This story is part of a series on how different types of libraries are using data to serve their patrons better. You can listen to the webcasts cited in this article in LJ’s archives at libraryjournal.com/webcasts/data-driven-libraries/ and keep an eye on this space for future coverage.
From learning what programs are working for patrons to being able to communicate the value of libraries to legislators and stakeholders more effectively, one thing is becoming more and more clear: having reliable data and the tools to analyze it are among the keys to a successful library system. Data can help to confirm suspicions, prove hypotheses, and offer evidence for the success of library programs. It can also dash expectations or surprise sleeping biases, forcing the rethinking or reinvention of a program that isn’t living up to its potential. Data, analyzed and contextualized, can also make it easier for librarians to tell their stories to legislators and stakeholders when the time comes to make the case for library budgets.
Perspective on patrons
One of the strategies for mining data is segmentation analytics, explored among other concepts in a series of webcasts LJ presented last summer (which can be heard in our archives here: ow.ly/vzNO6). Segmentation analytics enables librarians to serve their communities better by more carefully analyzing the demographics of sections of those communities. Recognizing that unique neighborhoods have diverse needs, segmentation analysis can help librarians understand those needs and learn how to meet them.
Gina Millsap, CEO, Topeka and Shawnee County Public Libraries, KS, relies on market segmentation to understand and reach her patrons. For Millsap, whose libraries serve a population of 175,000 out of one main branch and four bookmobiles, learning what customers want and being able to provide it nimbly is a particular challenge. To her, market segmentation isn’t about generalizing ideas about library patrons but learning what interests and traits they share. “People aren’t the same,” said Millsap in her webcast presentation, “but they do have a lot of things in common.” Learning what those things are, she says, helps her library better serve patrons by offering insight on everything from library programming to bookmobile stops.
“We wanted to get to know our community better,” Millsap said. “We wanted to know who used the library, and we certainly wanted to know who didn’t use the library.” To that end, Millsap and her colleagues worked with CIVICTechnologies, also a webcast series sponsor, to correlate anonymized data on library services with existing marketing segmentation data. What Millsap learned surprised her. The data showed, for example, that rural customers, who it was assumed checked out more films than books, actually read more books from the library than they watched movies. The actionable items the segmentation study brought into focus were useful, but the study also provided a needed reminder for librarians in Topeka not to take anything for granted about their patrons.
In Pierce County, WA, Executive Director Neel Parikh has used market segmentation for different purposes. Serving a population of more than half a million users spread over 1,700 square miles mostly in small, rural towns, Parikh and her staff have used market segmentation techniques and geographic analyses to help inform some tough choices for the 18 branches that make up the Pierce County Library System.
The data they gathered and analyzed helped them allocate resources at its Lakewood branch. That branch is near Joint Base Lewis-McChord (JBL), which is home to a significant population of military personnel and their families. While the population is large, though, those patrons do not make use of the library the way that other demographics do, partly because base housing is not geographically convenient to the library. As they adjusted services at the Lakewood branch, the data clarified that the branch needed a new set of priorities. Rather than spending time and money trying to attract new patrons from JBL, a group of users that they would probably never draw in large numbers, staffers narrowed their scope, putting an emphasis on improving the services they were already providing to other members of the surrounding community.
Patron data can also inform how librarians plan services budgets as they look into the future. Johannes Neuer, associate director of marketing for the New York Public Library (NYPL) presented his thoughts on how getting accurate measurements on the way patrons use digital media can help libraries improve marketing strategies, especially with regard to younger users. “Today, it’s not enough just to measure how many followers you have,” said Neuer. The real measurements that tell librarians how their digital content strategies are performing, Neuer told the audience, are of concrete outcomes like brand mentions, website visits, and community size and engagement. For him, that means constantly tracking online conversations to monitor mentions of NYPL and the sentiment expressed toward the brand, as well as correlating online popularity of content with real-world outcomes, for example by comparing Facebook likes for an event page against actual attendance.
New York’s Queens Library has used data to help identify new stakeholders and expand the variety of services the library can provide to them, according to Tracie Hall, VP for strategy and organizations. Hall and company have had to move out from a comfortable suite of output metrics like program attendance and circulation to pace horse metrics, which take those comfortable metrics and then benchmark them against peer libraries. “These are really important metrics to have,” says Hall, “because we want to know what aspirational models exist and how we can compare our own performance and motivate ourselves to be more efficient and more effective.” The next step in the evolution of data use at Queens Library, says Hall, was to move from measuring outputs to measuring strategically aligned outcomes—not only how many kids attend, say, an after-school program at the library but how those programs impact things like reading levels and grades for those kids who show.
A big picture look from Sacramento Public Library (SPL), CA, deputy director Denise Davis helped expand the conversation on using data to make smart changes to their service offerings. Data gathered on how patrons used the library—and how they wanted to—Davis notes, informed how SPL set priorities while upgrading its website and tech infrastructure to serve users better by offering access to in-demand services like streaming media and more support for wireless users. But keeping up with users is never a finished job, and learning more about how patrons use the library improves incrementally. By improving the infrastructure to keep pace with patrons, Davis points out, the staff also improved the library’s capacity for gathering user data and learning more from it in the future. With new wireless networks in place, Davis says, “We’ll be able to do better planning and determine expansion as we need it.”
David Singleton, director of North Carolina’s Charlotte Mecklenburg Library (CML), has also made data a big part of programming in that library system. Better data management has let the library leaders determine which of their programs are reaching the most people and having the biggest impact on them. Going further than analysis, CML staffers now use software like Zoho Creator to build digital dashboards for programs that have been successful, allowing librarians from other branches to customize the ideas for their patrons.
While taking cues from data can help libraries to get the biggest bang for their programming buck, Singleton tells LJ that it can also run the risk of making staff feel their creativity and hard work aren’t valued in the face of raw numbers. “We used to do a wide array of programs. They were all good, but not all of them were necessarily tied to programming needs,” says Singleton, pointing out that data was just being used to judge the effectiveness of programs, with putting those programs into action left in the hands of librarians. “Anybody can propose a program. They just have to tie it to community need.”
In short, programming may be better when it’s not hooked to staff desires. “We have an amazing staff, but programming is about the customer,” says Singleton. “If we’re not doing a program for a reason, we need to ask why we’re doing it at all.”
While data can make a good tool for setting policies and fine-tuning programs, it’s also useful for telling stories. Parsed properly, normally dry numbers can paint a vivid picture of the impact a library has on the community it serves. While human stories still mean a lot to legislators, the combination of stories about successes and data that show how they came about can make for a potent one-two punch when pleading a library’s funding case, though John Chrastka, founder of the library political action committee EveryLibrary and a 2014 LJ Mover & Shaker, says that data can only push funding so far.
“Proving that you did what you said allows for renewals, extensions, and expansions of grant funding,” Chrastka told LJ. “But with legislators, all the data in the world doesn’t mean anything if they are not hearing from their constituents about the library.”
And when it comes to getting library supporters in the community to speak up, data has an increasing role to play. Janet Griffing, marketing and public relations coordinator at Ohio’s Wadsworth Public Library (WPL), speaking in one of LJ’s webcasts, pointed out that data isn’t just for presentations. It can also help to maximize the impact of more grassroots efforts by using market analytics to tip the scales of getting a yes vote on a levy by targeting likely library advocates to get out the vote. Given only eight weeks to prepare for a vote in spring 2012, WPL saw its levy shot down and had to head back to the drawing board to devise a new plan for a levy for the fall ballot. With little money to spend and needing a win, Wadsworth’s team compared voter registration data against their own patron information to find neighborhoods with high rates of library use.
If people in these neighborhoods were already library users, library staffers assumed, they would be more likely to get out to the polls and to vote yes for libraries. Rather than changing the minds of swing voters, Wadsworth focused with an easier path to victory—making sure the people already won over voted. Operating on a shoestring budget, library boosters used the data to target specific neighborhoods with door-to-door campaigns, signage, and direct mail efforts. The work smarter, not harder tactic bore fruit in November, resulting in a nine point swing at the polls from the previous vote and securing the levy with 55.6 percent of voters in favor.
Better data can improve the lives and performance of collection development librarians as well, by giving them a better than ever picture of what patrons want to see on shelves. Baker & Taylor subsidiary collectionHQ is preparing to launch a new Evidence-based Selection Planning (ESP) service to bring more data to bear on collection development choices. ESP analyzes past library records on both branch and system levels and then compares them to records from similar library systems using the Title360 ordering service. And last year, Innovative Interfaces released its Decision Center Product, which also works to predict demand for new titles and can help librarians manage floating collections across multiple branches.
Phoenix Public Library has been an early adopter of collectionHQ in 2011, and Phoenix city librarian Rita Hamilton reports being happy with the new ESP feature. “What [collectionHQ] provides is data that measures the success of your choices and how you’re spending money, giving you direct feedback based on use,” Hamilton told LJ. “We have been using that data for the past three years, really honing our ability to make the best use of our resources. When they started talking about ESP—how we could further automate our selection processes—we were very interested in that.”
You can always have too much of a good thing, though, and falling in love with data gathering risks ending up with more data than a system can effectively use—or asking for more than library patrons are comfortable giving. Libraries will, naturally, only be able to analyze a finite amount of data in a given period of time. That means being choosy about what data you gather. Know what you want to know and why you want to know it before you ask questions, advises Singleton. Even at data-driven Charlotte Mecklenburg Library, staffers understand that they can’t bite off more data collection than they can chew. “We don’t take on a survey unless it is focused on specific feedback regarding something that matters,” Singleton told LJ. “About five or six significant surveys in a year is all the public can take, and it’s all we can process.”
At CML, Singleton and his staff use a local company called Market Wise to conduct phone surveys. Elsewhere, free and low-cost online survey-building tools including Zoho and Survey Monkey remain popular for librarians on a budget, while Google also offers tools for making simple surveys. For more heavy-duty information gathering on how patrons use their tech offerings and programs, librarians might consider the Impact Survey, a tool designed by the University of Washington and the Gates Foundation solely for libraries, while 2014 LibraryAware winners at the Wichita Public Library, KS, used the program MindMixer to develop their Activate Wichita tool, which lets users and nonusers alike weigh in on what services are important to them.
To ease that process, software options from companies like CIVICTechnologies can help libraries parse the data they have into meaningful information, while Envisionware’s Reporter software allows libraries to track, analyze, and visualize patron data gathered on its devices, tying together multiple reporting tools into one package. Other companies are entering the arena as well, with Gale preparing to launch its Analytics on Demand service this spring. Powered by analytics firm Alteryx, Analytics on Demand aims to let libraries analyze circulation data and demographic data together. By using high-powered standardization tools that can work around flaws in patron-submitted data like addresses, the service intends to speed up the process of running market segmentation reports for users, says Gerry Sawchuk, VP of new product development at Gale parent Cengage Learning. “They can simply extract files, load them into these applications, and immediately get results,” Sawchuk says, allowing librarians to run variations on a single report to home in on data points they may have otherwise overlooked while not taking up their entire day.
Simplifying that data-gathering process for staff will be valuable moving forward, as the importance of facts and figures is on the upswing, both in improving services and helping libraries tell their own stories by improving the number of patrons they reach and more accurately quantifying the outcomes those users experience. “The most powerful statements I make about our services have three components—reach, outcome, and an individual story of impact,” says CML’s Singleton. “If we can make those three points together, it’s a very powerful combination.”