Friction in a Marketplace
“How Lumosity Spiked Active Users 10% with Complexity, Not Simplicity” is one of my favorite First Round Review articles.
… they added several survey questions into the sign-up flow, including questions about their cognitive goals and the industry they worked in. This continued for several pages before users had the chance to register and start playing games.
Initially, the team assumed this would lower registration rates, but would be worth it in the long-term to help them learn more about their users. What they found didn’t just benefit their research — they discovered that users actually answered the extra questions in full. Surprised by the results, they decided to make the questions an intentional part of the sign-up experience.
…
With all of this in place, they decided to A/B test the sign-up flow again to see if things had changed. The results were largely the same: More users signed up without the survey, but those who took it ended up subscribing at a 10% higher rate.
As product folks, we some times get caught up in removing all friction. On mobile especially, we’re hyperfocused on eliminating friction because we know that users are “on the go” — so the faster they can do something, the better their experience. But as that article suggests, in some cases, friction when applied correctly and at the right time in the flow can actually yield in higher quality users.
We talk about supply being absolutely paramount for a marketplace company to get going (examples of supply: apparel @ Threadflip, drivers @ Uber/Lyft, hosts @ AirBnB). So as a marketplace operator you focus on removing all the friction that could possibly exist for a supplier to start supplying in the marketplace. Except, this could have a detrimental effect on your marketplace if you don’t police well.
I’ve seen some really weird things in marketplaces which make me then doubt the validity of the marketplace operator: I’ve seen listings on AirBnB where the profile of the host is a logo of the agent with their number underneath. I’ve seen the case of a watch being advertised as the watch itself on eBay. These kinds of actions basically propagate the notion that
- it is ok for others to behave this way in the marketplace and
- buyers should be careful.
If you want to remove friction, remove it from the buyer’s (or the demand’s) path. The last thing you want is for your demand to second guess their decision.
Friction!
There are subtle ways in which friction can be introduced on the supply side to ensure quality and trust in a marketplace.
- “Skin in the game”: The marketplace I’ve seen do this best is Flippa.


I really like this simple approach — it’s a one-time deal of asking the supplier to put their credit card on file, should something go wrong.
2. Curate Listings: If you are a marketplace like AirBnB or Threadflip, you could considering crowdsourced vetting of new or updated listings. In essence, this is what the Apple AppStore (and now Google Play) does for apps — while they don’t crowdsource vetting, they do have an approval process that ensures the platform is copacetic. Crowdflower (or Mechanical Turk) would be good places to employ curators.
3. Vet your suppliers: AirBnB does a great job of this and services like Checkr are actually making it easier for marketplace operators to perform background checks on your users — that might be a little much in certain cases, but if you are dealing with high value goods (like property or cars for example), it makes perfect sense.
If I could augment the conclusion from my previous article on marketplaces:
quality supply is your key to liquidity.