When Teddy Chan was running an online storefront in 2004, he faced a problem common to early e-commerce entrepreneurs: scaling customer support without drowning in “Where is my order?” emails. Drawing on his background in computer science, Chan wrote a simple program to automate delivery status updates—laying the foundation for what would become AfterShip in 2012.
Today, AfterShip has tracked over 11 billion shipments and expanded into a comprehensive post-purchase suite covering tracking, returns and multi-carrier shipping for global brands. Here, Chan discusses how he navigated two decades of e-commerce evolution, the role of predictive AI in driving customer retention and why machine-to-machine interactions will define the next decade of retail logistics.
Sourcing Journal: What originally inspired you to build AfterShip in 2012, and how has your vision for global e-commerce infrastructure evolved over the last decade?
Teddy Chan: AfterShip started as a fix for my own store. I understood merchants’ pain because it was my pain. I studied computer science. When I graduated in 2004, I started my own online store. I imported MP3 players and gadgets from Asia and sold them on eBay and Amazon worldwide.
Back then there was almost no software for e-commerce sellers. No Shopify yet. We worked hard and were lucky, and the business grew fast. At peak, we shipped around a million orders a year, with fewer than 20 people.
There was no help desk, no shipping tool, no tracking, no returns solution. Everything was manual. Every day we got thousands of emails asking, “Where is my order?” Hiring more people did not scale.
As an engineer, I believed that if a process could be standardized, it could be automated. So, I wrote a program that checked the courier status and told the customer first, before they asked. It cut our manual email work by more than 60 percent. Then I built tools for shipping and returns too. That was the first version of what AfterShip does today.
Customer loyalty went up. More than half of my customers were repeat customers.
My friends who ran online stores kept asking me how I did it. I told them. But year after year, they did not solve it. Most eCommerce founders are great marketers or product people. They are not technical. If people ask you for the same solution for years, there is a gap in the market. So, in 2012, we started AfterShip.
Our first logo was a parcel with a map pin.
Then we learned something. Shoppers did not care where their parcel was; rather, they cared when it would arrive. So, we dropped the map view and built an estimated delivery. We kept the logo to remind ourselves how important it is for us to truly understand the real pain points and needs of shoppers.
On how the vision evolved: I have been in e-commerce for more than 20 years. I watched it move from marketplaces to [direct-to-consumer] sites, from one channel to omnichannel, from online to offline and back, and from local to global. On day one, about 90 percent of our customers shipped domestically only. Today, close to half ship internationally.
Twenty years ago, a seller sold on one or two channels and shipped with one or two carriers. Today, to deliver faster and cheaper, a brand has to pull data from many 3PLs, many carriers and many sales channels into one place. The infrastructure is not just more fragmented. It is more complicated. AfterShip connects it into one view. As the market got more complex, we grew with it, from a tracking tool for small online sellers into a platform for global brands running tracking, returns and shipping across many markets.
SJ: As AfterShip expands its footprint across international markets, what is your primary strategic imperative for maintaining high platform reliability during peak shopping seasons?
T.C.: Peak is when carriers are most stretched. Volume jumps, and exceptions jump with it. For a brand, reliability is not only our uptime; it is also whether the shopper can still trust the delivery date.
AfterShip built for that from day one. We designed the product and the team for a global business. At 20 people, our team already spanned 14 countries.
We are a geeky company—reliability, security and uptime are part of our engineering culture. In 2017, we moved to Google Cloud to support the steep growth in parcel tracking, and Google published a case study on it. During peak season, our team runs 24/7 on-call, so our customers can always get help when they need it.
This peak season, we added AI. AfterShip Agent catches shipment exceptions early and groups them by carrier and issue. One carrier problem becomes one task, not hundreds of tickets. Brands get the capacity to handle their busiest season without hiring for it.
And the delivery promise stays accurate. Dr. Squatch hits 94 percent on-time estimated delivery accuracy across 99.98 percent of orders.
SJ: How are you leveraging predictive AI and machine learning to help enterprise merchants turn post-purchase tracking into a key driver of customer retention?
T.C.: Post-purchase is where a shopper decides whether to buy again. That is why we see it as a growth driver, not a cost center.
AfterShip has been in this industry for 14 years, with more than 1,400 carrier integrations. That gives us more than 11 billion tracked shipments and about 110 billion delivery checkpoints. We turn that data into predictions and decisions for merchants.
One example is the estimated delivery date. Most companies show it after the shopper has paid. We show an accurate date before checkout. The shopper knows when the order will arrive and picks the shipping method that fits their needs. That cuts “where is my order” tickets significantly while also lifting checkout conversion because the shopper is not guessing.
Returns are the other big moment, especially in fashion. A refund ends the relationship. An exchange keeps it. AfterShip Agent reviews return requests with risk assessment and customer history built in. At Naked Wardrobe, return reviews are about 40 percent faster, and 15 to 20 percent of returns now resolve automatically, with no manual review.
The agent recommends, and a person approves anything that touches money or the customer. Brands keep control of key decisions and get time back to work on the business while we automate the rest.
SJ: What do you view as the biggest operational friction point in today’s cross-border logistics, and how is AfterShip positioning its multi-carrier network to solve it?
T.C.: Every shopper wants the order faster. Every merchant wants to fulfill it cheaper. Shipping across borders makes both harder.
A cross-border parcel can pass through several carriers plus customs. Each handoff is a chance for tracking to go dark. Problems show up late, and the shopper often finds out before the brand does. With duties and surcharges already squeezing margins, brands cannot afford that extra cost.
Brands can’t fix what they can’t see. By connecting 1,400-plus carriers into one tracking stream, we give them full visibility, and our AI flags exceptions early so they reach the shopper before the shopper reaches them.
Most brands cannot work that out from their own data. They see their own shipments. They do not see the market. That is why brands come to AfterShip when they expand globally and ask which carrier or 3PL they should use.
We have seen billions of shipments across 14 years. We have a dashboard that shows, for every order, how far it traveled and how long it took, by carrier and by region.
What brands are really solving is inventory placement: the right SKUs, in the right quantity, in the right region, so each order ships the shortest distance. We do not run their warehouses, but we give them the insights to decide where their warehouses should be.
SJ: How do you cultivate a cohesive engineering and product culture across global teams while rapidly scaling AfterShip’s product suite—beyond tracking into returns and multi-carrier shipping?
T.C.: Candidates ask me this in almost every interview.
It starts with hiring. We do not hire for culture fit. We hire for culture add. Everyone who joins AfterShip has a responsibility to add something to our culture, not just fit into it.
Second, I always joke that putting yourself in the customer’s shoes is not enough. You have to feel the merchant’s pain. Many of our team members have run real e-commerce stores. We also send people to follow customers through their day. Our product managers go to the customer’s warehouse and watch the real operation.
Third, we keep one product principle: Simple and Scalable. If a feature is not simple for a small brand and scalable for a large one, we rethink it.
That is also how we grew beyond tracking. Returns and shipping were not new markets we picked from a slide. I built all three for my own store years ago, because the pain did not stop at delivery. Our team keeps following the merchant’s next pain. If our team understands the merchant’s business better than the merchant does, we can continue to grow far beyond tracking and returns.
SJ: Tell us about the future of AfterShip. What happens next?
T.C.: Next year is already here. AI is making most post-purchase operations smarter and more automated. AfterShip Agent handles exceptions and return reviews today, with a human approving every decision that touches money or the customer.
This peak season is the first where an AI agent is a real shopper at scale. Shopping agents will collapse the front end of the buyer journey: discovery, browsing, comparing, checking out. That is where brands have spent the most time and money. When an agent does all of it, the brand barely meets the shopper. Post-purchase becomes the brand’s best chance to build a real relationship with the customer. And merchants will need their post-purchase data to be readable by agents. That is why we ship Model Context Protocol (MCP) servers and connectors for Shopify Sidekick, Claude and ChatGPT.
At the same time, merchants are already tired of agents that promise a lot and deliver little. An agent is only as smart as the data behind it. If it cannot see the order, the carrier, the warehouse and the customer’s history in one place, it guesses. That is why we invested in one unified data stack across tracking, returns and shipping, built on more than 11 billion shipments. There is a lot of agent hype in the market right now. Brands should ask what is inside the agent before they trust it with their customers.
Five years: the shopper’s agent talks to the merchant’s agent. “Where is my order” becomes a machine-to-machine conversation. Humans handle only the exceptions.
Ten years is hard to call, and I will not pretend to know. My bet is that post-purchase becomes fully automated and predictable. The merchant’s job will be to design the experience, not to run it.
Our goal has not changed. We want to serve more than half of the world’s e-commerce brands.