A summary of the points that personally stood out to me from "A Guide to MARKETPLACES," a framework compiled by a VC. A Japanese translation by Yasuda-san, now a partner at DeNA Strategic Investment Office, is also available on SlideShare for those interested.
I'll add my own analysis of what drove flea market apps in Japan at the end.
What is "A Guide to MARKETPLACES"?
A compilation of insights and success patterns extracted from marketplaces around the world, created by Boris Wertz and Angela Tran Kingyens, partners at Version One Ventures.
From the introduction:
What follows here are some of the insights we've learned while observing, funding, and building marketplaces for more than two decades. There's no single way to build and scale a marketplace, but we hope this handbook provides some insight to help you in your journey. Now let's get to work. — Boris and Angela
Boris Wertz's background: a deep practitioner in marketplace businesses.
- Founded AbeBooks (a C2C book marketplace) and served as COO; acquired by Amazon
- Founded Version One Ventures; currently GM. Invests primarily in early-stage startups.
- Partner at a16z since 2014.
The document uses "marketplace" to mean a venue where third parties — not the platform operator — provide the services or goods.
What sets a marketplace apart from a standard e-commerce site is that the goods and services are supplied by a third party. In most cases, the marketplace platform acts as a type of digital middleman.
Now let's look at the summary.
5+1 key factors when choosing a market
In 2012, Bill Gurley wrote "All Markets Are Not Created Equal: 10 Factors To Consider When Evaluating Digital Marketplaces." Boris and Angela highlight five of those as most essential:
Bill Gurley of Benchmark Capital wrote the definitive post on which markets are the best fit for marketplaces: "All Markets Are Not Created Equal: 10 Factors To Consider When Evaluating Digital Marketplaces." While Bill laid out ten factors, we believe five to be the most essential.
1 / Fragmented market
Choose a space with many sellers and many buyers.
It's much easier to start a marketplace when there are a lot of suppliers and buyers. When there are just a few suppliers, they'll likely fight the arrival of a new intermediary in their market and won't want to share in the economics.
2 / Polygamous (non-exclusive) relationships
Markets where buyers are not locked to a single supplier — where "switching" is natural — increase the frequency and value of marketplace visits.
In those markets where buyers are fiercely loyal and use the same supplier every time, the value of a marketplace is reduced.
3 / High frequency
Low frequency means no brand recognition and no virality. It's important to distinguish between frequency of using the service and frequency of using the marketplace.
When evaluating frequency, it's important to differentiate between how often a buyer uses a service and how often they use the marketplace. For example, you might need a babysitter every day, but you only look for a new babysitter every two years.
4 / Expands the addressable market
Great marketplaces grow the market itself. Don't be constrained by current market size.
Keep in mind that the best online marketplaces create new value; therefore, the current market size does not necessarily paint an accurate picture of the opportunity.
5 / Own the payment
Collecting fees becomes easier. Owning the payment also improves cash flow, creating a significant advantage for reinvestment.
For sellers, it's much easier to tolerate fees when they never saw that money in the first place.
Angela added a sixth factor in a post in August 2016:
6 / Can you expand the market?
Great marketplaces create new transactions that didn't exist before. Airbnb's "let strangers stay in your home" is a perfect example.
Good marketplaces connect supply and demand in a more efficient manner, but great marketplaces create new transactions, and add new value to the entire market.
How to beat existing players
- Lower fees
- Target a vertical
- A product 10x faster (delivering the "wow" experience to users 10x faster)
- Unique inventory or merchants
Let me dig into point 2, verticals.
Go vertical
Vertical marketplaces are growing in Japan as well. Large generalist marketplaces get unbundled by vertical startups. The Craigslist-unbundling model grew from 2010 to 2012, and many of those businesses survive today.
A startup can unbundle a generalized marketplace and focus on creating the best product for a specific vertical. Many marketplaces have struck gold by picking one thing and doing it extremely well.
That said, succeeding in one category doesn't guarantee category expansion will work. When Boris was COO at AbeBooks, he tried to expand into new books. Despite strong selection at competitive prices, the category never met expectations — buyers stayed with Amazon.
Keep in mind that category expansion doesn't always work as planned.
Double down on hot spots during the growth phase
Not all transactions in a marketplace are equal.
One of the most important things you can do is identify and double down on the things that work in your marketplace.
Identify what's working. Evaluate it correctly, and double-bet on it growing further. Trying to improve everything evenly across the board is poorly suited to marketplace businesses — which have so many variables.
Airbnb found that hosts with better photos attracted significantly more customers. CEO Chesky personally went out with a camera to offer free photography.
Trust and safety at scale
Once a marketplace completes initial validation and enters a virtuous cycle of buying and selling, "building trust" becomes critical. Simply put: eliminate bad actors, and ensure transparency.
Your job as a marketplace is to fight the bad actors proactively, so both buyers and sellers have confidence in the platform. Transparency is one of the most effective ways to establish trust and credibility.
Uber filters out drivers with 3 stars or below — customers never have to see them.
Uber filters these drivers out, so customers don't have to sift through driver reviews before getting a ride.
Airbnb had a 24/7 customer service team by its third year.
In response to some specific incidents in 2011 (when host homes were ransacked and robbed), Airbnb created a dedicated Trust and Safety team that's on call 24 hours a day, as well as a neighborhood hotline where neighbors can report any questions or concerns.
Support your power sellers
The buying experience in a marketplace isn't dramatically different from standard e-commerce. The energy-intensive part is the seller-side experience. The platform needs to make it easy and fast for sellers to list goods or services. High seller satisfaction leads to more supply, which creates a virtuous cycle. Reducing friction for heavy sellers is key.
But for buyers, very little is actually out of the ordinary. Most people are used to buying or even ordering services online, either from Amazon or eBay. On the flipside, fewer people are used to shipping products, providing taxi service, or turning their home over to other occupants. For all these reasons, it's crucial that marketplaces focus first on devising an elegant, instructive and — above all, easy — experience for suppliers.
(My take)
What drove flea market apps in Japan
Fril created the category; Mercari expanded it; then came Rakuma focused on electronics, ZOZO Flea Market (now discontinued) focused on vintage clothing, and Bukuma focused on books (the mobile version of AbeBooks). Flea market apps in Japan have established a real market and continue to grow.
This success aligns closely with the theory in "A Guide to MARKETPLACES."
1 / Started in a fragmented, high-frequency market: women's fashion × smartphones
Fril's founder Horii documented the founding story on his blog:
In conclusion, we pivoted to "a flea market service where girls can easily buy and sell the clothes they no longer wear." I pivoted to this plan because I noticed girls selling clothes they'd worn in selfies on their feature phone blogs, and using Mixi and Twitter to sell clothes. It was a real insight — going back to the mindset of the feature phone era. — "Why I Created a Flea Market App"
Horii specifically attributed this market choice to "experience providing apps for women on feature phones." Choosing the right market appears to be one key reason Fril grew so quickly.
(Note: Horii-san is now a shareholder in 10X.)
2 / Vertical, and 10x faster
At the time, Yahoo Auctions and Mobaoku served the women's vintage clothing marketplace. But they had a problem:
Registration to transaction completion was long and complicated, not optimized for smartphone use in spare moments. — "Why I Created a Flea Market App"
Fril offered a platform where transactions literally completed 10x faster on a smartphone. It spread rapidly, riding the wave of smartphone adoption.
Mercari then expanded the effective market Fril had created with "all categories × even simpler transaction technology."
3 / Escrow creates "default trust" and a path to payments
Escrow — the payment format standard in Japanese flea market apps like Mercari and Airbnb's model internationally — means the platform holds sale proceeds until both parties confirm the transaction is complete, then releases the funds.
So ubiquitous that users accept it without question, this is, I believe, the single biggest factor creating trust in flea market platforms.
Looking overseas, there are surprisingly many services where the absence of escrow makes it hard to guarantee trust. OLX (Southeast Asia, centered on the Philippines) and Carousell (dominant in Singapore) are both flea market apps centered on in-person handoffs. When I asked friends who've used these services what's the biggest difference from Japanese flea apps, the consistent answer was: "There's no escrow, so it's hard to transact with anyone you don't trust." Anecdotal, but I think it's fundamental.
Also, flea market apps hold sale proceeds as escrow, generating significant cash flow. And the ratings and reviews users accumulate create new credit information — aligning with Boris's point about the "path to payments." Horii-san noted this too.
The rise of flea market apps in Japan makes a lot of sense when read against "A Guide to MARKETPLACES."
Overall impressions
I only summarized the points that personally caught my attention, but the full text covers not just recent platform names like Airbnb and Uber, but also older, more established services like AbeBooks, eBay, and Thumbtack — very educational.
Marketplaces govern transactions, and you need to keep learning the psychology and behavioral culture of the many users participating in them. As a builder of services, that's both the hard part and the most enjoyable part.
"Double down on hot spots" is a well-chosen phrase. It might suit people who can maintain the sensitivity to keep hunting for hot spots and notice them when they appear.
