A new framework for online marketplaces
Today, many businesses monitor their customers’ Lifetime Value (LTV), or the total revenue expected from an average customer over their lifetime. In an ideal world, we all want a chart like this:
But how?
About two years ago, I sat down to consider this for my business. I was a product manager at consumer peer-to-peer marketplace Rover, which connects pet parents with providers for services like dog walking and boarding.
I searched online for guidance on marketplace LTV optimization and user retention. The Airbnbs and Ubers of the world had been around for much longer and surely had figured this out, right?
The more I dug in, the more I understood that those marketplaces are quite different from Rover. Many user retention best practices were an ill fit, like trying to slam a square-shaped peg into a circular hole. I realized that companies like Upwork and Care.com had more in common with our model, and began to separate consumer marketplaces into two distinct categories. Depending on which category a business fits into, there can be significantly different dynamics for user retention and acquisition.
The Framework: Short-Term vs Long-Term Relationships (STR vs LTR)
In a short-term relationship (STR) marketplace, transactions occur over a very short time period, and demand/supply don’t build lasting relationships. In long-term relationship (LTR) marketplaces, as you may guess, the opposite happens: demand and supply typically form longer-lasting, ongoing service relationships.
This simple difference overhauls business dynamics.
Differences in how demand and supply find each other. In STR marketplaces, the demand side heavily relies on the marketplace in order to source and choose the supply. In LTR marketplaces, there are many more opportunities for off-platform discovery: through social media, word of mouth, or even in in-person interactions.
When I was planning a Paris vacation a couple of months ago, I started my accommodation search directly on Airbnb. When I need to get across town, I open the Uber app. But many Gumroad and Patreon customers discover products after becoming an avid Twitter or Instagram follower of a creator. Many Rover sitters have a mix of clients sourced from Rover and via personal referrals, friends, or even meeting at a dog park.
Differences in brand association. In STR marketplaces, the customer is more likely to associate the service with the company’s brand than the provider. I’ll say to my friend, “I’m taking a Lyft”, not “John is giving me a ride right now.” In LTR marketplaces, the opposite happens: the customer associates the service with the provider. I’m a Rover sitter, and once I asked an owner how they feel about Rover. “You’re amazing — Yoshi has a great time with you,” they said. They conflated the company’s brand with me, the provider.
Differences in what high-value customers care about. In STR marketplaces, liquidity and diversity of services are key, and high-Lifetime-Value customers are relatively agnostic to service provider identities themselves. In LTR marketplaces, high-LTV customers primarily care about maintaining relationships with their preferred providers.
This is the most critical distinction to understand in order to create a compelling LTV strategy.
STR marketplaces should focus on meeting customer needs in many diverse circumstances. Doordash wants to be able to serve me when I need a quick work lunch in the middle of downtown, as well as a midnight snack out in the suburbs. Amazon wants to be my one-stop-shop not just for a kitchen table today, but also my toothpaste next week.
Meanwhile, LTR marketplaces should focus on creating value for the supply.
In an LTR marketplace, if a supplier wants to take business off-platform, the customer generally follows. Customers may have some brand loyalty to the marketplace, but generally, they care about maintaining a relationship with a supplier.
Once the relationship between demand and supply is established, long-term relationship marketplaces must convince suppliers to keep repeat business on the marketplace, and keep giving a cut of the transaction, rather than preferring Venmo or personal billing.
Put differently:
To optimize user retention (and thus LTV), long-term relationship marketplaces should prioritize features providing value to suppliers.
That makes sense. But how do I do that?
For long-term relationship marketplaces, I believe there are just two simple steps:
Step one: Build a meaningful relationship with your suppliers. User relationship building is key to any business. But it’s especially crucial for LTR marketplaces to build this with their supply base, given their business dynamics.
Make suppliers partners in your company’s success. Airbnb maintains a vast Host community network. Gumroad hosts creator dinners, where the CEO and other team members meet platform superusers and solicit their feedback. Other companies create superuser advisor panels.
One of my favorite online community forums is Spotify’s; they do a great job of collecting feedback, sourcing feature requests, and generally helping their users feel heard and valued. There are so many ways to create ongoing, bidirectional communication channels with users.
Step two: Build tools to help suppliers manage their business. Your marketplace connects demand to supply, which is extremely useful — congrats! It also likely already has a built-in capability to collect and track payments. What else does your platform do to make business management easier?
Now that you’ve set up a feedback mechanism with your users (refer to step #1!), you hopefully have a good idea of what problems bug your suppliers.
Maybe some suppliers use your marketplace less than others, and want more flexibility with fees: customize fee models to fit a more diverse supplier set. Perhaps creators want more engagement avenues with their audience: build interactive tools like creator/audience forums, creator update feeds in-product, or audience → creator feedback mechanisms. Maybe they want more ways to make money: diversify products and services available on your platform.
One example is Upwork’s Work Diary. For hourly projects, the Work Diary allows freelancers to continuously record the work they’re doing, and offers the ability to automatically take screenshots of the work so that the freelancer can, without a doubt, show clients they are hard at work.
Rover offers the ability for providers to track walk routes in real-time, record the number of times a pet peed or pooped, and easily send photos. Gumroad and Patreon offer creators the ability to customize product promotions, and analytics tools that enable simple business performance tracking.
Many marketplaces enable tipping. This is a creative technique because suppliers may be uncomfortable to ask for tips themselves, but are happy for the marketplace to ask on their behalf.
What does your marketplace do for suppliers that they can’t easily do themselves?
How do I know I’m doing the right thing?
Plugging a feature in and then seeing revenue increase over time may seem like a formulaic process, but business stewardship is more an art than a science, and dependent on your organization’s culture.
Apple doesn’t test its features broadly before widespread launch; their product development processes emphasize developing taste and user empathy. Google A/B tests everything, down to the shade of blue for search result hyperlinks. Both companies build great products!
At Gumroad, CEO Sahil Lavingia says, “We keep things simple.” In order to determine if they are building in the right direction, they measure just one thing: how many creators are making how much money?
“Basically we empower creators to earn more money. Everything else we do is second to that. So that’s the big thing we track,” he said.
I’m inclined to skew towards an A/B-testing heavy strategy, and to believe that if something is good for users — even the smallest of features — I should be able to observe a measured change, and ideally, tie it to business revenue. It’s difficult to tie incremental initiatives to a high-level metric like customer LTV — but it’s possible. You can adopt practices like building multi-step, explicit hypotheses; stating fail/success paths prior to development; and tracking lower-level metrics like feature engagement and user sentiment (refer back to “Build a meaningful relationship with your suppliers” above!) to enable quick feedback loops.
Compared to supplier retention initiatives, almost any project will be easier to tie incremental revenue to: conversion rate optimization, expanding services, demand-side UX improvements. Still, just because something is easier to measure doesn’t mean it’s the right choice. For LTR marketplaces, prioritizing supplier retention is critical for long-term success.
Putting it all together
Consumer retention for any company is a tricky beast, and two-sided consumer marketplaces are especially unique. To date, best practices have been discussed broadly, but it’s now clear that business dynamics are distinct depending on whether platforms cultivate short-term supply/demand relationships (STR) vs long-term relationships (LTR).
Understanding which category your marketplace falls into can help you determine whether to focus on liquidity and diversity of services offered (the STR strategy) or to create value for the supply side of your marketplace (the LTR way). Of course, in the lifetime of your business, you’ll probably invest in both! But this may be a helpful framework when stack-ranking various roadmap options.
Building value for suppliers is an iterative two-step process: form a robust relationship with your users; then, build tools to help them manage their business.
There is no one “right” way to measure success. The important thing is your organization has a philosophy that makes sense. Finally, just because an initiative’s impact is easy to measure, doesn’t mean it’s the right initiative to implement. Important things can be hard (but still critical) to measure.
Anything I missed? How do you tackle these problems for your business?
Thanks to Naoko Takeda and Lina Colucci for providing feedback on this post.
