While digital transformation is top of mind for almost every enterprise CEO, few would think a cloud-native B2B eCommerce company would need to launch a digital transformation journey after only 7 years in business. Zoro.com did.
“It’s almost unheard of for a company to need to upgrade so quickly,” says Tim Daneliuk, Zoro’s Director of Architecture. He joined the company in March 2018 to help lead the technical transformation. The issue, he explains, was that the business was such a high-speed success that they were on track to expand from offering about a million products (Stock Keeping Units - SKUs) to offering 5 million by the end of 2020. “To get to that scale and still deliver an exceptional customer experience, Zoro needed flexibility in its architecture.”
When Zoro launched in May 2011, their mission was to deliver an eCommerce site that could support the needs of medium and small industrial buyers. Their strategy was to capture the customer by providing a fast, easy way to procure critical Maintenance, Repair, and Operations (MRO) goods.
From the outset, Zoro understood that these buyers are not typically technology-centric. So the team designed a site focused on the needs of the industrial buyer. “These buyers are busy fixing and running things. They want to find, buy, and get what they need quickly, and move on with their day.”
The site, which debuted in 2011 with 180,000 SKUs, quickly generated sales. Daneliuk attributes its instant popularity to its high-touch customer service, rapid fulfillment, and the fact that it carried hard-to-find items. “Zoro’s vision was to give industrial and business buyers the same low-friction experience that had become the norm for consumer purchases.”
The difference, though, is that Zoro is not a traditional third-party marketplace. The customer’s entire journey is with Zoro. From finding the goods, to ordering, to delivery, through customer support, the customer gets a “Zoro experience”—regardless of who actually ships the products. Goods arrive in Zoro packaging, with Zoro paperwork. Daneliuk notes, “One of our competitive distinctions is that customers go one place for everything. In fact, our customer service organization consistently gets top marks from our buyers.”
From day one, Zoro experienced rapid growth. They continually added new SKUs, gained new customers, and hired more staff. By the end of 2014, they had added a range of traditional office products that customers needed to run their business, and by 2016, the site had hit the 1 million SKU mark.
However, by 2018, the Zoro technology platform started to show stress fractures. “Our system was struggling to keep up with the scale of our increasing sales volume. We became a victim of our own success.” Late in the year, the issue became a front-and-center concern. The company announced plans to create an “Endless Aisle” of goods by offering Zoro’s platform to new partner suppliers. Daneliuk recalls, “We were basically told that we were going to bring in hundreds of new suppliers and millions of new SKUs, and that we needed a platform that would enable, and scale, to accommodate this.” He laughs, “No pressure at all.”
“To make it happen, it was critical to have not only funding, but an executive team that embraces risk,” says Daneliuk. “Luckily, we had both. Our President, Kevin Weadick, told us, ‘Move fast. We know this is hard. You’re going to mess some things up. But if you do, do it quickly, fix it quickly, and move on.’” He says this attitude, and the culture it cultivated at Zoro, was key to his ability to accomplish their first digital transformation project in a mere nine months.
At the time, Zoro’s backend was a monolith that handled front-end eCommerce, financials, and supply chain fulfillment. So, the first problem to tackle was to decouple the guest buying experience from that system. “We set out to build a set of microservices that could fully price and take guest shopper orders independently,” he explains. “We didn’t know it at the time, but all the cool kids were calling our new design ‘Headless eCommerce’—in our minds, we were just making changes to the architecture to improve our customer experience.”
By early 2019, the Zoro team had started an API, microservices, and automation journey. The existing staff was critical to the process because they understood the current system. But Daneliuk realized Zoro also needed additional internal support and external vendors who specialized in APIs and microservices. “I wanted practical architects who could do more than just design stuff. I looked for people with one foot in the implementation and one foot in the design world.” Zoro also looked for vendors that delivered on their promises. “I’ve seen too many consultants come into companies, spend millions of dollars, and leave without delivering real solutions. So we required our potential vendors to do proof of concept as part of the qualification process. They had to take a fairly simple problem and show us a running version. We were a tough sell, I suppose.”
Over the next nine months, the Zoro team took the things they knew about APIs, microservices, automation, and databases and figured out how to bolt them together to work for Zoro. Daneliuk notes that even when the new system was ready, they didn’t just flip the switch and hope for the best. “We knew there would be hardening to do. You rarely hear about hardening, but with a big, complex system like ours, built at this speed, you don’t know what you don’t know. Hardening exposes that.”
To harden the system, he explains, you begin to slowly throw volume and traffic at it. “You start at 10% on the new system and 90% on the old and keep adjusting the amount going to each. This stress testing of the system exposes the flaws in the environment, the things you missed, and the ability of the system to run at scale.”
Daneliuk reinforces multiple times that the commitment of leadership was critical to transformation. “The main reason we were able to do what we did and do it so fast was because of their mindset toward true transformational improvement.” He shares a story that happened as the new system was about to be fully released. “I was sitting with Andy Goodfellow (Zoro’s CTO) when I found out there was a problem that would prevent full 100% traffic flow just one day before it was supposed to go live. Andy notified Zoro’s President, Kevin Weadick, via text and his response was simply, ‘We’re learning.’” Daneliuk shakes his head, “That attitude is extraordinary and incredibly rare. Both Andy and Kevin were true advocates who understood the complexity and risks of this journey and stood by us as the work unfolded.”
It took five months to identify, fix, and fully harden the new system. The results, though, exceeded expectations. “Not only did we solve an immediate problem—decoupling guest buying from the backend systems—more importantly, we created a platform for the future. Now, everything in our next generation has a place to run.” He points out another big win was getting the guest checkout process down to 5 seconds. “That’s the magic number customers notice.” Daneliuk notes, “The system has scaled as expected and, most importantly, protected our customers’ ability to ‘get on with their day’.”
It’s been more than a year since the initial project delivered the first microservices, and Daneliuk recognizes Zoro still has a long way to go on their journey. While he’s proud Zoro solved the guest checkout problem internally, he says they’re not going to try to do the rest of the MACH transformation themselves. “We are going to buy a product, partner with a vendor, and enable the full eCommerce buying experience with our own customer master and headless implementation.”
Even though it was difficult, he believes launching the journey internally was the right choice. “We made a point to keep people across Zoro informed about what we were doing. Ultimately, it spurred a change in thinking that has spread across the organization. Headless, APIs, and microservices are now common in the Zoro vernacular. Today, not only does our guest checkout run on microservices via an API, so does our search abstraction, and some of our supply chain capability.”
While Zoro has yet to recognize the full benefits of MACH-driven composable architecture, Daneliuk feels they are well on their way. “The cool thing is, when we solved the guest checkout problem, we also solved payments and a large part of cart management, tax, and shipping calculations. Our headless journey won’t be easy, but it will be easier because we already have the frameworks in place that do all these things.”
Words of Wisdom
Working as an engineer, architect, and IT exec for more than 30 years, Daneliuk has a few observations which have guided his decisions and shaped his career.
· Never build infrastructure unless you are solving a business problem at the same time. Keep the customer in mind ahead of the “roads and bridges.”
· Never cede control of a project to a third party. The governance of a project belongs to the buyer.
· If you hire well and hire people who know things you don’t, you’ll always be successful. Never settle just to fill an opening.
· Technology is complex and people are complicated. Plan for the ongoing care of both.
Knowledge Comes from Experience
“You get good at something by being bad at it first.” — Tim Daneliuk
1 – Involve stakeholders from every department from the very beginning.
“You may not think you need input from customer service, accounting, operations, etc. until you have something to show them. Not true. We could have avoided some of our problems during the transition and hardened more quickly had we known more about the intricate details of departmental processes and needs. In retrospect, I should have involved all the stakeholders earlier than I did.”
2 – To transform at scale, you need multiple SI vendors with expertise in specific areas.
“It takes different vendors to solve the range of problems found in a system of Zoro’s complexity. For example, we discovered we needed one kind of vendor to handle the microservices and API components, but we needed a different vendor who was an expert in our ERP.”
3 – Focus on the essential technologies that matter most first.
“If something isn’t important, just pick a solution that works and live with it for a while. When we started, we knew we wanted a world-class API gateway. Ultimately, we realized we didn’t need it yet—it was just too early—so we picked an open-source gateway. Now that everything else is running, we have migrated to an enterprise-class API product.”
Finding the Right MACH Match
In their quest to find the expertise and support needed to solve Zoro’s technology problems, the Zoro team met with more than a dozen vendors. “I told the same story over and over to the point where I was sick of hearing it,” Daneliuk laughs. He points to four qualities that can give a vendor the competitive edge.
(1) Show an ability to understand our needs quickly.
“We’re going to tell you our story and where we need to go. If you can’t quickly grasp the problem, then you don’t understand our space.”
(2) Speak and act with credibility.
“Explain what you are good at and how you think you can help. Be ready with a plan for how you’ll work with us, as well as a transition plan and an exit plan.”
(3) Have a partner mentality.
“Promote your willingness to collaborate. Embrace partnering behavior—acknowledge in the beginning you’ll be doing most of the work and we’ll do less, but by the end we’ll do the majority and rely very little on you.”
(4) Be honest.
“Don’t promote skills and capabilities you don’t have. Sell us on your real expertise and be willing to admit it if something isn’t in your wheelhouse. Tell the truth. Be able to say ‘No, we can’t do that.’ That’s how you impress us.”
MACH: Business Technology for 2020 and Beyond
”66% of developers find that maintaining and “paying for” technical debt (aka the money, flexibility, and opportunity lost to a bad technical investment) associated with outdated technology is bad for their productivity at work." State of the Developer report 2109, Devada