Marketplace Liquidity: Measure Match Success on Both Sides

Marketplace liquidity is the ability of eligible buyers and suppliers to find a useful match within a defined time. Measure the request funnel and the available supply separately. A response, a booking and a completed transaction are different outcomes; counting them as one “match” hides where the marketplace fails.

Specify what counts as eligible demand

Choose the unit: a unique job request, room night, delivery or another customer task. Write exclusions such as test requests or duplicates before looking at the result. A hard request that went unfilled should not disappear merely because it makes the rate worse.

a16z separates match rate, market depth and time to match. Your measurement window and successful event must reflect the product promise, not whichever event is easiest to log.

Work a fictional request funnel

Fictional service marketplace: 100 unique eligible requests enter a cohort. Eighty receive a suitable quote within 24 hours, sixty book and fifty-four complete within the defined service window. Six bookings cancel. All requests have had enough time to reach the reporting cutoff.

The quote-to-book rate is 60 / 80 = 75%. Booking cancellations are 6 / 60 = 10%. These use different denominators from the request-level rates. Put the denominator in the metric name or note.

Measure supply in the unit that can clear demand

Suppose suppliers offered eighty genuinely available service slots matching the location, skill and time requirements. Fifty-four delivered jobs use 67.5% of those slots. Counting supplier profiles instead would not tell you whether usable capacity existed.

Do not assume every supplier listing is available or that one supplier can fill unlimited simultaneous requests. Match supply and demand by the constraints that change the experience: time, geography, service type and price range.

Inspect the zeros and the waiting time

Twenty requests had no suitable quote within 24 hours. Twenty quoted requests did not book. Six bookings canceled. Those are three different failure groups, requiring different evidence.

• No quote: inspect coverage, availability and matching constraints.
• Quote but no booking: investigate fit, price or buyer timing without assuming the reason.
• Cancellation: distinguish buyer, supplier and operational causes.
• Long wait: show the distribution, not only a fast average.
• Incomplete observation: keep recent cohorts pending until the window closes.

A national average can conceal a weak city or time slot. Compare meaningful segments without multiplying tiny samples into confident conclusions.

Copy the two-sided liquidity worksheet

Customer task and unique request unit:
Eligibility and exclusions, set in advance:
Request cohort and observation cutoff:
Quote/match event and time window:
Booked / completed / canceled definitions:
Counts and denominator for each rate:
Available supply slots with matching constraints:
Time-to-match distribution and pending requests:
Unfilled groups and documented reasons:
Segment, sample limits and next evidence task:
Owner and next complete-window review:

Exercise: a shorter window changes the question

Seventy of 100 eligible requests receive a match within one hour, while ninety receive one within a day. Which match rate is correct?

Answer: both can be correct: 70% within one hour and 90% within one day. They measure different service windows. Choose the window tied to the user promise and show the longer-window result as context, not a replacement that hides waiting.

Connect the metric to a founder decision

The finished worksheet should identify whether the constrained side is usable supply, eligible demand or the matching experience. More signups alone do not resolve that diagnosis. The business-model guide explains the broader model, while the marketplace investor guide serves capital research.

Keep the approved liquidity brief with your fundraising materials in Finta. Ask Aurora to help explain the funnel and the next evidence task from your reviewed counts. Confirm the data and sample limitations before using the story in investor outreach.

Finta Field Guide. By Finta Editorial Team. Sources checked November 14, 2026. Examples are original, fictional teaching exercises, not customer results, benchmarks or individualized financial, accounting, legal or tax advice.