Author picture Phil Leggetter

90% fewer API calls, 3,000% better CX: How Hookdeck took Wineshipping from polling to an event-driven architecture

Published


Wineshipping key facts

Shipping goods is a complex business. To deliver on a two-day shipping promise, logistics companies must bring together multiple players, from e-commerce platforms to last-mile delivery companies. However, those trucks, warehouses, and schedules work in sync only if the platform that coordinates them can receive, process, and deliver software events reliably and at the right moment.

Since 1998, Wineshipping has grown to become the logistics backbone for wine producers across the United States. With ten warehouses and over six million shipments handled annually, Wineshipping relies on Hookdeck to orchestrate millions of events every month, ensuring that each part of their logistics chain works in lockstep.

The challenge

Wine isnt just another product—it's perishable, heavily regulated, and highly valued. This means Wineshipping must meet high standards for accuracy, timing, and compliance, all while coordinating a complex ecosystem of e-commerce platforms, shipping partners, and regulatory bodies.

Each of the company's millions of shipments generates around 10 unique events, such as when a package is packed, quality-checked, or loaded onto a carrier's truck. Historically, Wineshipping managed these events using a homegrown, polling-based platform. However, as the company scaled to process over six million shipments a year, this approach revealed severe inefficiencies.

Pawel Smolarkiewicz, Chief Innovation, and Experience Officer, explained how the inefficiencies grew: “Every partner involved in a shipment was constantly polling other systems—'Have you shipped it yet? Have you shipped it yet?'”

That's because the architecture relied on polling at regular intervals, but when problems occurred, those schedules fell apart. “Even when we asked partners to poll less often, things would get out of sync. To fix it, people just increased the polling frequency.” This approach became unsustainable as the number of shipments and integrations grew.

We were receiving events from every corner—carriers sending updates, wineries pushing order changes, and third-party partners needing tracking information. Trying to handle this with polling meant we were wasting a lot of compute power.

Seasonality further strained the system. “Wine is sensitive to heat, so during summer, orders can sit for weeks waiting for the right weather. But we had systems that were polling every few hours for months at a time, constantly asking, 'Have you shipped it yet?' only to get the same answer: 'No, it's still hot in Texas.'”

And seasonal issues weren't limited to summer. The period from Thanksgiving through to New Year sees massive spikes in demand. “Two years ago, during peak shipping season, systems in the industry became so overloaded that customers were receiving shipment notifications two days after their wine had already shipped,” Pawel said.

The team recognized that what had worked for the company and the industry 15 years ago was no longer fit for their needs.

The solution

Faced with the growing inefficiencies of their polling-based platform and the increasing complexity of their integrations, Wineshipping's engineering team knew it was time for a change. The old system simply couldn't scale to meet the demands of millions of shipments or handle the diverse tech stacks of their partners.

The search for an alternative

When Pawel joined Wineshipping, one of his top priorities was to shift their architecture to an event-driven model. Along with his team, Pawel began by evaluating a variety of alternatives, including options from MuleSoft, Boomi, and Jitterbit. However, although they had a lot to offer, it was clear that they weren't quite aimed at solving the problem Wineshipping faced. Pawel described it as feeling “like taking a steam engine to the grocery store. Webhook management in these solutions seemed like an afterthought—a bolt-on feature that didn't offer the ease of use or reliability we needed.”

The team also considered Azure Service Bus and similar tools on AWS, but they, too, seemed to miss the mark.

We would've needed to spend six months to a year developing internal tools to give our team the visibility and configurability required. We couldn't afford that type of delay.

Finding Hookdeck

When Pawel came across Hookdeck, the platform's event-driven approach seemed like the perfect fit for Wineshipping's needs.

Hookdeck stood out because it was designed to guarantee that no events would be lost, which was a foundational requirement for us. We can deal with delays in event processing if that happens, but missing an event is not an option.

Evaluating Hookdeck

The Wineshipping team started small, testing Hookdeck with a subset of their event traffic. “We began by connecting some of our external carriers and wineries to see how it handled real-time updates. Almost immediately, we saw a difference in performance and reliability,” Pawel said.

One of Hookdeck's standout features during the evaluation was its real-time visibility into the event lifecycle. The team could track each event as it moved through the system, apply filters, and test scenarios in both production and development environments.

Testing an event-driven system is notoriously difficult because of the complexity of simulating realistic scenarios across multiple external partners, like carriers and winery platforms, each with different event formats and schedules. With Hookdeck, we could easily adapt real-world events into our test environment, filter out sensitive data, and replay events as needed—making testing much simpler without compromising security or accuracy.

Having seen the transformational impact Hookdeck had during their evaluation, Pawel and his team made the decision to replace their homegrown platform entirely.

Creating an event-driven architecture with Hookdeck

With Hookdeck, Pawel and his team transformed Wineshipping's architecture from a polling-based system to an event-driven one using the Hookdeck event gateway, addressing three key areas of functionality.

1. Inbound events

Transitioning to an event-driven architecture with Hookdeck has enabled Pawel and his team to address one of their most significant challenges: managing the influx of events from a multitude of external sources.

Under their old polling-based system, not only was this process fragile and inefficient but the team had to maintain multiple integrations to take account of the diverse formats used by their partners.

With Hookdeck handling inbound events in the form of both webhooks and API requests, the Wineshipping team can now focus their engineering efforts on innovation and developing new features, knowing that Hookdeck is managing the ingestion, processing, and routing of inbound events.

We receive events from every corner—carriers sending updates, wineries pushing order changes. Hookdeck allows us to ingest these external events seamlessly, despite the differences in how each partner operates.

2. Outbound events

As orders move through Wineshipping's system, keeping partners updated in real time is crucial. These updates include shipment statuses, delays, and any issues that require attention, ensuring that wineries, e-commerce platforms, and other stakeholders are always in the loop. Previously, handling these outbound notifications diverted engineering effort away from other tasks, as the team had to build and maintain custom integrations for each partner.

With Hookdeck now in place, Wineshipping no longer needs to invest time and resources into those custom-built solutions. The event gateway's flexibility allows them to deliver updates as webhooks with signed payloads or through HTTPS requests with other forms of authentication. This has freed the engineering team to focus on solving customer problems.

3. Service to service communication

Hookdeck isn't only for connecting Wineshipping with its partners. Within their own infrastructure, Wineshipping uses Hookdeck as a reliable way for internal services to communicate asynchronously. This decoupled architecture has improved the resilience of their platform in two ways. First, each component can now operate independently, meaning that updates or maintenance can be performed on individual services without affecting the entire system. Second, Hookdeck's queuing, retries, and rate limiting ensure that even when individual components suffer outages or degraded performance, no data is lost.

Pawel describes it as “a beautiful thing to be able to pause an event feed and know when you restart it that you're not actually going to miss anything.”

The result

Replacing their polling-based system with Hookdeck had an immediate impact on Wineshipping's ability to serve customers efficiently.

90% reduction in unnecessary API calls

Pawel's plan to shift to event-driven architecture involved a complete overhaul of their existing infrastructure.

“Wineshipping has been around for 25 years, and up until 18 months ago, I think it's fair to say we were never known for being a tech-savvy company. But we recognized that to better serve our clients, we had to be more efficient, better at customer service, and more flexible with different workflows.”

The result was the Awesome Client Portal (ACP), a new platform that takes care of every aspect of Wineshipping's consumer experience. And building the ACP was possible only because the team had adopted Hookdeck.

Everything in our portal is connected through events. There's no more internal polling, and Hookdeck, as the event-driven backbone, is a key part of that. Without the reliability and scalability that Hookdeck provides, there's no way we could have even thought about developing the Awesome Client Portal.

This shift has improved efficiency, with a 90% reduction in unnecessary API calls.

If we were still using our old polling system, we would've gotten maybe a tenth as far as we have with Hookdeck over the last two years.

For customers, the result is a 3,000% improvement in Wineshipping's customer service metrics.

Hookdeck accelerates the delivery of new functionality

Hookdeck has transformed Wineshipping's ability to develop and test new features. Before adopting Hookdeck, testing real-world event flows was a major challenge, especially when dealing with sensitive data or complex event chains.

“We faced two key challenges with testing,” Pawel explains. “First, event-driven systems make it very difficult to simulate realistic scenarios. Fortunately, we don't lack real-world data to test with. However, adapting those events into a test environment and stripping out sensitive information was extremely hard with other platforms we evaluated. Hookdeck made that process easy.”

With Hookdeck, the team can now easily filter out sensitive data, throttle traffic, and control the flow of events by pausing, restarting, or replaying them. This level of control has greatly simplified the testing process and accelerated their ability to deliver new features.

We can test more thoroughly and efficiently now. Being able to pause, retry, and replay events has made a complex system manageable, and that's something we simply couldn't do before Hookdeck.

This improved testing process has allowed Wineshipping to iterate faster and confidently roll out new features, knowing their platform is thoroughly tested and ready to scale.

Instant scaling, especially during the holidays

Adopting Hookdeck has eliminated the bottlenecks that once slowed down Wineshipping's operations during peak demand. Previously, their polling-based system struggled to keep up, causing delayed notifications and overwhelmed infrastructure. With Hookdeck's ability to scale seamlessly, Wineshipping can now handle surges in demand without missing a beat.

The future

As Wineshipping continues to scale and refine its operations, Hookdeck remains central to the company's growth. The team is particularly excited about Hookdeck's support for continuous integration and delivery (CI/CD) pipelines, using the Terraform and the Hookdeck Terraform Provider to automate more of their development process and manage releases more smoothly.

For Pawel, the relationship with Hookdeck isn't just about functionality or reliability. What makes the partnership even stronger is how closely Hookdeck has listened to Wineshipping's needs. Pawel notes,

The Hookdeck team has been amazing at listening to us and evolving the platform to support our use cases. Their responsiveness has allowed us to get even more value than we could have hoped.

Wineshipping now serves more customers more efficiently with Hookdeck

By switching to Hookdeck's event-driven architecture, Wineshipping overhauled its entire logistics platform, achieving measurable improvements:

  • 90% reduction in unnecessary API calls, resulting in greater operational efficiency and reduced infrastructure load.
  • 3,000% improvement in customer service performance, ensuring customers receive timely, accurate updates on their orders.
  • Real-time event orchestration, processing millions of events and delivering shipment notifications within minutes.
  • Faster feature rollouts thanks to Hookdeck's flexible testing, throttling, and event replay capabilities.

Wineshipping's engineering team is now able to focus on innovation and customer satisfaction, confident that Hookdeck can scale with their growth and handle peak demand without a hitch.






Try Hookdeck for free or share your EDA requirements and book a demo

The Hookdeck Event Gateway