Retail Partner Conversations: FenixCommerce improves profitability using AI

After the pandemic’s hyper-growth in ecommerce sales, online retailers and direct-to-consumer (DTC) brands are now having to compete in a higher inflation environment and uncertain economy. One area gaining increasing attention is shipping and delivery—either how to reduce cost or offer a better delivery experience (often both).

In the latest installment of our Retail Partner Conversations blog series, Amazon Web Services (AWS) talked with Akhilesh Srivastava, Founder & CEO, and Don Vangeloff, VP of Customer Success, at FenixCommerce (Fenix). FenixCommerce is an artificial intelligence (AI)-based delivery management software as a service (SaaS) provider helping ecommerce merchants deliver better, every step of the way.

In this blog, they share their views of how applying AI and machine learning (ML) to shipping and delivery helps retailers significantly increase profitability while improving their customers’ shopping experience.

AWS: Help our readers understand why shipping and delivery are so important now, to online retailers and brands. What part does Fenix play in this space?

Don Vangeloff: FenixCommerce provides a robust delivery management platform to help simplify shipping and delivery decisions for online retailers and DTC brands. We use predictive AI-based models incorporating data from a variety of sources, including the retailer’s own, to communicate clear delivery promises from pre-purchase to doorstep.

Delivery matters, 68% of shoppers abandon online carts today because of either unknown arrival times or shipping that is too expensive or slow. Fenix helps retailers solve both problems, thereby lifting ecommerce conversion rates at checkout and reducing their shipping costs.

AWS: So, your platform’s AI-based modeling requires a lot of data, including the retailer’s?

Don Vangeloff: Yes. When clients onboard Fenix, we integrate their inventory-by-location data to determine the best origin to ship from, for each SKU. This is critical for retailers shipping from multiple warehouses, stores or utilizing drop-shipping, as Fenix will route an order to the optimal location that also has all items in an order. To further improve accuracy, Fenix uses operational data including daily order cut-off, pick-pack and carrier pickup times for each available ship-from location to calculate, in real-time, actual estimated delivery dates (EDDs).

This is how Dermalogica gained a 19% increase in checkout conversion and a 92% decrease in split shipments (reducing postage, labor and packaging costs), within 30 days of launch.

AWS: You mention delivery accuracy. In addition to a retailer’s own data, what makes Fenix EDD’s more accurate than everyone else? Especially when carriers already provide delivery dates through real-time API calls?

Akhilesh Srivastava: Fenix utilizes multiple data sources to provide accurate estimated delivery dates, going beyond just the carrier’s committed SLA, or carrier transit time APIs.

First, is client data from our own platform. Fenix accesses current and historical performance data from 100+ different brands, comprising of 3+ million monthly shipments. This includes all data attributes, such as location, weights & dims, and packaging, processed through our platform, providing up-to-date information on delivery times.

Second is carrier-related data, both current and historical. Utilizing real-time carrier transit time APIs is quite common today, yet it’s not enough. With Fenix we take it further by also incorporating historical performance of the carrier(s) for that specific order type, route, and such, into our models, using our deep integration with 100+ carriers.

But carrier transit time APIs do not cover all their services, (including widely used, low-cost hybrids such as UPS SurePost or FedEx Ground Economy) or provide only static zone-based transit times. That’s why Fenix uses a dynamic, data-centric, AI-based approach to predict delivery date promises for each carrier service, including the hybrid services. Our carrier-related data also leverages weather-related APIs to factor real-time conditions impacting delivery times.

Third, Fenix pulls third-party data services, like Nielsen, offering industry-wide historical shipping data. Since this data is not specific to individual retailers, we gain valuable insights into average shipping times based on origin, destination, and service type. This enables Fenix to incorporate broader industry and seasonal trends into our date calculations.

Fenix utilizes a triangulation approach, analyzing and combining all these different data points to create a more comprehensive and reliable delivery prediction for every single order.

Relying on just carrier API data alone, will leave you with 65% accuracy rates overall, on average, for their services. But considering all these data points, Fenix’s predictive AI model provides EDDs with 93-95% accuracy, across all carrier services. With the ability to further fine-tune the desired accuracy level.

AWS: Beyond offering accurate delivery times, pre-purchase, shoppers also want to know the delivery status of their order, once purchased. Does Fenix provide a better ecommerce delivery tracking experience than typically provided by carriers?

Akhilesh Srivastava: Yes, Fenix provides more detailed tracking information, including branded tracking pages. Some solutions claim to present scans when available. Our differentiator lies in our deep access within carriers’ inner data layer, allowing us to present more specific package movements within their network.

By storing and analyzing every scan we receive and comparing it to historical scan data, Fenix tracks and predicts the progress of packages more intelligently. Our model maps various shipment routes and touchpoints for each order, enabling us to identify if a package is stuck in a particular stage and whether it’s experiencing delays beyond the expected delivery date.

However, Fenix’s software also intelligently waits for a few hours to allow the carrier to catch up on any predicted delays; a frequent occurrence. This avoids premature alerts while ensuring customers still receive timely, accurate updates about their orders.

Fenix can provide internal alerts to retailers as soon as our models foresee a delay, enabling them to stay proactive. And that’s the point—providing a better service to our clients and a better experience to their customers.

AWS: You indicated that Fenix also helps retailers lower shipping costs by optimizing carrier selection decisions. Can you explain more please?

Akhilesh Srivastava: Ideally, automated carrier selection, also called Carrier Rate Shopping, uses multi-carrier shipping software to compare shipping rates, service levels, and special offers across different carriers. Then it intelligently selects and displays only the best and cheapest delivery options at checkout. This is a more involved topic that we’re happy to cover in another session.

High-level, Fenix already calculates accurate EDDs pre-purchase. So, we use our same AI-engine to also select the best carrier service that will meet that expected delivery date of that order, at the lowest possible cost, before the customer makes their purchase.

Don Vangeloff: To add, this is significant because it’s this dynamic decision-making capability (instead of traditional static rules) that empowers retailers to offer the lowest possible shipping at checkout, free or paid. Other multi-carrier parcel management solutions focus on selecting lowest cost carrier services at the time of label printing, several steps after the purchase transaction. This does nothing to help retailers incentivize customers to buy at checkout in the first place.

AWS: Any predictions on how retailers will act in this space?

Akhilesh Srivastava: Think about all the money and time spent by retailers in product optimization, trying to find that perfect merchandising mix. The retailers that will thrive in the years ahead will need to expand their focus to take shipping optimization more seriously in their growth plans.

AWS: Thanks for chatting with us, Don and Akhilesh. We appreciate your insights and expertise.

We hope you enjoy our blog series. You can learn more about Fenix through the AWS Partner Network (APN) and by visiting Fenix’s AWS Marketplace Listings.

Contact an AWS Representative to discover how we can help accelerate your business.

Further Reading

● AWS last mile solution for faster delivery, lower costs
● Build conversational experiences for retail order management
● Building a Serverless Event-Driven Retail Order Management System

AWS Partner Spotlight

FenixCommerce uses inventory data, carrier historical transit times, the retailers’ operational performance history, and its proprietary AL algorithms to predict accurate delivery dates and personalized shipping costs on branded, direct-to-consumer ecommerce sites during the pre-purchase stage on the product, cart and checkout pages.