Overview
Tutorspree helped families find, compare and book local tutors. Founded in 2010 by Aaron Harris, Josh Abrams and Ryan Bednar, it joined Y Combinator’s Winter 2011 batch. Its starting promise was straightforward: make a trusted tutor easier to find than a personal referral. [1] [2]
The company closed in September 2013 after raising about $1.8 million. Contemporary reporting said it still had cash and planned to return remaining funds after paying employees and creditors. Closure therefore needs a more specific explanation than simply running out of money. [3]
Founding Story
Harris had spent a year at Compass Advisors and three at Bridgewater, where his work included internal product design. He and Abrams began working on Tutorspree after he left finance. [4] [1]
The problem came from experience on both sides of tutoring. Harris told Fueled: “The idea itself grew out of our shared experiences finding tutors in high school and as tutors in college.” Tutors needed clients, while families lacked an easy way to compare recommendations. [5]
Abrams led marketing and Bednar engineering. Bednar had worked at SeatGeek, a DreamIt graduate; Harris and Abrams found him through a Hacker News hiring thread. Early hire Paul deGrandis came from OurShelf and Code for America, and freelance designer Brent Meyer was a Drexel graduate. These connections helped assemble the product team. [1] [6]
Timeline
- 2010: The founders start Tutorspree. [1]
- 2011: The W11 company launches and expands its city coverage. Early investors include Founder Collective, Sequoia, Lerer Ventures, SV Angel and Thrive Capital. [6] [3]
- 2012: The product moves toward assisted matching. [7]
- 2013: Resolute invests a further $800,000; the company subsequently closes. [3]
- 2013: Harris joins YC as a partner. [16]
- January 2014: Wyzant acquires Tutorspree’s assets; the price is undisclosed. [8]
What They Built
Tutors created profiles with subjects, educational background and hourly prices. Families searched by location and subject, contacted a tutor, scheduled a lesson and paid through the service. Reviews and peer feedback supplied additional signals. The product also described administrator controls over accessible tutors and visibility into lesson activity. [5] [4]
Launch coverage described one unnamed high-school integration; it did not establish institutional revenue. [9]
A profile solves discovery, but it leaves a family to judge teaching suitability. That distinction matters when the buyer cannot readily assess the service before purchasing it. An introductory conversation, a relevant explanation and help choosing a tutor can support that decision. They also require time, so the economics depend on how much coordination each completed lesson consumes.
Market Position
Target Customers
Families seeking academic help were central buyers, while independent tutors supplied the lessons. The service accepted learners of all ages. A contemporary business article also reported growing demand from executives learning software, which shows that its customer base extended beyond schoolchildren. [5] [10]
By August 2011, city coverage included New York, San Francisco, Los Angeles, Washington, Boston and Philadelphia. Bednar said, “Philly is an area of focus for us,” emphasizing SAT preparation and programming. The relevant supply was consequently more specific than a national tutor count: someone needed the right subject, location and time. [6]
Market Size
The 2011 YFS profile quoted a $3.5–5 billion tutoring market estimate. Its geography and underlying method were not explained, so it offers period context rather than a reliable market-sizing basis. A rebuild should count reachable families and usable lesson hours within a chosen cohort before projecting national revenue. [1]
Competition
Families could use personal referrals, local agencies or free school tutoring. Participants in the shutdown discussion described these alternatives, without measuring their share of Tutorspree’s lost demand. [15]
These options competed for the same decision: whom should this family trust with a lesson? A directory could improve comparison without automatically replacing the reassurance of a recommendation.
The continuing category is competitive. Wyzant became part of IXL in 2021 and offers online lesson tools. IXL also publishes math tutorials made with Wyzant instructors. Preply explicitly advises tutors to use introduction videos. Video and online delivery therefore provide useful capabilities, but neither is an uncontested modern advantage. [11] [12] [13] [14]
Business Model
Fueled’s October 2011 interview reported a 50% commission on a new student’s first lesson and 12% on subsequent lessons. The difference matters: first-session acquisition economics and repeat-session economics were not identical. This report does not establish that those terms remained unchanged throughout the company’s life. [5]
Once two people meet, they can arrange future lessons directly. Pando’s contemporary account highlighted this risk. A platform needs continuing value—coordination, payment records or dependable support—to justify repeat fees. Lower fees may reduce the incentive to leave, but they do not prevent it. [3]
Traction
In June 2011, Tutorspree claimed a 60% rebooking rate and said its top tutor earned $2,000 in one month. Those figures describe early company-reported activity; they do not establish sustained retention across all customers. Contemporary closure reporting cited about 7,000 tutors. Registered supply alone does not show how many were available, responsive or completing lessons. [1] [3]
Post-Mortem
Better conversion did not ensure enough demand
Harris’s account says parents distrusted profiles and messaging. A March 2012 agency pivot doubled revenue within a month; revenue then tripled over the next six months. Margins rose from 15% to 40%, and the company recorded a profitable month. [7]
Human assistance can therefore improve a marketplace rather than merely add cost. The unresolved question was whether the improved service could acquire enough suitable customers. A product can work for individual buyers while remaining difficult to grow.
The search shock exposed a narrow source of suitable leads
In March 2013, traffic fell roughly 80% overnight. Harris suspected a Panda-related change but did not claim certainty. Most traffic recovered by June, with poorer lead quality. Paid acquisition moved toward positive returns, but volume was constrained by the locations and subjects where Tutorspree had sufficient tutor density. [7]
This mechanism is more precise than counting marketing channels. A new channel must deliver buyers whom available tutors can serve at an acceptable cost. More visitors can make an assisted service busier without producing enough completed lessons to cover that work.
The evidence has limits
A tutor posting as chimeracoder described scheduling friction and a request to arrange subsequent lessons directly. Harris replied that leakage was not their problem: customers repeatedly booked lessons and packages. He said the limitation was growth at the desired rate and economics. These accounts describe different perspectives. The tutor’s experience raises an operational question, but it cannot establish leakage as the company’s main failure mechanism. Public cohort data would be needed to compare them. [15]
The founders’ closure statement said they could not make Tutorspree the company they wanted. Its later asset sale does not identify the founders’ financial return or prove that tutoring itself was unviable. [3]
Key Lessons
- Count serviceable demand. Track inquiries that match available subjects and times, then completed lessons. Traffic and registrations are earlier steps.
- Test assistance as part of the product. Compare conversion and contribution after coordination costs. Automation is useful only when it preserves the help buyers need.
- Prove repeat value. A first match is not enough to earn recurring fees. Track whether families use the shared record and return through the platform.
- Test channel loss before expansion. Measure which customers remain reachable if the largest source stops producing suitable leads.
Sources
- YFS — founder interview, June2011
- YC — Tutorspree company profile
- Pando closure reporting, reproduced by Bamboo Innovator, September2013
- Enterprise Podcast Network — Aaron Harris interview
- Fueled — founder interview and lesson pricing, October2011
- Technical.ly — Philadelphia expansion and team, August2011
- Aaron Harris — founder post-mortem
- TechCrunch — Wyzant asset acquisition, January2014
- TechCrunch — launch and unnamed school integration, January2011
- PROFIT magazine — tutoring opportunity article, hosted by School Is Easy
- IXL — Wyzant joins the family, May2021
- Wyzant — online lesson tool
- IXL — Wyzant instructor tutorials, updated2025
- Preply — introduction-video guidance, updated2026
- Hacker News — shutdown discussion
- YC — Aaron Harris partner since2013