Case Study: Tutorspree

In 2011, Tutorspree emerged from Y Combinator calling themselves “like Airbnb for tutors.” If you were a parent looking for a math tutor for your kid, you could find one on Tutorspree. Similarly, if you were a grad student looking to make some money tutoring, you could list your services on the platform. 

It was a huge potential market—their research indicated that tutoring was a $5 billion market and growing. The revenue model was simple: Tutorspree would take a fifty-percent cut on each tutoring session booked.

The economic model Tutorspree presented to investors looked very solid, especially because the three founders were SEO experts, so they knew how to make the website show up at the top of Google searches for tutoring services without having to actually pay for any marketing or advertising. The recurring nature of tutoring sessions is such that Tutorspree expected the lifetime value (LTV) of a new customer would be reasonably high. They figured the average new customer would book a total of six tutoring sessions, so if they made an average of $25 on each one, then their LTV would be $150. Their CAC would be nearly zero (since the founders were SEO experts they figured they’d get Google search traffic for free). What could go wrong?

The touted their “proprietary algorithm” that would match students and tutors based on a variety of factors including location, background, preferences, and style. 

Based on their very solid economic model, they raised $1.8 million in capital from leading investors, including Y Combinator, Sequoia, and Quora co-founder Adam D’angelo. 

They signed up more than seven thousand tutors in the first few months after they launched, and soon customers began to appear, finding the site through Google and booking sessions with tutors. 

But the tutors weren’t very happy about giving fifty percent of their fees to Tutorspree, so at the end of the first tutoring session they would often tell the student, “Hey, just contact me directly next time instead of going through the platform,” so Tutorspree’s estimate of a six-session customer lifetime value (LTV) turned out to be a bit off. 

In March of 2013, Google changed its algorithm, and Tutorspree’s SEO volume dropped by eighty percent overnight. It looked like they would need to actually spend money on paid advertising, and that wasn’t in the economic model they had prepared. 

By September, they had ceased operations. The investor money was gone in less than two years.

Key Takeaways:

• Every venture distills down to one simple mathematical expression: CAC < LTV. Every entrepreneur underestimates their customer acquisition cost (“We’re SEO experts, so we don’t need to spend any money on advertising!”), and every entrepreneur overestimates the lifetime value of a customer (“Once we have a customer, they’ll come back and buy from us again and again and again!”). Finding out too late that your CAC is too high relative to your LTV will bring your startup to its knees faster than a tutor could have taught you the basic math you needed at the onset.
• As we explored when we discussed Marketplaces, one of the risks of a digital marketplace model is the threat of disintermediation. In other words, the marketplace model requires that you be in the middle of the transaction so that you can take a cut – that’s where your revenue comes from. If people want to go around you and transact off your platform, then the model fails. Successful marketplaces like Airbnb and Upwork have worked hard to provide benefits that discourage buyers and sellers from going arount them. Tutorspree failed to do this.