E-commerce has come a long way in the last 10 years. Same day grocery delivery, catalogues spanning hundreds of millions of products, seamless payment gateways, ad retargeting, dynamic repricing and predictive analytics have changed the landscape considerably for both consumers and businesses. As average consumer spend increases and as businesses strive to keep pace with the innovations of the last decade, it helps to ask — how is e-commerce likely to change in the next 10 years?

The future of e-commerce?

In this two-part article series, I would like to give my take on how a few select components of e-commerce are likely to shape-up in the year 2025. In part 1, I will tackle this question from the consumer standpoint, focusing on evolution in delivery time and prices. In each case, I would like to consider two scenarios — expected, i.e., the minimum outcome that we should expect, and ambitious, i.e., the outcome were more ambitious possibilities to materialize.

I. Delivery Time

State of Affairs

Here’s how the delivery process currently works:

Step 1 — Intention: Customer hits the purchase link.

Step 2 — Package Identification: Package to be delivered is identified and marked for shipment.

Step 3 — Warehouse Delivery: Delivery from the origin fulfillment center to the final delivery center.

Step 4 — Last Mile Delivery: Delivery to the customer’s doorstep.

We wait and wait.


Step 2, package identification, will help slash delivery times significantly as e-commerce becomes more ubiquitous. With an increase in e-commerce activity, supply will increase too, increasing the likelihood that a unit of the average product will be available at the customer’s nearest fulfillment centre. This in turn will reduce the average distance that products will have to travel to reach their destinations. Therefore, where previously a customer in California would’ve had to wait for his or her package to be shipped from New York, in the future, chances are that a different unit of the same item will be available in California.

Step 3, warehouse delivery, will gain in efficiency as UPS, Fedex and other logistical networks achieve further economies of scale and routing algorithms improve.


Step 1, intention, is in many ways where the bottleneck lies. The entire delivery process is stalled until the user makes his or her final decision, and everything that comes after is a rush to minimize time. We’ve already seen early signs of anticipatory shipping to kick off the delivery process even before the user makes a purchase. This sort of anticipation of intent could become a fixture as companies develop more complete models of customer personas and interests. This will come in the wake of three separate phenomena:

— Wearable computing will result in more data about our lives and thoughts being available online.

— Data exchange between the various silos in which our online behavior is stored will become more seamless.

— The availability of more data will lead to more accurate predictive models of customer intent.

It wouldn’t be unthinkable that in 10 years, e-commerce companies will be able to predict a sizeable percentage of our actions, even before we’ve decided upon them ourselves.

Step 4, last mile delivery, might be overcome in two critical ways:

Drones: As autonomous delivery technology gets better and laws around their use take form, e-commerce last mile delivery via air is a natural next step forward, as has been widely forecasted.

Coming soon: 60-minute delivery

Uber / Lyft: These companies are likely to look beyond the transportation of humans to the transfer of goods (or the two of them in tandem), especially if growth in their traditional markets begins to taper off. Lower cost and quicker delivery is a natural outcome. We may even come to see innovative models in which consumers are delivered to their goods, or in which goods are rerouted in the last mile to other consumers who’ve placed similar orders.

II. Prices

State of Affairs

The hunt for the lowest prices is currently precisely that, a “hunt”. Presently, once a consumer identifies a product for purchase, he or she is likely to look up that product on a search engine (Google), alternative e-commerce sites (Amazon.com), price aggregators (Google Shopping) or on a product lookup app (RedLaser). The final decision on where the product should be purchased from is driven by a combination of data gathered, gut feeling, and related factors such as delivery time guarantees and customer support. This is far from ideal; once a user chooses a product, all else should be abstracted away and the customer should be confident that he or she has won the best deal. This ideal state of operations doesn’t exist currently because product and price data is stored in non-standardized silos, and efforts to share information across boundaries face daunting challenges.

The first challenge pertains to the manner in which data is collated. Currently, e-commerce information is aggregated either via a pull approach, or a push approach. Services such as Google Shopping rely on merchants pushing their product information to them. The shortcoming in this approach is that not all merchants submit feeds to Google Shopping. More damningly, the feeds that they push aren’t always comprehensive due to technical restrictions or concerns around data sharing.

Companies such as Semantics3 (disclosure: I am one of the co-founders) attempt a pull approach, i.e., send web crawlers out into the open to gather and index information, just as the search engines of today do. While this approach is more universal and relationship agnostic, it has its drawbacks too. Traditional search engines treat web pages largely as blobs of text that can be searched against; an e-commerce centric search engine would have to go the additional step of translating HTML into structured fields such as name, brand, price, color, size, which is a challenge since e-commerce sites seldom follow popular standards.

The second issue pertains to the matching of products once product information has been gathered. Unique identifiers such as UPC/EAN/GTIN cannot be relied upon, since they lack strict enforcement of standards, leading to confusion and misuse. Merchants cannot be blamed because there is often little consensus on what the UPC of a product is. Even when UPCs are well documented, they may not do a good job of distinguishing color/size variations, leading to inaccurate stock availability and price information. One of the key value-adds we offer at Semantics3 is assigning a single “ID” to overcome some of the shortcomings of UPC, but it is hard to say if another standard is the solution to these problems.

More standards?


In 10 years, consumers will undoubtedly have access to better information. I expect that a single authority will emerge as an oracle for e-commerce information, by adopting a hybrid model of the “pull” and “push” approach, along with a certain degree of crowdsourcing. Consumers will learn to trust information provided by this oracle as accurate and thorough, eliminating any buyer’s remorse caused by price or lack of information. This would, in turn, reduce the time gap between the desire to purchase a product and the completion of the purchase.


What if this “oracle” of e-commerce information were to become more than just a more accurate, comprehensive and trustworthy version of the price aggregators of today? What if this oracle were to become so trustworthy that it began to reflect demand and supply, the way stock markets of today do. The ambitious result would be that the markets would have a more direct hand in setting the price of e-commerce goods, which is currently largely the domain of retailers.

Here’s how I see this happening:

1. A private company, with no retail interests of its own, will build a great e-commerce oracle using the “pull” approach, and eventually supplement it with “push” relationships.

2. The oracle will eliminate gaps in information, leading consumers to find the lowest price more often. As a result, companies that offer higher prices with no perceived benefits will suffer.

3. In an attempt to offer the lowest prices, retailers will plug in to the information in this oracle to algorithmically modify their prices, in an attempt to maximize profits. With this feedback loop in place, the oracle will not just eliminate gaps in information, but also begin the influence the prices that it aggregates.

A credible oracle needs to withstand attempts by retailers to unfair advantages, so it will have to remain unbiased and incorruptible. This leads me to hypothesize that oracle may be stored on de-centralized block-chains.

These are just two avenues along which e-commerce is likely to face significant change. Businesses that wish to keep up need to track these avenues and several others to keep pace with changes in the market. In part 2 of this article, I will explore the potential impact of some of these trends on businesses.

The author, Govind Chandrasekhar, is a co-founder of Semantics3, a San Francisco-based startup which curates products and pricing data.