Under the trend of SaaS + AI, AfterShip showcases a new paradigm, deeply integrating AI technology into all aspects of the enterprise for systematic exploration.

Author | Dou Dou, Pi Ye

Produced by | Chanyejia

"AOV increased by 45%, total revenue increased by 5%, 60x ROI"

"AOV increased by 70%, total revenue increased by 4.5%, 40x ROI".

— These are two sets of data from AfterShip customers, driven by their AI-powered Personalization product for personalized product recommendations & search.

However, despite such impressive figures, in contrast to many service providers' fanfare around AI, AfterShip, as the "producer," demonstrates more calmness and restraint amidst the AI wave.

For customers using AfterShip's products, they have a distinct feeling: even though they use these products daily, at every moment of their business, they don't realize they're AI-powered. Instead, they intuitively feel more products sold, increasingly accurate delivery times, better user feedback, and ultimately, faster revenue growth reflected in their books.

What is the true value of AI? Is it parameters, technology, a concept, or a hidden productivity boost within products and organizations?

For many SaaS enterprises today, this is a problem difficult to plan and even harder to execute.

For every enterprise, the emergence of new technology presents an opportunity, but whether they can seize it and ride the wave depends on their own courage, accumulation, and investment and attitude towards new technology.

AfterShip is becoming a worthy study case of "SaaS + AI." James Hong, CTO of AfterShip, tells us, "The integration of AI has indeed brought tangible value to us."

From an architectural perspective, this international e-commerce SaaS enterprise now has a mature AI data department. From a business standpoint, AI has become a core direction and strategy within this leading enterprise in its niche, and AI is now a core component of both AfterShip's new and existing products.

How can SaaS enterprises navigate into the AI era? Where are the difficulties, pain points, and anchors for AI to deliver value? How can we leverage the momentum of the old era to sail towards new horizons? In AfterShip, we seek answers.

I. 60x ROI: The Visible Real Demand for AI

"Where is my order?"

This is a question AfterShip's customers frequently ask. To be more specific: Where is my package? When will it arrive?

Delivery prediction is a "standard" feature for domestic e-commerce platforms. But in the overseas e-commerce market, dominated by independent sites, it's a long-standing challenge.

Unlike the comprehensive domestic e-commerce ecosystem, only a few giants like Amazon and Walmart possess this capability overseas. The independent site sector remains a blank slate.

"Before forming the AI team, there were already many similar demands from customers," Loring Liu, VP of Data at AfterShip, told Chanyejia.

In response to this rigid demand, AfterShip launched AEDD (AfterShip Estimated Delivery Date), which uses AI technology to predict delivery times, enhancing customer purchase intent and satisfaction.

One statistic is that it achieves nearly 90% accuracy in its first prediction upon shipment creation. EDD reduces logistical uncertainty before purchase, increasing the likelihood of customers completing their purchases; post-purchase, it optimizes the shopping experience, helping merchants attract more new customers.

Clearly, AEDD is not just a backend logistics tool; it's a brand productivity enhancer and even a sales tool.

Besides EDD, AfterShip also incubated Personalization, an AI-driven product that delivers greater average order value and revenue growth.

Take Dime Beauty, a beauty DTC brand, as an example. Early on, it leveraged social media to influence consumer minds and drive traffic, becoming popular in North America. To better monetize this traffic, Dime Beauty started building a refined brand operation system and considered enhancing customer loyalty and retention through personalized shopping experiences.

Powered by Personalization, Dime Beauty offers 1:1 personalized Product Discovery Journeys. AfterShip's self-developed Universal Recommendations model helps merchants provide tailored shopping experiences. Additionally, AfterShip acts as a strategic advisor, providing insights into industry standards and best practices, helping Dime Beauty understand trends and optimize its business strategies.

Notably, Chinese e-commerce giants like Taobao and JD.com already use advanced data analysis and machine learning to offer highly personalized shopping experiences. However, the overseas e-commerce market lacks such an ecosystem. Based on the independent site model, its overall construction is still in the early stages of personalized recommendations.

Personalization is precisely the product of AfterShip's combination of China's advanced e-commerce experience and its deep insights into the overseas e-commerce ecosystem and customers, truly empowering overseas merchants' business growth with AI.

Under Personalization's multi-touch cross-selling capabilities, Dime Beauty's AOV increased by 45%, total revenue by 5%, and the overall ROI of using Personalization reached 60x.

A clear demonstration is that Personalization achieved real sales growth for Dime Beauty. "For after-sales products, the focus is on cost reduction and efficiency enhancement. While some sub-scenarios can bring GMV to customers, they're relatively limited and indirect compared to pre-sales scenarios," James candidly told Chanyejia.

Overall, AI empowers AfterShip to strengthen its product models, helping customers reduce costs and increase efficiency. It also amplifies AfterShip's value to the sales and transaction sides, addressing not just after-sales service issues but also enhancing customers' transactional capabilities.

II. "Don't Do AI for AI's Sake"

"Our starting point for AI is not to provide the technology to customers in some way. Instead, it's about understanding what our customers need," Harvey, Data Director at AfterShip, emphasizes repeatedly.

This same viewpoint is echoed by Loring. "Our AI capabilities revolve around customer needs, enhancing them based on AI within our existing products. Initially focused on AI for after-sales, we now help enterprises enhance their transactional capabilities."

In 2021, Loring and Harvey joined AfterShip as VP of Data and Data Director, respectively. Prior to their arrival, AfterShip, an international e-commerce SaaS, had accumulated vast amounts of data. They began building AI and data teams, turning this data into assets and monetizing it.

In essence, this lays the foundation for AI to empower business and meet customers' real needs.

A phenomenon observed in past technological revolutions is that whenever a new technology emerges, enterprises rush to integrate it into their products to sell better. But is this really the right approach?

Since the AI boom, SaaS service providers have jumped in, yet there's little customer feedback. Despite their grand AI reinvention efforts, SaaS providers have few notable achievements. Customers buy SaaS+AI products but see little growth... Beyond factors like long technology implementation cycles and maturity levels, we should consider: What kind of AI products do customers need?

In late 2022, ChatGPT was officially launched. The ensuing AIGC wave prompted AfterShip to reconsider the value of AI technology. Soon, a consensus emerged within AfterShip:

"Don't do AI for AI's sake." Rather than blindly pursuing AI integration, it's crucial to understand customers' real needs and direction.

Meanwhile, they focused on building AIGC's infrastructure and data infrastructure (including data asset governance) from the ground up, ensuring they can seize opportunities quickly when they arise. As Harvey puts it, "If you want to do a good job, you must first sharpen your tools and lay a solid foundation."

For AfterShip's internal team, how can we maximize AI's use? This deepens everyone's insights and understanding of AI, which, in turn, drive business and product development.

"Initially, many people may not have a clear idea of what AI can do," Harvey believes. First, they need to spark interest in AI and help everyone better understand it, ensuring everyone is on the same page.

To involve everyone in creating AI products and experience how ideas turn into potentially commercializable products, James led an AIGC hackathon and Hack-Day forum within the company, fostering AI enthusiasm and enabling cross-team, cross-role communication, connections, and co-creation.

The hackathon brought unexpected rewards to AfterShip. "Over 60% of our colleagues signed up for the AIGC hackathon, with over 30 teams shortlisted. Essentially, all departments participated," James noted, adding that AI enthusiasm surpassed expectations.

Additionally, AfterShip built internal platforms like AfterShip ChatGPT and AfterShip GPTs Agent, and procured AI tools like GitHub Copilot, enabling everyone to use AI in their work, boosting efficiency.

Moreover, AfterShip's AI team launched the AI Playground internal experience platform, housing multiple online AI demos, helping employees experience AI capabilities more intuitively.

With this organizational-level AI culture, AfterShip established a unified understanding of AI use, not just within individual work processes but also in cross-departmental collaboration. "Some even explore on their own, discovering things that impress us."

Today, AI-powered smart assistants assist various functions like product development, marketing, sales, and design, enhancing efficiency.

From AI culture building to AI adoption by everyone, to AI empowering organizational efficiency, and finally, to AI driving business. Under James' and the AI data department's leadership, with the ChatGPT wave, AfterShip was the first to complete an inside-out, bottom-up "AI organizational reinvention" while domestic enterprises were still discussing concepts and imagination.

The customer-centric technical philosophy of "Don't do AI for AI's sake" has also been unified across AfterShip's work scenarios that have been efficiently transformed.

III. The "Data Acceleration" Behind the AI Strategy

A thought-provoking question is: What underpins AfterShip's AI strategy implementation? For many domestic SaaS enterprises, despite attempting AI strategies, execution results are underwhelming.

In reality, whether it's EDD or Personalization, these AI-powered products rely on one crucial element: data.

As a hidden champion in its niche, AfterShip has accumulated vast amounts of valuable data over the years. Initially, these data were like scattered pearls, not yet strung into a necklace.

When Loring joined AfterShip in 2021, he faced this situation. His priority was to assemble a team to integrate data and AI, unleashing their potential. From 2021 to 2023, AfterShip invested heavily in building data infrastructure.

"The company divided data asset accumulation and definition into stages, gradually building and improving the data system from the company level to the customer level."

After completing these two systematic steps, AfterShip's data gradually became valuable and asset-like.

Essentially, the continuous accumulation and categorization of data drive AfterShip's product improvements and operational optimizations. For instance, EDD is based on its vast data foundation and repeated model tuning; currently, the AI EDD model is trained using 4.4 billion logistics data points.

Based on this data, AfterShip further launched multiple AI e-commerce solutions, such as Catalog AI, Discovery AI, and Logistics AI.

Specifically, Catalog AI leverages AfterShip's accumulated billions of product data and combines current multi-modal large model technology to build a product knowledge base and industry-wide large model, enabling deep product understanding to enhance product management efficiency.

It automatically categorizes products, generates high-quality product materials like images, making products more attractive and conducive to marketing; it also understands different e-commerce platform specifications, helping sellers ensure product information meets platform requirements for seamless listings.

If an Amazon or independent site merchant with tens of thousands of SKUs wants to expand sales to TikTok Shop, manual listings could take a month. But with Catalog AI, this process can be shortened to three days, significantly boosting operational efficiency.

Discovery AI benefits from Catalog AI's strong product understanding and cross-channel user behavior tracking, accurately extracting user interests and preferences to provide more precise product recommendations, ideal for D2C merchants focusing on off-site traffic and product testing.

Discovery AI can also be applied to scenarios like search, enabling precise matches between users and products; through personalized product distribution, it improves marketing conversion rates, driving more potential customers to purchase.

According to AfterShip's publicly released data, Discovery AI can help sellers increase conversion rates by an average of 30% and AOV by 10%.

Logistics AI, besides collecting logistics company data, also actively acquires weather, traffic, and social event data to provide comprehensive logistics information. By cleaning and parsing raw data, Logistics AI leverages standardized data to provide accurate delivery time predictions, helping sellers plan logistics operations in advance, enhancing service responsiveness and customer satisfaction.

These product rollouts further propel AfterShip in satisfying customer needs.

More importantly, at AfterShip, data isn't just for technicians; it concerns everyone. By fostering a culture where data informs every decision, as Loring says, "We aim to establish a top-down culture of data-driven decision-making."

From data-driven organizations to intelligent products, AfterShip's underlying logic lies in optimizing business processes through data-driven decisions and AI technology application. This logic is evident not just internally but also in every updated product and service process.

Whether it's years of accumulated valuable data, early platform-based technical strategies, or a more open and inclusive organizational culture, these are the core reasons why AfterShip can rapidly build a data system and unleash AI's productivity value.

Closing Thoughts:

In the AI wave, SaaS enterprises face the challenge of deeply integrating AI technology with their business. However, there's no definitive solution to this challenge yet.

But through AfterShip's practices, we're seeing some answers.

For instance, its open and innovative organizational culture encourages employees to embrace AI-driven changes and leverage new technologies to enhance internal efficiency and creativity. It also consolidates infrastructure, establishing rigorous data governance mechanisms to ensure data accuracy, completeness, and security. Furthermore, its demand-first product philosophy better meets customer needs with AI.

Under the SaaS + AI trend, AfterShip showcases a new paradigm: deeply integrating AI technology into all aspects of the enterprise for systematic exploration. This exploration, from organizational culture empowerment to internal efficiency enhancements, infrastructure building, to "not doing AI for AI's sake," is grounded and systematic. It focuses not just on technology but also on how it serves the enterprise's long-term development.

AfterShip's AI exploration and practical experiences demonstrate that customer-oriented, seeking scenarios with real value for customers, is the key to integrating AI with SaaS enterprises and truly driving business growth. As James says, "We're constantly seeking 'new scenarios' to create real value for customers, whether by boosting revenue, enhancing efficiency, or reducing costs. Identifying these new scenarios and leveraging our data will unleash even greater value."