What you should pay for a tutor

posted on 1/3/2012 Tagged as: Tutoring teaching tutoring demand tutoring supply data 18 notes

Aaron is a cofounder of Tutorpsree, and a tutor. He’s the resident numbers dork (well, so is Paul), and writes about what we’ve learned and how it influences what we do

The start of the new year means it’s time to take a look at overall trends in the tutoring universe. We’ve sliced up the profiles of 10,000 tutors to get a sense of pricing dynamics across the US. We’ve also pulled data out of nearly 80,000 subjects listed by tutors on Tutorspree as a way of understanding what it is that people want to teach.

The Price of Tutoring

One of the more opaque pieces of the tutoring market is, of course what tutoring actually costs. To get a sense of what people should or should not charge, we analyzed listed prices across our largest markets. As most of these are large cities, we expected to see fairly similar dynamics across them. The regional differences, however, are pretty extreme on a percentage basis. Maybe LA’s prices are as depressed as they are thanks to all of the unemployed actors?

Along with the average, the range of prices charged gives us information on market dynamics. We know that, based on our demographics, we are not targeting the extreme high end of tutoring (the $1000 an hour base - yes, it exists). We do, however, see that there is a natural bottom to pricing - just north of $10 an hour. Logically, that price clearing minimum wage makes sense, however seeing it in action across the country demonstrates that people, without data, have a gut feeling for what their time is minimally worth, even if they don’t have an exact sense of their maximum worth.* The data also raises questions about the overall value provided by tutors clustered at the extreme high end relative to the middle and low ranges - which we will look to tackle as we aggregate and analyze more information.

What Teachers are Paid

According to the National Bureau of Labor Statistics, the overall median hourly rate for educators is actually $21.97, which is less than 50% of the hourly wage listed on Tutorspree, and less than 8% of the max price in NYC (the highest priced market). Of course, tutoring does not come with health insurance, pension, etc. Still, the disparity between the pricing of the two, between what people are seeking as tutors (along with what they are actually being paid), and what they are paid as teachers within the classroom is stark. It also raises valid questions about what the actual price point of an hour of learning should be given the need and the benefit it offers. We don’t yet have the data to accurately evaluate that - but it is a line of inquiry that we want to continue pursuing.

What Tutors Tutor

Based on our understanding of the overall tutoring market, we expected that the top categories listed as competencies by tutors would skew towards Math. STEM tends to be a confusing area for students, and one that seems to be particularly well suited for 1 to 1 instruction (though we’d argue that all subjects are better in 1 to 1). Additionally, tutors capable of teaching one area of Math are often capable of teaching related fields. You’ll see quite a few tutors capable of teaching Calculus, Algebra, and Statistics. As a result, Math as a whole gets a boost as a subject. So combining the strong tradition of Math as a tutored subject with the breadth of subjects on offer, you get a significant plurality of subjects on offer for Math.

What Students Want to Learn

Tutors are listing the subjects they want to teach based largely on what they know. What’s missing from this equation is a consideration of what people want. That makes sense - students are less likely to actively post their needs than they are to look through what is available via traditional methods (Craigslist, flyers on telephone poles, etc.). If they do not find what they want, they generally give up. There is, therefore, no demand feedback into the marketplace. To replicate that loop, we did an analysis of our lesson requests as a proxy for overall demand in system.

We noticed a few surprising things here:

Requests for programming tutors are a much larger category than pure supply would indicate. Whether this is a function of a fundamental shift in what people are trying to learn, or a function of what people looking for tutors online are trying to learn is a question we can’t yet answer
The small size of test prep as a demand source is immediately noticeable, especially relative to the overall dynamics of tutoring. However, it makes sense given our stronger focus on non-testing subjects - we know the tests are important, but academic topics have been our main target.
The “Other Category” is really large. It represents a broad array of subjects, from music to accounting, that merits its own analysis down the road. There’s definitely a lot going on in there

Summing it all up

2011 has given us a lot to think about with regard to 1 on 1 learning. Our marketplace has provided us with a fascinating trove of data, and we find new sources/uses for it daily. There are a huge number of questions we still want to answer - questions on payment patterns, demand dynamics, etc - and we’ll be sharing those findings with you as we come up with them. Overall, though, we’re learning that the lack of transparency in the market leads to gross inefficiencies. Those inefficiencies need to be solved in order to make tutoring a fair and viable resource for students that need it. Using data to suss those problems out, and build solutions for them, is what we live for.

*Looking at the pressures on the supply demand curve, we’ve seen something endemic to immature marketplaces, especially the “collaborative consumption” variety: there is upward pricing pressure, but little downward pricing pressure. We’ve seen repeated examples where a tutor prices him/herself below the local average, receives too many requests, and begins raising prices until the requests fall off. They can find equilibrium by starting low. On the other end of the spectrum, tutors who initially price too high do not get indicators - instead they get no contact which creates a perception of inactivity within the overall marketplace. The solution to that lies in creating greater transparency - hence one of the reasons for this post.