While starting a tutoring marketplace is easy, scaling is often where the troubles begin. Tutoring marketplaces require a base of tutors that have the bandwidth and empathy to work with students across different learning styles, goals and comprehension levels. The nuance means that fast scale isn’t foolproof and can lead edtech startups into a classic marketplace downfall: the inability to grow consistently while also providing definite outcomes.
But, as 2020 showed edtech, the demand for quick and convenient help is high. To win post-pandemic, the sector needs to think bigger about the way it can reach more students in an effective and savvy way.
Innovation from Quizlet, Chegg, Course Hero and Brainly shows that the future of tutoring might not look like a 30-minute video on Zoom or Google Hangouts. Instead, modern-day extra help might take the form of an AI-powered chatbot, a live calculator or tech more subtle than either.
Regardless, the rise of tutoring bots over marketplaces illustrates that some of the biggest decision-makers in edtech are taking a scalpel to the way that tutoring used to work and hope to scale faster by doing so.
The businesses driving the change
On January 31, Chegg will close its standalone tutoring service, which matched vetted tutors with students, relaunching it into a live chatbot that answers students’ questions. The move from a tutoring marketplace to chat interface, according to a spokesperson, will help Chegg “dramatically differentiate our offerings from our competitors and better service students.”
“Ever since Chegg Tutors was launched in 2014 we have seen what a powerful tool synchronous tutoring is for learners,” the company said in a statement. “What we have also learned is that the real need for learners is contextualized help directly in the experience of their actual learning environment.”
The closure of a marketplace isn’t necessarily a failure; the company says that live tutoring was never a big part of its business. Still, it’s clear that Chegg didn’t see enough opportunity to match students and tutors live and saw more promise in a chatbot approach. Plus, it goes well with Chegg’s theme of self-directed learning. CEO Dan Rosensweig was unavailable for comment.
While Chegg’s next move is in the works, Quizlet has been tweaking the way it tutors students for over three years with a tool called Quizlet Learn.
Quizlet Learn is trying to recreate human tutors in a digitized way. First, the tool asks students to share their goals, then it assesses them on current comprehension levels. After those metrics are analyzed, Quizlet can create a custom study sequence and path to answer those goals. As a student goes through the system, Quizlet Learn consistently assesses students to see where they are making mistakes — and where they are making progress.
“If every student had their own personal tutor, they would probably do better, right?” says Quizlet CEO Matthew Glotzbach. “But it’s obviously impractical, too time consuming and not very cost effective to try to pair every human on the planet with their own expert tutor.”
Quizlet used its data around flashcard sets questions and trained natural language processing tools to understand how students might respond to certain prompts. Artificial intelligence gives the company a little more flexibility in understanding the different ways a student could correctly answer the same question.
“It obviously doesn’t yet replace and can’t come anywhere close to replacing a human, but it can provide that guidance and point you in the right direction and help you spend your time in the right places,” he said. “Just even helping you set goals is such a critical step in learning.”
Quizlet Learn is now the company’s most popular offering.
Two other quick examples that underscore validation for quick and fast Q&A? Brainly, a homework-helper startup recently announced that it has grown to 350 million users in 2020 from 150 million the year prior. It raised $80 million to expand into the United States in December. Additionally, Course Hero scooped up Symbolabs, a math equation calculator, to help boost up its Q&A services.
“You can’t just [do this] in a short amount of time,” says Course Hero CEO Andrew Grauer. “You’re looking for how do we get the right, accurate answer. But then, how am I going to get, not just accurate, but step-by-step solutions that are actually helpful.”
What does this mean for early-stage startups?
A future of chatbot tutors is a future that bets largely that the rise of self-directed learning will continue. There are two complications that can arise with current innovation.
First, there is a specific type of student who prefers quick help versus extensive conversation and debate around a topic. Take early learners, for example, who are experiencing learning loss during the pandemic because second-graders aren’t used to learning through Zoom. A chatbot won’t work for a student with special needs or someone who needs to be handheld a bit more. Quizlet is used by two-thirds of all high schoolers in the United States, and Chegg’s core users are adult learners. The specific needs of these populations might be high-level enough that speed can be prioritized over personal 1:1 learning.
Second, speed tutoring can only work for a specific set of subjects. Matt Burr, the CEO and co-founder of Nomadic, says that all learning doesn’t have to be personal, especially within the topics that have only one right answer.
“If you misunderstand an equation in math we can send you backward to review and map this out literally,” Burr said. “But when it’s something like, how do you communicate well or how do you motivate in the middle of the pandemic” the conversation is more difficult.
As bigger edtech companies focus their tutoring plays on faster Q&A, it could bring competition to early-stage startups that offer tutoring as their core and only service. After all, there’s a long list of tutoring marketplaces that have had to shut down because of inability to acquire customers, generate demand or prove outcomes.
Private startups that are still bullish on tutoring marketplace plays will need to differentiate by finding hungry and needy end users that need heavy support.
Juni Learning, for example, is a seed-stage startup that connects kids to math and science tutors. It frames itself as a “full-stack experience” for platform, curriculum and instructor information. Plus, it builds dashboards to monitor if a student shows up to class and texts or calls people if they’re not in class to check in. The deeper integrations are part of its competition advantage, Vivian Shen previously explained to TechCrunch, and creates stickiness with the platform.
In 2021, tutoring platforms can’t simply be middlemen that take a cut; they have to be extensive, smart and responsive.
Comment