It’s a Friday night, and you’ve just opened your Zé Delivery app. The time shown for your ice-cold beer to arrive is more than 3 hours: it’s possible you may be looking for other options to meet your demand.
It’s a Friday night, you’re a beer distributor, and you are intrigued by the low demand, as shown by the volume of orders, on Zé Delivery: you begin to wonder whether it’s in fact worthwhile doing your beer distribution through this app.
In both cases expectations were not met, and this generates a direct effect on the platform’s liquidity – the fact that transactions on the platform are not being completed.
When we’re talking about Marketplaces we are talking about network effects – and not having liquidity is a latent threat for construction of a scalable business: people stop coming back, or even indicating the platform as a solution to other clients.
This is why focusing on liquidity, and not only on the growth of the user base, is fundamental to ensure that the platform will have a strong and sustainable network effect over time.
____
As well as monitoring key metrics for validating the product, the market and network effect, it is essential to monitor the probability that a seller or a buyer will find what they are looking for on your Marketplace. As we said in our most recent publication “The complexity of marketplaces and the quest for PMF”, this is very closely related to a Marketplace’s Product-Market Fit.
In practice, liquidity is a Marketplace’s most vital metric. The reason it’s the most important is because without it, there are no transactions, no PMF, nor even a network effect.
Julia Morrongielo of Point Nine Capital produces a very interesting definition in her article on the subject:
“Liquidity is the lifeblood of marketplaces. It is the efficiency with which a marketplace matches buyers and sellers on its platform. One could say that a marketplace without liquidity has no real product because the ability to transact on the platform IS the product.”
In other words: liquidity is the product that a Marketplace is providing to its users – and is also a central factor in indicating a network’s ideal minimum density (the ‘Atomic Network’).
An interesting example to illustrate this is Craigslist: although its UX is terrible, the platform has gigantic liquidity – which is what has sustained it until today.
And whether you are a Marketplace that started by concentrating on buyers or sellers (we talk about this in our post “The Hard Side”), after the first clients enter the platform, monitoring indicators of liquidity becomes a key factor for measuring the efficiency with which people are in fact transacting on it.
Importantly, there are numerous ways of measuring liquidity. They vary widely, depending on the sector or the business model (e.g. B2C, B2B, C2C, etc.), but a unanimous conclusion in our survey was that measuring liquidity and orienting strategic decisions about the product, pricing, positioning, etc., based on liquidity, are fundamental for achieving maximum growth and efficiency in building the network.
We know that answers that begin with the words “it depends” are always an added complication when the aim is to learn rapidly – and entrepreneurs expose themselves daily to this uncertainty as they make their journey – but let’s try to make things simple.
As a first step, before trying to say what these KPIs are (and they ‘depend’ a lot), it’s worth understanding the various types of Marketplace.
Josh Breinlinger, of Turtle VC, separates them into two main groups: “Double-commit” and “Single-commit”.
Types of Marketplace:
1) Double-commit Marketplaces
- Confirmation of an order requires mutual approval – by both buyer and seller.
- Examples: OLX, Tinder
- Challenges:
- They tend to have lower liquidity, due to the need for double approval.
- Inefficiency in transactions: Usually the processes of transaction between the two parts are long, and have many stages. Time and energy are frequently wasted in processes that are not converted into transactions. And this of course impacts the experience and quality perceived by both buyers and sellers.
- In these Marketplaces, the ‘curating’, the filters and the matching are done by the users themselves. Also, sellers may need to meet some very personalized needs of buyers, in a process in which both sides are able to communicate and negotiate independently.
2) Single-commit Marketplaces
In these, one of the parties to a transaction is able to execute it without the active approval of the other. This category further subdivides into two:
2.1) Buyer-picks Marketplaces
- In these markets the sellers (by more or less automated means) insert their availability of products or services into the platform. Buyers are able to see the sellers, and what products or services they have available, and contract them without any form of friction.
- Examples:
Enjoei, iFood, Zé Delivery, Cayena, Doji, Mercê do Bairro.
2.2) Supplier-picks Marketplaces
- In this model, buyers publish their demands, for products or services, and it is up to the sellers whether or not to meet them. This this can be done with a reasonable degree of automation (for example, the 99 ride app offers the ride to drivers that are closest to the user), but at this point there is the element of acceptance or agreement by the seller.
- Examples:
99, GetNinjas
Note that many authors on this subject define this category as “Marketplace-picks Marketplaces”, based on the fact that it is the platform itself that allocates the demands, in an automatic and centralized way.
Most of these models have a strong component of automation and matching of the supply with the demand, but it’s still up to one of the sides to give the final acceptance for the transaction to be confirmed. We see Josh Breinlinger’s segregation of Marketplaces into the two basic categories (‘double-commit’ and ‘single-commit’) as the most useful for discussing liquidity metrics.
Having defined these two categories, how do we measure the liquidity of a platform and its main drivers?
To facilitate building a logical framework, below we separate the key metrics by type of Marketplace. Clearly you can adapt these metrics to the specific context of your Marketplace, but as we said above, our idea here is to simplify:
1) Double-commit Marketplaces – where both sides choose
Since these models depend on many stages of internal interaction between buyers and sellers (who make searches, interact via chats, negotiate price, etc.), liquidity is highly correlated with the process. So mapping the stages of the conversion funnel, and monitoring barriers, is extremely important.
Double-commit Marketplace
Search to Fill Ratio
# of transactions completed
# of searches or visits
Source: Astella.
Search to Fill Ratio: Of all the searches made in the Marketplace, how many result in transactions?
Here it’s worth looking at the conversion rate in each of the separate stages of the funnel. Since double-commit Marketplaces usually have a longer and more delayed funnel, when we want insights on users’ behavior and how to improve the product, it’s important to understand which phases have the highest fallout rate.
2) Single-commit Marketplaces – where the buyer chooses
For buyers to have multiple, good options to choose from, it’s necessary that sellers should be engaged in offering good products and services. That’s why the key liquidity metrics for these Marketplaces focus on this point:
Source: Astella.
2.1)
Utilization Rate:
What percentage of everything that is offered in the marketplace over a given period is converted into sales?
Share of Wallet:
How much of the whole of a seller’s billing in a given period comes from the Marketplace?
These are excellent metrics for understanding the product fit on the sellers’ side. This is because the lower the utilization rate, the less motivation suppliers have to offer products in the Marketplace (due to the low conversion rate for their offerings) – which of course sets off a very negative effect in the network.
Another important piece of information that a high Share of Wallet can indicate is how important the Marketplace is for the supplier’s sales, or even whether those sales are dependent on the Marketplace.
2.2) Supplier Utilization Rate
Of all the sellers on the platform, what percentage are Heavy Users?
Note that before embarking on calculation of this metric you need to define what is considered to be a Heavy User for this Marketplace. This can vary depending on the company’s phase and moment (the ideal is that it should evolve over time), but as an example, you might establish a definition of a Heavy User as the top quartile of users in terms of how many hours they spend on the app per month/or day.
Whatever the definition, measuring how suppliers are using the platform is important in monitoring the importance of the solution for the user – clearly, the more the Marketplace is generating qualified leads and conversions for the seller, the more the seller will tend to use the platform.
2.3) Buyer to Supplier Ratio
How many sellers can a buyer serve in a given period of time?
Monitoring and understanding this pattern, and the turnover of clients, is important for predicting problems of imbalance between supply and demand. Each marketplace has its own specific dynamic, but it’s very important to understand how this dynamic functions within the universe of your Marketplace, and make adjustments when necessary.
3) Single-commit Marketplaces – where the seller chooses
For these marketplaces, although the sellers are the final point for closing the transaction, they have less bargaining power than the buyers. This is because the order arrives already closely specified (within the range of possibilities offered by the seller), and all the seller has to do is accept the demand, or not. Thus in this case there is a high degree of standardization.
There is another point: When buyers experience any problem with their order, it is much more the Marketplace’s image than the seller’s image that is at risk. In this context, monitoring usability of the platform and maintenance of its standard of quality is a key metric.
Source: Astella.
3.1) NPS: What is the quality of the service being provided or the product being sold?
As well as using NPS, other methods of researching users’ satisfaction with their experience can provide useful quality metrics to monitor. One example is the use of instant surveys, such as the umber-of-stars approval rating used by a ride app such as 99, or the service satisfaction assessment used by iFood.
3.2) Time to Fill Rate: How fast are buyers’ demands being met by sellers?
Ideally this metric should be monitored by geographical area, by category of product or service, and/or even by time period.
