PostNL's parcel service is growing rapidly each year. IT plays a critical role when it comes to scalability. This involves technology, people, and processes.
"Deliveries stand still if the IT goes down," says Gerrie de Jonge, CIO for Parcels and Logistics at PostNL. "What's more, all the software that we use in the stack, with one exception, has been built to our specifications. That was necessary because of the high volumes, the intricacy of the process, and integration with third parties such as web shops. The result is that we are in fact now almost a technology company."
As part of a policy for sustainable growth and development, PostNL had already taken steps relating to logistical infrastructure and personnel policy. "Change was needed in IT. Three years ago, when I was appointed, I found a fragmented application landscape. We're in the process of transforming that, step by step, into a more logical structure. The paradox is that, if you make the systems as easy to change as possible, you get stability as a bonus."
"On the operations side, each application supplier has provided its own application," Gerrie de Jonge continued. "That was an appropriate choice in the initial phase, when we were more like a startup, but it's less suited to the kind of company we had become. Our growth means that we are operating almost continuously with near real-time information flows to the customer. At the same time we continuously expand our service portfolio with for example evening service and fresh food deliveries. As a result, recovery times needed to be as short as possible, which meant that the suppliers' stacks had to be upgraded to a mission-critical regime."
In September 2016, Gerrie de Jonge contacted Schuberg Philis for advice and support. Then, when the company suffered an IT failure at a pivotal moment in the holiday season, the top management was more than open to the promise of 100% functional availability. The first assignment was to make a scan of the IT landscape. "Gerrie wanted to hear the whole truth from us," says Peter Siepel, Customer Director for Schuberg Philis. "They didn't want a good news story."
"During the scan we looked at each application and we asked ourselves whether we dared to go live with that application ourselves, given that we give a 100% uptime guarantee," says Mark Reijn, Lead Engineer for Schuberg Philis.
"PostNL's infrastructure was already on Amazon Web Services (AWS), and they wanted to keep that. We assessed critical points such as monitoring, availability, scalability, and security. That gave us an estimate of the quality and stability of each application." In April 2017, work started on an 'AWS landing zone' for the existing and future application landscape. Gerrie de Jonge: "Nowadays, holidays begin on Black Friday – and there was Cyber Monday in late November (when major retailers start their sales online at midnight). That was our deadline. We got through without any major upsets and delivered record breaking number of parcels without any major disruption. It was a tremendous result."
According to Gerrie de Jonge, the implementation would not have succeeded with any other supplier. "Especially not at the point in the summer when we hit a rough patch, mainly because of project dependencies and the need for different ways of working involving multiple parties. But the all-too-familiar disputes about price, scope, quality, and timing never arose. Behind the scenes, they scaled the number of people up from nine to sixteen or seventeen people without a hiccup. That's also down to the quality of our own team, but it's quite an achievement."
Schuberg Philis acts mainly as integrator and orchestrator. Peter Siepel: "We work bottom-up, from the level of the AWS cloud infrastructure. By creating uniformity, we ensure that every application runs within the same structure and parameters. There's a uniform toolset that we, working with the software suppliers within the ecosystem, constantly update and develop. Thanks to stack monitoring it's clear at the process level which application and supplier is responsible for each element. That means that problems can be dealt with more quickly."
Gerrie de Jonge got his wish to have 'all the systems in one room' before the festive season, and with all the necessary expertise. Close cooperation led to synergy and shared responsibility. "That meant that we could respond quickly and effectively to any possible incidents in all logistic applications, including those that were not the responsibility of Schuberg Philis."
PostNL's vision and approach to its working method and governance were also a good match with those of Schuberg Philis. "Our involvement and contribution has been cultural, as well as technological," says Mark Reijn. "It was more about facilitating leadership than issuing instructions as well as giving the teams end-to-end responsibility. Gerrie de Jonge: "We've always had DevOps as our overall vision and that's been strengthened by what we've achieved with Schuberg Philis. We've been through some pretty tense times together, delivering good results, and together we've established a good foundation for the future."