Business Models

PlanetMath - Business Model #1: Classes

Here's our pitch:

We have a great "Version 1.0" of Engineering Calculus that you attend for $5 per class. If you buy a 10 session block of classes in advance, we'll discount that to $45. This is considerably less than most calculus textbooks, and you get live help and a dedicated group of classmates. You can always access the learning materials on PlanetMath for free.

The class will have 10 meetings: Monday, Wednesday, and Friday at 6PM GMT, and will run for about a month. Meetings will be conducted via Big Blue Button (a free/open source video conferencing system with an an integrated whiteboard).

Key data

• Projected revenue: $450
• Overhead: 10%, $45 per month paid to PlanetMath
• Instructor's fee: $405
• Project start date: September, 2013

PlanetMath - Business Model #2: Patronage

Here's our pitch:

$5/month will buy you:

• An enhanced homepage, with your own blog and social network integration
• the ability to create private projects and articles on PlanetMath
• the ability to have your posts appear on the front page of PlanetMath by default
• the ability to vote on development priorities
• the ability to vote in organizational polls (e.g. board elections)
• "flair" to decorate your username with and link to from other sites
• other benefits to be discussed/added...

Key data

• Projected revenue: $2000 per month (400 Good Patrons at any given point in time... once we get things going)
• Overhead: 100%, $2000 per month paid to PlanetMath
• NB. This money could then be spent on paid development work (or other things).
• Project start date: September, 2013

Scaling up

For $500 we can run a class.

For $5000 we can run 2 classes and do 2 person months of development work.

NB. this is actually the "non-scaled" version of the two-stream model described above, running both models for two months. We can consider this to be a "baseline for survival", if we want (or need) to move beyond the purely volunteer-driven mode of production that we've been using so far.

• In addition to Engineering Calculus, we would add an Introduction to Discrete Math.

• This would enable us to upgrade the software to include everything from the "minimal plan" (or some equivalent amount of work later on).

For $50000 we can run 20 classes and do 20 person months of development work.

NB: 4000 Good Patrons and 100 students per month does not seem entirely unrealistic, although it would definitely require us to be delivering some nice products and services!

• We would have a class for the 20 most popular PlanetMath Outlines (See https://github.com/holtzermann17/planetmath-docs/issues/59)

• This would enable us to upgrade the software to include the "top 10" features in the Planetary tracker.

NB.: The other option is to get some investment that will get us to the $50K mark - at which point we would want to talk about ROI. I think the key feature of ROI is simply that it would help us get from the $50K level to the $500K and $5M levels described below.

For $500000 we can run 200 classes and 200 person months of development work.

NB: It seems pretty unlikely that we'd get half a million dollars coming in without being part of a grant project or adding some other business model, see Reflections below. Furthermore, it seems somewhat unlikely that there would really be demand for 200 classes (= 2000 students) -- although that's not certainly not impossible, either (there were 2000 students at my high school!). Presumably, if we were operating at this level of revenue, we would run somewhat fewer classes, and do things like upgrade the hardware, sponsor some paid content work, and start to invest in research - see Joe's draft of a PlanetMath.org Research Prospectus.

• we would add 30 more course outlines and run up to 4 sections of each class in parallel.

• This would enable us to upgrade the software to include pretty much all of the requests from the Planetary tracker and the legacy PlanetMath feature requests.

Of course, if we manage to get half a million coming in directly from users/clients, we could use some of the surplus to pay dividends to investors. If we want to work with investors, we should create a much more detailed treatment of how this potential $50K→$500K transition could work: as noted above, presumably it would involve implementing and running an additional business model.

For $5000000 we can run 2000 classes and do 2000 person months of development work.

• Since 2000 classes per month is a pretty ridiculous number, we would probably just stick with 200 classes per month, and pour more money into paid content work and hardware.

• Further, assuming we had already cleared out all of the tickets from the Planetary tracker, we could hire a full time administrator to keep things in good working order, and pour money into applied research, sponsoring 10 or so Ph. D. students or postdocs to do research relevant to PlanetMath's mission (see Joe's draft of a PlanetMath.org Research Prospectus)

Reflections

The sketch above makes various assumptions that might not be entirely realistic.

1. It assumes that we're getting most of our early content work done on a volunteer basis. That seems reasonable but definitely has its limits. It might be more realistic to add paid content work earlier in the "scaling up" outline.

2. It assumes that we can find teachers and developers willing to work for these rates.

3. We will hopefully continue to find volunteers who are willing to work for less than these rates. We will have to consider how volunteers and paid staff will relate. It's probably safe to assume that volunteers are students or interested in working on student-level research/development projects, so they may consider "publication" to be a valuable reward.

4. It assumes that there is actually a demand for these services.

5. It does not deal with other potential business models, like paid tutoring, print-on-demand publication, mathematical consulting, and software/systems consulting. We should expand the model to include these at some point, and it would probably be necessary if we're going to get to the more highly "scaled" levels of revenue.