Even a year ago, before city officials improved their recordkeeping, before the mayor held a press conference announcing new staff and philanthropic funds, people in Mobile, Ala., knew their city had a blight problem. They just couldn’t say how big.

About 16 percent of Mobile's 90,000 housing units are vacant, according to estimates from the U.S. Census Bureau. That's higher than the national average but lower than Detroit, where the Census pegs the vacancy rate at close to 29 percent. Of course, not all blighted properties are vacant, and not all vacant properties are blighted. Anecdotally, Mobile’s code enforcement officers could describe what they see while driving around the city: slumping roofs, boarded-up windows, weeds that cover walls and more exposed wood than paint.

When Mobile Mayor Sandy Stimpson ran for office in 2013, blight wasn't a top issue for voters, but it affected much of what residents do care about -- declining property values, economic development and crime rates. So Stimpson decided to make it a priority for his administration.

"I refer to blight as being like having pancreatic cancer," Stimpson said, explaining that it has far-reaching impacts that are hard to detect at first. "What blight does to a city, it’s devastating."

Last December, Mobile won a three-year $1.65 million grant from Bloomberg Philanthropies to pay for a municipal “innovation team,” a handful of people who seek fresh, data-driven solutions to longstanding urban problems. Stimpson applied for the grant specifically to reduce blight. Joan Dunlap, who now directs the team, was Stimpson’s first hire. As soon as she got the job, Dunlap realized her team had a measurement problem. They wouldn’t be able to say how much they were reducing blight because they didn’t have a baseline.

“If someone had asked four months ago, [how many blighted properties are in the city],” Dunlap said, “nobody would have been able to answer the question.”

Twelve U.S. cities, including Mobile, have innovation teams funded by Bloomberg Philanthropies. The teams range from three to eight people, drawing from a mix of public- and private-sector expertise, and their job is to address problems of their mayor's choosing. While the teams work in the mayor’s office, they usually coordinate with multiple agencies on a range of issues. The first cohort of innovation teams in Chicago, Atlanta, New Orleans, Memphis and Louisville tackled homelessness, homicides, wait times for business licenses and neighborhood revitalization.  (For more on Bloomberg Philanthropies, read this.) While the teams have different mayors and different focus areas, they are alike in that the foundation has a specific definition and methodology for municipal innovation, beginning with a call to investigate the problem. Often, that involves both in-person interviews and an assessment of existing data.

But in Mobile, the data were scarce. All Dunlap’s team had was a log of 311 complaints made by citizens who were bothered by the look of someone else’s property. A code enforcement officer had to respond to the call within 72 hours, but usually the response amounted to a verbal warning to the property owner.

Kathryn Pettit, a senior research associate at the Urban Institute, said Mobile's predicament is familiar to city officials in other places.

"Most cities do not have comprehensive information" on their blight problem, she said. Like Mobile, "most cities are more reactive to complaint calls."

Pettit directs the National Neighborhood Indicators Partnership, a collaboration between the Urban Institute and local organizations in 30 cities. Many of those communities are focused on improving data collection on blight and vacancy rates, in part to decide where the city should invest in neighborhood revitalization. Not only do they need to know where blight is and how much exists, Pettit explained, but officials need to discern between the parcel with the unmown lawn -- a minor problem -- and the parcel with the fire-damaged house in danger of collapsing.

"You need a way to prioritize properties," Pettit said, "because there’s usually far more blighted properties than resources."

Screenshot of an Instagram photo, which shows a blighted property in Mobile. (Instagram/City of Mobile)

In Mobile, Dunlap's team researched tools for quantifying blight and decided to try Instagram, the photo-oriented social media app. The city could create an account and send code enforcement officers to take photos of blighted properties. The app would automatically document the general location of the photo and its mapping function allowed the mayor’s team to see where the biggest concentrations of blight were. Unlike the 311 call data, it was easily accessible from any workers’ computer and could be updated as officers found more properties.

Stimpson liked the idea enough that he offered to set aside city funds to buy smartphones for 14 code enforcement officers, but it turned out not to be necessary: The employees' had work phones that were eligible for free upgrades. Within two weeks, the officers fanned out across the city with their new smartphones and recorded images of 926 blighted properties.

Paula Hillery, a code enforcement officer for Mobile, had been working on nuisance and abatement issues when the Instagram sweep occurred. At the time, she knew the city had about 60 blighted properties turned over to the city, and that figure was her best estimate for the extent of blight in Mobile. After the photo-gathering initiative, she was struck by how much higher the actual number is. “There are a lot more out there,” she said.

Screenshot of an Instagram map, which shows the number of blighted properties in each area. (Instagram/City of Mobile)

As much as the Instagram images improved city intelligence on blight, the data didn’t automatically tag addresses and officers couldn't always tell if a structure just needed some repairs or was on its way to demolition. Dunlap credits the software with providing enough initial information to stir interest. “We had to break the ice,” she said. “Instagram did that.”

Since the city created the inventory, Mobile's Geographic Information Systems department provided code enforcement with a better app that allows them to pinpoint the exact locations of each property. Within six days, officers went back out to collect addresses for 1,256 blighted properties. Unlike the original Instagram files, these allowed different agencies to tag information to help the city decide which properties it should target first. For example, if a parcel has unpaid taxes and has extensive fire damage and is in imminent danger of collapse, it would be a good candidate for city action. In fact, the innovation team is testing a comprehensive scoring system of properties located in two designated "blight zones." Based on a combination of agencies' records about parcels, the city should be able to categorize properties as either on track for demolition, in need of stabilization or in need of further review.

Beyond gathering the raw data, Mobile has worked to understand why blight occurs and how it became so pervasive. Over the next two years, the city will raise the profile of blight as an issue, Dunlap said, and look to businesses, nonprofits and community groups to invest in neighborhoods and help residents keep their properties updated.

For his part, Stimpson knows that the Instagram initiative only helped define the blight problem.

“The solution of fixing blight is a much greater challenge than fixing the process of collecting data,” Stimpson said, “but it’s not often that you can completely change a city process that is old and bureaucratic at almost no expense. It’s been hugely successful from that standpoint.”