Marketplace: Metrics and Growth Mechanics

Marketplace cannot be viewed as a funnel of one type of user: an improvement for buyers can worsen the economy of sellers and vice versa. Start with the unit of trade, each party’s constraints, and local liquidity—the probability of a successful match in a particular category, geography, and time.

Base: what to count in the marketplace

Key Marketplace Metrics

The main task of the marketplace is to ensure effective transactions between sellers and buyers. So it’s important for you to follow the metrics at the intersection of supply and demand. Here’s the base for any CPO:

• GMV (Gross Merchandise Value) is the total volume of orders through the platform. It shows turnover well, but not margin and not repeated purchases.
• Take Rate – What percentage of GMV becomes market place revenue before cost is taken into account. A small change in rates can have a big impact on margin and participant behavior, so it cannot be measured separately from volume and retention.
• Number of active sellers and buyers. Retention and dynamics are the litmus of platform health.
• Conversion Rate – How often visitors turn into real buyers
• Retention Rate – How many users return for a new order.

Example: Why GMV is not always relevant

Marketplace services with high turnover, but low repeatability of orders (for example, one-time repairs) can report a large GMV, but in the long term lose in economics to leaders with lower revenue, but high LTV and retention. Always consider how GMV converts into revenue and repurchases.

Classical mechanics of marketplaces

Balance of supply and demand

The main risk of the marketplace is skewed either towards sellers or buyers. It is usually difficult to synchronize the growth of both sides at the start.

A popular mechanic is subsidies or promotional codes for one party. There is a situation when you artificially increase the attractiveness of the platform for early users. The main thing is to track the points of failure in time and gradually withdraw from subsidies without a strong subsidence of the metrics.

Example of grant withdrawal

When Uber launched in new cities, the company often gave discounts and bonuses to both drivers and passengers. Conditions were then tightened, and heavily inorganic users were screened out of retention and repeatability metrics. It is important to switch from growth at all costs to tracking the return on investment in stimulus.

Level Two Metrics: Funnel, Trust and Quality

Deal funnel and loss points

A good CPO builds a clearly visible transaction funnel: viewing the product – adding to the cart – payment – delivery – return. At each stage, look for bottlenecks. Like a lot of abandoned baskets? So, there are barriers: payment difficulties, insufficient photos of goods, non-obvious delivery conditions.

Trust metrics and user quality

In the marketplace, the quality of sellers and goods directly affects the retention of buyers. It is important to track:

• Proportion of bad/failed orders
• Average seller ratings, fractions of low ratings (e.g. NPS below 6)
• Time to resolve disputes and prompt support

Example: Quality control of sellers

On marketplaces where it is easy to create an account (for example, Avito, OLX), there are many one-day sellers. Entering ratings, reviews and service standard requirements helps clean up the market and improve quality for loyal customers.

How to calculate viability: unit economy and LTV/CAC

Initial formulas: count “on a napkin”

The basic formula for success is that if a customer’s Lifetime Value (LTV) is larger than the Cost of Acquisition (CAC), the model is viable. And look at that.

• CAC Payback Period – How many months does it take to recoup the cost of attracting a user?
• Marketing Spend % of GMV is the share of marketing costs in total turnover.
• Contribution Margin – How much margin each order is after deducting all variable costs (logistics, payments, support).

Example: A classic unit economics miscalculation

If the marketplace spends ₽500 to acquire a customer and receives ₽2,000 in lifetime revenue before costs, the model may have room to grow without permanent subsidies. If CAC remains higher than LTV, the acquisition or monetization assumptions need to change.

Mistakes, anti-patterns and checkpoints

Top 3 Mistakes of Product Teams

1. Dumb copying of metrics of Western players, not taking into account the specifics of local markets.
2. The emphasis is only on GMV and MAU, ignoring the real unit economy and return on investment.
3. Targeting too early is mass marketing before the liquid market problem is closed and conversion/retention metrics are set.

Example: how to avoid scaling the hole

It was observed on niche marketplaces of services: attract many sellers without checking the solvency of demand. As a result, the site has to artificially support sellers - losses grow, quality falls, users leave.

FAQ

What is the GMV and what is the take rate?

GMV is the sum of all confirmed transactions. Take rate is only compared to sites of a similar category and service model: logistics, payments, returns and subsidies radically change the economy of the same commission.

What are the first metrics to run in the MVP marketplace?

First of all, GMV, number of transactions, conversion rate, retention of users, number of active sellers and buyers. For starters, these indicators are enough.

How do you know if the balance of supply and demand is broken?

Signals are long queues without deals for one party, long waiting periods, a lot of rejections. Try to stimulate the weak side through promo and analyze the channels of attraction.

What role do ratings and reviews from sellers play?

Critical. Rating directly affects the choice of the user. Review systems help filter scammers and set a standard of quality.

Do you always need subsidies at the start of the marketplace?

Nope. Sometimes it helps to start in an organic niche (for example, b2b products with a pronounced demand). But often without first incentives can not do.

How can we move from growth to efficiency over time?

The signal to scale is stable repeat trades in target cohorts, positive order contribution after variable costs and a channel that leads both sides of the market reproducibly. The LTV/CAC threshold is chosen based on the payback period, risk and working capital needs of a business, not on the universal coefficient.