How to Build a Three-sided Marketplace
Headswap is a three-sided market. Balancing the supply and demand of a three-sided marketplace during a launch is clearly difficult. In the words of Bastian Lehmann from Postmates in his talk Building a Three Sided Marketplace:
“Don’t do it!”
But we are. And, after Jithamithra G T’s article, How Uber solved its chicken and egg problem (and you can too!), we call it the chicken, egg and nest problem. For us it is Mentors (people who want to teach), Learners, and Hubs (venues).
Why? Because this is the critical factor that needs solving in the logistics of education to make knowledge more transferable.
So how will we do it? Well when we started thinking about this it was clear that we have to be able to offer a value proposition to the first users without the need for the other sides of the marketplace.
Later a wonderful blog post on Free SaaS Enabled Marketplaces — A Novel Go-To-Market For Software Startups by Tomasz Tunguz dropped into my inbox, describing how google build up their Adwords marketplace in a similar way.
Having talked to potential users we know that many of the problems faced by freelance teachers (or teachers that would like to go freelance) can be catered for without a marketplace. These include:
- Having a website to display a profile, activity information, and blog about their area.
- Using this to build a brand and promote themselves.
- Having a way to manage their learner enrolments, scheduling and payments.
- Having a way to engage current learners with resources and lesson plans.
This allows us to position the platform as a SaaS (Software as a Service) model for freelance teachers. Built correctly this will allow teachers with current students and venues to consolidate their activities on the platform.
Allowing us to scale independent of the geographical constraints of an online-offline marketplace, and access the two other sides via the first.
We will also include some hacks to get things moving. We can manually provide the venues in the first cases, incorporating it into a plan to recruit Hubs.
We can also identify needs (such as companies that want a teacher for yoga or language lessons) and provide the teacher for them.
Hacking marketplaces by Sal Matteis is a good watch for tips on thinking about this.
As you can see from Slide 9 our strategy is to recruit manually via a number of locally valid methods that don’t scale (Read the classic Do Things That Don’t Scale by Paul Graham).
However over this time we will work very closely with our users to get as much feedback as possible and drive this into an agile development team to iterate quickly on the product towards a strong product-market fit. See the Lean methodology.
Once this is achieved we will push the growth via shares and recommends (i.e. virally — not ‘going viral’ but engineered virality.).
The first step then is identifying an accessible segment of the target market. We are starting with a ‘guerrilla marketing’ campaign, following the method prescribed by Mitchell Harper in his article How I got 50,851 views on Slideshare and 706 email subscribers for less than $350.
I am using the method and campaign to assess the interest for positioning our brand and the easiest way to reach current and potential freelance teachers.
I will write this up in length when we have finished the experiments and can provide more valid data. I was however encouraged to read Busy, but not productive by Satya Patel which I felt validated my approach to experimenting this way on the basis that one should do business based on hypothesis and experiments.
If you have questions — from my survival to the project — please let me know at james.murray@headswap.org. Always happy to discuss.
