Officials in Columbus City Schools were looking for a solution last year to some of the educational fallout of the pandemic — and they thought they found it in Paper, a popular virtual tutoring company that says it offers high-quality support for students at a lower price point.
The district spent $913,000 in COVID relief funds for Paper to provide its middle and high school students with access to 24/7, on-demand tutoring.
But Columbus quietly cut ties with the company in September because too few students were using the tool. District records obtained by Chalkbeat show that less than 8% of students with access logged on last school year. Half of those students used it just once. In some schools, not a single student logged on.
“I’ve had personal experience with it with my student,” school board president Jennifer Adair said of her rising seventh grader at a June meeting. “It was frustrating, and annoying, and she didn’t want to use it again.”
School districts across the country have spent millions in COVID relief dollars to purchase services from virtual tutoring companies to try to plug pandemic learning gaps. Paper, an eight-year-old company based in Montreal, has emerged as one of the most popular players in the market. It holds multi-million dollar contracts with some of the nation’s largest school districts and has splashy billboards in cities like Tampa and Chicago.
But educators and officials in districts that were among the first to contract with Paper say its text-based tutoring service often frustrates the students who need the most help, isn’t easily used by the youngest students, and can go unused altogether.
Philip Cutler, the co-founder and CEO of Paper, says the company has made several changes to respond to student and district feedback, with more in the works. Paper is piloting a voice notes feature aimed at helping younger children and English learners more easily use the platform. And the company has taken several steps to try to boost usage.
At a time when many students need academic help, Cutler says his company has proven it can deliver that on a large scale.
“It would be fantastic if we could have a tutor who sits next to a student for eight hours a day while they’re in class and helps re-explain everything to them,” Cutler said in a November interview. “Are we able to do that for 60 million students? I don’t think so. We need to make sure that there is something that actually can be applied to millions of students, that they can take advantage of.”
Still, schools’ reliance on programs like Paper worries observers like Allison Socol, a vice president at the education civil rights group The Education Trust who wrote a guide to spotting quality tutoring programs. Even with staffing challenges, she says, schools can do better.
“I am purposely not going to call it tutoring, because it’s not,” said Socol, of on-demand virtual help. “It doesn’t mean it’s not useful to some students. But is it useful at the scale that we need, and is it worth the amount of money that a lot of districts are spending? My gut says no, and a lot of the emerging data also says no.”
Why on-demand tutoring, and Paper, took off during pandemic
Paper traces its origins back to when Cutler saw firsthand how private tutoring can fuel academic inequities. While attending a teaching program at McGill University, Cutler ran a tutoring business that catered to children of wealthy families. “The other 90% needed the help the most but didn’t have the resources at home,” he said in an interview last year, “and no one was serving that side of the market.”
Cutler co-founded Paper in 2014, shortly after he graduated, and within four years the company had some district clients. But Paper really took off during the pandemic.
Schools, flush with COVID cash, wanted to offer tutoring to their students, but often struggled to staff and schedule those programs. Many districts couldn’t find enough teachers, who were often too exhausted and stressed to tutor for extra pay. And in a tight labor market, other adults were hard to recruit, too.
Paper offered a solution: It found and hired the tutors, and connected them to students whenever they needed help. Paper said its on-demand model could help schools reach struggling students who had to work or care for siblings after school, or who didn’t have a parent at home to help them with assignments.
“Paper aims to address the inequities facing all students, especially those from marginalized groups,” the company said.
Paper says it now works with 400 districts across the U.S. and Canada. Among its clients are four of the nation’s 10 largest districts: Los Angeles Unified, Clark County in Nevada, and Palm Beach and Hillsborough counties in Florida. Other big clients include the school districts in Boston; Prince William County, Virginia; and Jefferson County, Kentucky. Together, those contracts are worth $24 million and counting, records obtained by Chalkbeat show. (Los Angeles’ contract has yet to be finalized, Cutler said.)
Paper also holds statewide contracts worth $12 million total to provide virtual tutoring to students in grades 3-12 across Mississippi and to high schoolers in Tennessee.
Here’s how the service works: Students log on to Paper, type in a question, and get matched with a tutor. Students chat with the tutor over text message, and they can draw a problem on a virtual whiteboard. But the student can’t see or hear the tutor in real time, since there’s no live audio or video.
Even Paper’s marketing materials illustrate why that setup can be hard for some kids.
In transcripts of real tutoring sessions Paper provides to potential clients as “exemplary,” the company includes a session in which an elementary schooler needs help with basic math.
“I need help taking away,” the student types.
The tutor asks if the student knows why they’re having a hard time with subtraction.
“10000 - 0872,” the child responds. Drawing on the virtual whiteboard, the student reaches an incorrect answer: 2666.
“Can you explain what you did on the top with the 0’s?” the tutor asks. The student struggles to explain, starting with, “well I crossed it out.”
“Yeah! Do you know why you had to do that?” the tutor asked.
The student then left the session before getting guidance. But Paper noted they “left a glowing review for the tutor.”
Costs balloon when few students use Paper’s online tutoring
In Columbus, 14-year-old Zion Holbert used Paper last year to get feedback on a writing assignment comparing themes in “The Hunger Games” books to historical events. He found the site confusing at first, but eventually he figured out how to upload his work and he took some of the tutor’s editing suggestions.
His mother, Lucetta Holbert, appreciated how quickly the feedback came in. “It was really nice because I’m not a really good writer,” she said, “so that took the pressure off me to try to figure out how to do all this comparing and contrasting.”
But Zion never tried Paper again, though he was struggling with some math concepts, like fractions, that he learned when school was remote. When he wanted math help, he’d stay after school to review problems with his homeroom teacher or visit the library to work with a volunteer tutor.
His advice to other students? Paper can be helpful for English class, “if you’re stuck and you got work you want someone to read and you’re at home.” But if you have a more complicated question, seek out a teacher at school. “In person, you can just show them and they can help you,” Zion said.
One of Paper’s biggest selling points is that districts can offer unlimited virtual tutoring to many students at a fixed price. A Chalkbeat review of 13 recent district contracts show the cost per student can range from $21 to $183, though the median price was around $40. (Cutler says rates vary based on district size, contract length, and how much help the district needs to get started.)
But when students like Zion don’t return to the service or don’t use Paper at all, the true per-student cost is much higher.
Paper charged Columbus schools $38 per student for tutoring access, but district officials noted that the cost ballooned to $446 per student who actually used the service.
Santa Ana Unified in California paid Paper over $1.1 million last year to provide access to nearly 41,000 students. But just over 1,000 students logged on for tutoring or essay help from December 2021 to May, district records show, ultimately costing the district nearly $1,100 per child. (The district is no longer using Paper.)
Cutler says he believes Paper’s product is worth the cost when the number of help sessions roughly equals the number of students with access to the tool. Even by that generous standard — which can count the same student multiple times — the company often falls short.
In Hillsborough County, Florida, around 16,000 students used Paper from September 2021 to this September, or just under 14% of the middle and high schoolers with access. Those students logged around 47,000 tutoring sessions and essay reviews, district officials said — less than half of what Paper had projected. The usage was so off that the company ended up owing the school district over half a million dollars.
Usage looked similar in Palm Beach County, Florida. Some 104,000 middle and high schoolers had access to Paper, and they completed around 53,000 tutoring and essay review sessions last school year, district records show. That was within the projection in the district’s contract, but under what the company considers a good deal.
But school board members there had another concern: students from high-poverty schools used Paper less than their peers at more affluent schools.
Paper says usage rates improve when the company and district make concerted efforts to reach out to teachers, students, and families to tell them about the service. But others say the numbers reflect a problem baked into Paper’s opt-in model.
“It’s not necessarily the virtual part of it,” said Socol of The Education Trust. “Online homework help puts the responsibility on the student to say: ‘I don’t understand this individual question on my homework, let me reach out to a potentially random adult who I don’t have a relationship with.’”
Research released last month seems to back that up. In California’s Aspire charter school network, only 1 in 5 of the middle and high school students in the study used Paper in spring 2021. But higher-achieving students were almost twice as likely to use the platform as students who’d gotten at least one D or F the prior semester — the exact students the charter network had hired Paper to help.
More struggling students did try Paper when school leaders urged them and their parents to do so, but “take-up remained low,” the researchers wrote.
“If you expect them to bring their questions to the tutoring, that’s very difficult, too, because many students don’t quite know what they understand or don’t,” said Susanna Loeb, an education professor at Brown University who co-authored the study. “As a strategy for supporting students in need, it’s not a good strategy.”
Jillian Eichenauer, a middle school math teacher at an Aspire school just south of Los Angeles, has seen that in her classroom. Last year, she had her eighth graders redo problems they got wrong on their tests with Paper, but she didn’t turn to the tool if they were struggling with a concept, like writing an equation.
“I usually try to direct them to do that when it’s like a check for your answer, rather than get help,” Eichenauer said. “Some of them require re-teaching, which Paper is not very beneficial for.”
Paper’s shortcomings for young kids, English learners
Young students and students learning English as a new language have an especially hard time using Paper, educators say, though the company markets itself as being accessible to both. Paper employs tutors who can speak Spanish, French, and Mandarin, which has been a draw for many district clients.
But in several places, usage was especially low for those two groups of students. In Santa Ana Unified, a mostly Latino district where 40% of students are English learners, just four students used Paper in Spanish, data provided by the district for last school year show. No first or second graders logged on, and only two third graders did.
In Palm Beach County, only about 1% of tutoring and essay help sessions were conducted in a language other than English last school year, though 11% of students who had access were English learners.
Several districts, including Boston, Clark County, and Los Angeles, are paying Paper to use with children as young as 5, though experts in early literacy say kindergartners and first graders typically aren’t able to read and respond to a virtual tutor over text-based chat. Struggling readers in second grade are likely to have trouble, too.
Amanda Samples, the executive director of academic support and school improvement for DeSoto County schools in Mississippi, which uses Paper in grades 3-12 through the state’s initiative, says that when she reviews tutoring session logs, she can tell some younger students don’t realize the virtual tutor is a real person.
A student “might say: ‘I need help with vocabulary,’ and so the tutor will ask a question back, and then they may just not respond,” Samples said.
In Chicago’s west suburbs, middle school teacher Hannah Nolan-Spohn has used Paper to help English learners practice their conversation skills. But some have found the platform challenging without a voice option. The speech-to-text feature hasn’t helped much, either.
“The bot doesn’t always understand what it is that they’re trying to say,” Nolan-Spohn said, “and then they get frustrated.”
Paper makes changes as more schools seek online tutoring
Paper has acknowledged some of its shortcomings and says it’s working to improve.
“We realized for the younger students, in particular, they don’t necessarily have the ability to sit in a live chat,” Cutler told Chalkbeat earlier this summer, acknowledging that setup can be challenging for students with disabilities and English learners, too. “They’re just learning a language.”
As a fix, the company introduced a voice notes feature in Los Angeles’ schools this fall. It allows students to upload a recording of themselves speaking, but they still can’t have a live two-way conversation with a tutor. Cutler said it’s shown promise so far, and students who use the voice memos are more likely to return to the service. Paper intends to test it out in Boston before making it widely available next year.
Many virtual tutoring competitors now have live audio and video options, but Cutler says Paper doesn’t plan to change that part of its model because student focus groups haven’t shown a demand for that.
The company says it’s also stepped up efforts to make students and families aware of its services, running contests with prizes for schools that use Paper a lot and hiring staff who train teachers and demonstrate the tool for students.
Asked why students from high-poverty schools used Paper less, Cutler said that challenge is not unique to Paper. The company has worked with some districts, such as Clark County, to launch virtual tutoring in high-need schools first.
“What we need to do, and we are doing, is really focusing a lot of the messaging on: How do you support students who don’t really trust the system?” Cutler said this summer. “They are not the first ones to say, ‘Hey I think this is going to help me.’”
In the meantime, school districts are deciding how long they should give Paper to prove its worth as the deadline for spending federal COVID funds looms. Earlier this month, for example, the Hillsborough County school board renewed its contract for 10 months at a significantly reduced rate after raising concerns about low usage rates. District officials said they’d drop Paper if the plan to get more students logging on didn’t work.
“This is a lot of money,” Superintendent Addison Davis said, adding he wanted “to make certain that we’re getting the return on investment.”
How effective Paper is at helping students also remains an open question. The Aspire study found that when students and families got extra nudges to use Paper, and did, those students were 4 percentage points more likely to pass all their classes. Paper is involved in other ongoing research, but there’s not much else to go on for now.
Cutler maintains that Paper is “a critical piece to recovery.”
“I would 100% disagree with the fact that it’s not a solution that can address learning losses,” he said. “It absolutely is. And it’s being used that way by districts across the country.”
Leaders in districts like Mississippi’s Jefferson County schools are banking on it. Superintendent Adrian Hammitte jumped on the chance to use Paper through the state’s initiative.
“Coming from a school district with not many resources,” he said, “with it being free and offering 24/7 support, I was pretty much sold.”
Elsewhere in Mississippi, districts are using Paper to approximate “high-dosage tutoring” — a highly effective strategy in which students attend multiple tutoring sessions per week, during the school day. Cutler and the company’s marketing materials say Paper can be used in a high-dosage way, though its model is missing key components of that research-backed strategy, such as providing students with a consistent tutor.
Others say the kind of help Paper offers isn’t enough to catch up struggling students — and the fact that so many districts have turned to it raises questions about the country’s capacity to truly help the students who need it most.
Tony Solina, who oversees 16 Aspire schools in California, says it’s unrealistic to think Paper is going to “close the learning loss gap” the way most schools use it. His strategy to do that was to make sure the schools he manages had an after-school program staffed by educators who build relationships with students and their teachers.
“That’s, to me, the gold standard,” he said. “I don’t believe any online system is going to do better than or trump that.”
Kalyn Belsha is a national education reporter based in Chicago. Contact her at kbelsha@chalkbeat.org.