
Scale back new retail zoning. As discussed above, there will still be a need for new retail development in growing cities, but it will be less than has historically been the case. If e-commerce is siphoning off approximately 15 to 20 percent of retail sales, cities should plan for a roughly equal reduction in new retail square footage. This is particularly true for large scale shopping centers aimed at a community-wide or regional market. These centers have traditionally been supported primarily by department stores and apparel stores which have been particularly hard hit by e-commerce. There will still be some demand but cities should be skeptical of grand plans that seem to harken back to the early 2000s. Cities should be especially leery of developers requesting public subsidies or up-front public infrastructure commitments. Over-zoning for retail development will simply cannibalize existing shopping districts causing premature decline.
Expand loading zones. Everybody loves on-street parking because it is convenient to use and either free or wildly underpriced. But it is arguably the least productive use of the public right-of-way, particularly in dense city-center locations. Instead of parking, cities should block off at least 50 to 60 linear feet of loading zone space on virtually every block face. This will lessen the congestion caused by double-parked delivery vehicles and facilitate the growing utilization of robotaxis and network transportation services (e.g. Uber, et al). This is likely to be very unpopular politically – local businessmen will swear that any reduction in parking will force them into bankruptcy – but cities should at least start a pilot program to gather actual data on the impact on both business and traffic.
Encourage package delivery centers. Cities should strongly encourage any large residential complex or high-rise office building to have a ground floor package delivery center that offers a convenient and secure way for packages to be delivered and stored. Developers are already doing this voluntarily because it is a valuable amenity for tenants, but cities should push to make this practice as universal as possible. Even suburban subdivisions should think about providing a secured package delivery station in order to thwart porch pirates. The technology is already available to make this workable and it would cut down on the number of packages delivered to office locations simply because people don’t want packages left unattended on their doorstep.
Finally, several delivery companies are testing out delivery via unmanned drones, either airborne helicopter-type vehicles or sidewalk based robots. I don’t think any of these delivery devices are ready for prime time and they all have major potential pitfalls. If cities are approached by companies wanting permission for this kind of service, they should just say “no.” Let some other city be the guinea pig until the technology is proven.
Notes:
1. Sandy Gosling, et al; “What do US consumers want from e-commerce deliveries?”; McKinsey and Company; February 2025; https://www.mckinsey.com/industries/logistics/our-insights/what-do-us-consumers-want-from-e-commerce-deliveries
2. “Amazon sets new Prime delivery speed record in 2025”; Amazon News; February 2026; https://www.aboutamazon.com/news/retail/amazon-prime-same-day-next-day-delivery-2025
3. Allison Smith; “Sales from major brands on TikTok Shop nearly doubled in 2025, drawing Ulta and Sally Beauty”; ModernRetail; March 2026; https://www.modernretail.co/technology/sales-from-major-brands-on-tiktok-shop-nearly-doubled-in-2025-drawing-ulta-and-sally-beauty/
4. Caroline Mackey; “Nearly 60 Percent Use AI to Shop – Here’s What That Means for Brands and Buyers”; University of Virginia, Darden School of Business; June 2025; https://news.darden.virginia.edu/2025/06/17/nearly-60-use-ai-to-shop-heres-what-that-means-for-brands-and-buyers/
5. “The E-Commerce Boom Isn’t Over: Implications for Logistics Real Estate”; Prologis; March 2025; https://www.prologis.com/insights-news/research/e-commerce-boom-isnt-over-implications-logistics-real-estate#:~:text=E%2Dcommerce%20penetration%20growth%20translates,of%20net%20absorption%20by%202030.
6. Melissa Repko; “Walmart hikes sales and earnings forecast as it attracts shoppers across incomes”; CNBC; November 2025; https://www.cnbc.com/2025/11/20/walmart-wmt-q3-2026-earnings.html
7. Noah Baustin; “One year, 14,000 tickets: How delivery giants shrug off fines and flout SF’s parking laws”; The San Francisco Standard; July 2024; https://sfstandard.com/2024/07/31/ups-fedex-delivery-parking-tickets-2023/
