Simon Property Group Monitors Hundreds of Digital Assets
It’s the eve of Black Friday. The promotional emails are piling up as high as the Thanksgiving dishes. Shopping has already begun at many of Simon Property Group’s malls, while others await the stroke of midnight. Somewhere in Indianapolis, Patrick Flanagan, senior vice president of digital marketing and strategy, is contemplating what’s about to happen in the company’s $7 billion brick and mortar world. He’s also hyper-focused on the online world he oversees, which is always open.
Both will be subject to increased traffic and interactions with users and more digital everything—bandwidth demand on in-mall Wi-Fi, social media posts, mobile app usage, mall signage interface, location queries, and transactions from the country’s largest e-commerce gift card program. All touchpoints must be scalable, secure, optimally performing, and highly available every day, but especially at key moments in time—like the biggest shopping day of the year. “I lead the digital marketing efforts across the B2B and B2C portions of our company—simply put, anything external,” says Flanagan. “That falls under three main pillars of web, social, and mobile.”
What’s at stake are numerous consumer and business-facing websites, many microsites, landing pages, three native apps, 1,000-plus on-site digital screens, SMS, MMS, push notifications, and other background services, plus one Facebook page, Instagram account, and Twitter handle for each of Simon’s 225 shopping centers. Integrations with third-party applications like Yelp, Trip Advisor, Waze, Apple Maps, and Foursquare must be functioning too.
This digital bounty is run on its own infrastructure stack by a comparatively modest internal team. In fact, Simon is “surprisingly lean for a Fortune 100,” according to Flanagan, who says the company once won an award for the highest stock cap per employee.
“My team is small, but it’s a very efficient team. We use next to no agencies. Almost everything we do is built in house, including very foundational pieces like our CMS and site engine,” he says. Considering the model, how do they keep digital marketing up and running so well?
For one, “almost everything” doesn’t include website monitoring. “One person can do an awful lot if they’re really well equipped with great tools and partners. Pingdom from SolarWinds is one of those core foundational pieces for us. It enables us to deliver on the lean, efficient staffing model. Pingdom checks 280 different things once a minute, from around the globe and around the clock for us,” Flanagan says.
Doing More With Less
Flanagan describes his function as “full service,” which translates to a hybrid of the typical digital marketing efforts already mentioned and an IT department. A separate, traditional IT infrastructure and team handle corporate applications and business systems.
“We work in tandem with them, but they don’t host, run or manage the marketing sites, services, and apps,” he explains. “Everything our shoppers interact with publicly is powered through an entirely self-managed infrastructure stack. We host a majority of applications in Rackspace and have a number of other cloud add-ons, plus providers for content storage and distribution.”
The environment must be PCI-compliant, so network and system engineers, developers, and database administrators are part of the team who are “deep in the weeds of running a full stack”—i.e., managing complexity and a high-level of scale, speed, and security to handle the retail world, according to Flanagan.
“I’m very much a peer marketer who helps create strategy and messaging, and measure results. But equally important is where the campaigns drive to. What the site looks like, what it does, how fast it is, what it’s built on and how it’s hosted—this is all part of the job for me and my team,” he says.
Even though there’s no room for digital downtime, digital marketing is not the first priority at Simon, which makes Flanagan’s responsibilities all the more complicated. He explains, “At Simon, we’re a physical company first and foremost, making $7 billion primarily from rent. The digital piece is significant, but it’s not our core business. If a website goes down, it’s important. But if our malls close, it’s disastrous.”
With less revenue, comes less budget, so Flanagan required a cost-effective monitoring solution able to grow with the evolving digital era. “Eight or nine years ago we had a far smaller digital footprint,” he says. “But I had a vision of where it would end up going. We needed a solution that would scale with that vision. Also, I’m not a systems engineer, so it had to be easy to install. That’s why we selected Pingdom. It was affordable, reliable, scalable, and turnkey.”
“The site is slow.”
Once in place, SolarWinds® Pingdom® helped define what this vague but inevitable statement means: “The site is slow.” Slow to the user could be connectivity, for example, but the reality on the backend is frequently different.
Flanagan says, “We needed one source of truth. Pingdom measures performance once a minute from many locations at a time, so we can be very precise. If we receive a report of slow, we can get real status immediately. 99.9% of the time a site is performing, and the device, network or browser on the user’s end is at fault.
No human could ever keep track of whether 280 sites, services, tools, APIs, and apps are running, running well or running fast. Pingdom does that and does it really well. Of course, sites will falter and occasionally fail, but the solution notifies us in near real time so we can take action.”
The tool alerts a range of marketing stakeholders, many of whom are non-technical. They don’t necessarily know what’s responsible for the notifications, but understand what to do when they get one—someone in the field, for example, or a member of the gift card team. They’re often engaging directly with the customer or consumer. Knowing status helps them provide better service.
Through visibility from Pingdom and by applying metrics, Flanagan says sustained performance has gone up since implementation. They started at ~95% uptime years ago. Now, 100% is the standard. “That was not a small journey for us. The same applies to speed. We had plenty of sites well north of three seconds per page, and now the vast majority are under three seconds,” he explains.
While stress and load tests precede launches, there’s no certainty until in production. Pingdom helps measure performance changes after launches. Flanagan says, “One of the most gratifying things is seeing the moment of launch and being able to lower the plateau to run at a new faster baseline.”
Pingdom is also helping optimize for future customer and consumer demand. It will be used for more complex transactional checks, getting deeper and deeper into multi-step, multi-page journeys. He sees the perpetual expansion happening in layers.
Simon started with Pingdom on its two main sites. From there, it was applied to secondary, tertiary, and microsites, then to landing pages and partner sites. The next layer involved integrations like APIs, plugins, and widgets. While Simon doesn’t control and can’t fix the thirdparty services it employs, some site features depend on them running—weather, movie ticketing, and map APIs, for instance.
Even though it’s not Simon’s fault if the services are down, to the end user it looks that way. “So, we need to make sure our SaaS providers know something is wrong,” Flanagan points out. “Then we take it one step further to the tools our employees need to do their work.”
He sees the Pingdom value as more than just a question of up or down. It’s about the entire ecosystem in which he works—whether his team can troubleshoot or not. “Even if I can’t put a hand on a server to fix it, I can call. I can use my SLAs to get the response times I negotiated for.”
When things are going well, Pingdom acts as a silent partner. “On good days, I don’t think a lot about it. When things don’t go so well, it’s comforting that I can quickly get what I need and take action. It’s just a great visibility tool,” Flanagan says.