Blog

Predictive analytics helped us weather Hurricane Ian

by Everstream Team | October 5, 2022

The sandbag manager was paying close attention.

His retail company’s entire executive team* was huddled on a crisis call, listening to an expert meteorologist explain where and when Hurricane Ian would hit the next afternoon. The company had been tracking the storm for a week, and now they snapped into action.

Store by store, the meteorologist predicted which of the retailer’s hundreds of locations in Florida would get winds strong enough to break windows, which would face dangerous storm surge flooding, and which would have no issues.

Minutes counted. The sandbag manager took careful notes. They were ready. This is the story of the eight days before Hurricane Ian, and how one company secured enough lead time to weather the storm.

Tuesday, September 20

Nine days before Ian landed, a supply-chain weather specialist at Everstream Analytics sat looking at his three giant computer monitors. The spiraling wind pattern he had been watching in the Atlantic Ocean was the kind of thing that meteorologists get excited about. Not because of the drama of impending damage, but for helping people prepare. The cluster of wind spinning away from the African coast had all the warning signs that this applied meteorologist had spent a career spotting and warning people about.

The path tracked across areas of the Caribbean that tend to feed big storms, and atmospheric conditions in those zones were rich – low shear and warm water. After decades of watching storm tracks, the Weather Expert knew this was a swirl you wanted to watch.

It was time for the Everstream Analytics team to shift into Storm Mode.

Key members of Applied Meteorology and Intelligence Solutions cancelled their regular meetings and started writing briefing notes. And drinking coffee.

From locations in Spain, Germany, Minnesota, Chicago, and more, the Storm Mode team steeled themselves for some very long days briefing clients from Singapore to Silicon Valley. Retailers, pharmaceutical companies, food and beverage, and other Everstream clients with supply chain connections in Florida would need to know all the risk factors – both upstream and downstream – as soon as possible.

Although the Storm Mode team couldn’t yet predict the exact track or ferocity, they were confident that Ian was most likely racing towards Florida. Every day that clients in those areas had to prepare would lower their risk and keep employees and stores safer.

Wednesday, September 21

When the sandbag manager got the first Watchtower Alert email from Everstream warning of tropical storm Ian’s potential, he read it closely. With hundreds of retail stores in Florida, his company was already too experienced with hurricane preparedness and recovery.

Employee safety was the top concern, but which employees were at risk? And for what? Where would winds be strongest, and at what time? Where would storm surges lead to flooding, and when? Knowing the key details far enough in advance to predict specific risks at certain stores could save lives and prevent damage.

What the sandbag manager wanted to avoid was boarding up windows at a store location that experienced flooding instead of high winds. Or closing a store an hour too late, exposing commuting employees to flooded roads.

He sent the Watchtower email to the rest of the retail company’s crisis team, including the CEO. “We need to keep an eye on this over the next few days,” he said.

Sunday, September 25

The retailer’s Crisis Center Manager sent the email Sunday evening, which was early Monday afternoon in Europe for Everstream’s meteorologist. The California-based retail client was concerned about the escalating chance that Ian would strengthen and aim toward their Florida locations. Did Everstream think it was appropriate to convene the crisis team?

“Absolutely,” the meteorologist said. He had known for almost a week that Hurricane Ian had “crisis” written all over it. Especially for this client, who had so many stores based in the storm’s path.

After working with this client for years, Everstream’s team knew the information to gather: Which stores had the highest risk of wind damage and flooding. Managers would use that information to drive operational decisions like where to send teams to secure cash registers, where to board up windows, where and when to send home employees, and where to stage sandbags.

Monday, September 26

As the Everstream weather team studied the models, they decided that Ian was moving to the west coast of Florida. Scanning the retailer’s store network, the team mapped each store to evaluate levels of different types of weather-related risk. The team synthesized multiple data streams to give a ranking of 1-25 for each risk at each store.

They were creating similar reports for 20 other clients as well, all eager for recommendations to help them protect operations. The media had also caught wind of the team’s insights, with print and broadcast news outlets calling for interviews.

Behind the scenes, the team was also working closely with the American Aid Logistic Network (ALAN), a nonprofit that mobilizes aid to get support to communities affected by crisis. ALAN wanted to schedule a call with mobilization managers for details on high-risk locales.

Time for more coffee.

Tuesday, September 27

“Sandbag Manager” is not the real title, of course, but this retail giant does have a crisis team which includes someone who makes sure that sandbags get to the stores with a high risk of flooding. The company’s crisis team includes the sandbag expert, the CEO, and all executive leaders, and they were all huddled on Tuesday’s crisis call.

The only outsider on the call was Everstream’s meteorologist, who is a composite of the experts on Everstream’s Storm Mode team. On Tuesday, Tropical Storm Ian graduated to Hurricane status, and the models showed it was headed to either Tampa or a bit further south to Fort Myers.

The meteorologist shared the storm’s path, timing, intensity, and what stores most at risk. After the presentation, the questions came hard and fast.

“At what point at store X do you begin to see Hurricane force winds hitting?”

“Will there be inland flooding in Orlando?”

“Where will we see a combination of high winds and flooding?”

Store by store, factor by factor, risk score by risk score, Everstream’s experts walked the retailer’s crisis team through the likely scenarios as they decided how to prepare.

Wednesday, September 28

Our meteorologist’s Wednesday started right after his Tuesday ended. He was a guest on a virtual panel in Singapore which started at 2:00am his time. Done by 4:00am, he turned back to the storm models and saw Ian would bypass Tampa for Fort Myers.

Then the client calls started – not only Everstream’s regular Wednesday briefings, but the 20+ other clients looking for Florida risk updates specific to their locations and business operations. Then the ALAN briefing for the recovery experts, more press calls, and a live Q&A call with almost 100 small business owners in the pharmaceutical industry.

Ian made landfall Wednesday afternoon at 3:15pm local time at Cayo Costa, Florida – a thin stretch of barrier island off the coast of Fort Myers. Sandbags were in place.

Thursday, September 29

After a storm, clients want to know the weather forecast for cleanup missions, and if there are any additional storms on the way. That wasn’t a risk for Florida — Ian’s third landfall was predicted for North and South Carolina, fortunately both at much lower risk for damage.

Intelligence Solutions experts stepped in to shift Everstream’s work from forecasting weather to assessing impact damage. The team started collecting information from local sources that would pinpoint where damage occurred and who was without power. Clients needed accurate and timely information about which economic areas of their operations were the most affected.

Eight days of lead time gave the retailer enough advance warning that they were able to protect multiple locations by trucking in sandbags, boarding windows, closing stores, and other proactive measures.

And with a clear forecast ahead, cleanup and recovery in Florida could begin.

*Company details changed to protect client privacy.

Get our ebook to learn how predictive analytics can accurately identify and rate business risks.

Share this post