A marketplace founder pitches two audiences and keeps two promises. Sellers join because buyers are there, and buyers come back because sellers are good. Until both sides show up in the same place at the same time, nothing works, and the numbers you show investors mean little.
That's why the tooling question for a marketplace is unusual. You need finance views that separate what passes through your account from what you earn, and a sales process for each side of the market. The parts that touch money moving between strangers need specialist infrastructure that a company operating system shouldn't pretend to be.
Liquidity comes before everything else
Andreessen Horowitz's 13 metrics for marketplace companies measures liquidity in three ways. Match rate shows how often the two sides find each other, such as fill rates in a jobs marketplace. Market depth asks whether there's enough supply to fit what buyers need. Time to match asks how long a connection takes. The same piece suggests counting the zeros, the searches or requests that ended with no match.
None of those numbers lives in Nautis by default, and that's fine. They come from your own product database. What Nautis can do is make the team look at them on a schedule and act on them.
- Add match rate, time to match and zero-result searches as custom metrics on the Metrics tab of a GTM Planner workspace, with a target for each and a snapshot every week.
- Create one workspace per launch market, since the guide recommends one workspace per segment. A city that's 60 days old and one that's two years old shouldn't share a health score.
- When a metric slips, design a fix on the Experiments tab: "If we guarantee new hosts three bookings in their first month, we expect time to first booking to halve, because hosts who earn early stay."
GMV, net revenue and take rate in one table
The a16z piece is blunt that GMV does not equal revenue. GMV is the total value of what transacts through the marketplace in a period. Revenue is the share you keep, mostly from transaction fees and sometimes from ads or sponsorship. The take rate links the two. Mixing them up is one of the fastest ways to lose an investor's trust in a first meeting.
| Number | What it means | Where it goes in Nautis | Common mistake |
|---|---|---|---|
| GMV | Total value transacted through the marketplace in the period | The Growth section of the monthly investor snapshot, written in by hand | Calling it revenue in a deck or a model |
| Net revenue | Fees, commissions and other income you keep | The revenue field of the Startup Health check-in, if you label it clearly (see below) | Counting the seller's share because it touched your bank account |
| Take rate | Net revenue divided by GMV | Unit price of a commission revenue item in the Financial Planner | Raising it before liquidity is solid |
| Supply concentration | Share of GMV from your top sellers | Snapshot Growth section, which already asks about revenue concentration | Ignoring it until a big seller leaves |
| Cohort retention | Whether newer buyer cohorts come back more often than older ones | Your analytics tool, summarised in the snapshot | Measuring logins instead of a core action such as a booking |
A note on Startup Health. Its check-in asks for monthly recurring revenue, and its guide says to leave one-off sales out. Most marketplace income is transactional. The honest workaround is to pick one rule, such as "monthly net revenue, excluding GMV", use it every month, and say so in the snapshot's Overview. Runway and burn still come out right, because they depend on cash and spend. Don't put GMV in that field, ever. At a 10% take rate it would make you look ten times bigger than you are, and the growth alerts would mean nothing.
Choosing a take rate you won't regret
Bill Gurley's essay A Rake Too Far argues that the most dangerous strategy for a platform is to price too high. A high rake pushes suppliers to look elsewhere and leaves room for a cheaper rival. His examples: eBay's marketplace rake ran just under 10% in 2011, and Amazon's marketplace fees ranged from about 6% to 15% by category. oDesk cut its commission from 30% to 10%, which cut revenue to about a third at first, and it later passed its nearest competitor.
You don't have to agree with Gurley to use his test. Before you change your fee, write the decision down where you'll have to face it later.
- Open the Financial Planner and model your commission as a revenue item, with the unit price set to average order value times take rate.
- Clone the project, change the take rate in the clone, and lower forecast volume to reflect sellers you expect to lose. Use Compare to put both versions side by side.
- Log the call in the Decision Loop under Finance, mark whether it's reversible, and set a 90-day review with a measurable expected outcome such as "net revenue up 15% with seller churn under 5%".
- At the review, record what happened. If you logged it with high confidence and it went worse, the Insights tab will count it.
Two pipelines, one for each side
Early marketplaces usually grow supply by hand. Someone calls restaurants, recruits tutors or signs up warehouses one at a time. That's a sales process, and it deserves its own pipeline instead of a spreadsheet tab.
Supply onboarding
In the Leads CRM, create a pipeline called Supply with stages such as Contacted, Applied, Verified, First listing live and First sale, and mark the last one as Closed Won. Track source (Referral, Event, Cold Email) so you learn which channel brings sellers who stay. An automation rule can raise priority when a seller reaches Verified, so nobody forgets to help them publish.
Seller applications
Build the application in Custom forms with Stepper Mode, a File Upload field for a portfolio or licence, and a short URL or QR code for events. Applications arrive in the Entries table, which exports to CSV. Keep identity checks out of it: verifying who a seller is belongs with your payments provider.
Demand partnerships
A second pipeline holds the demand side's larger deals: an employer buying shifts in bulk, a property manager listing every unit. Different stages, different owners, and the pipeline overview shows each side's value separately.
Money between strangers needs specialist rails
Stripe's guide to marketplace payments lists what a payments provider handles for a platform: verifying seller identities to meet know-your-customer rules, payouts to sellers in their local currency, splitting payments between you and the seller, chargebacks and refunds, and tax reporting. It also notes that payment transactions come with legal and regulatory requirements. A startup should not rebuild any of that.
Nautis doesn't touch that flow. Its Invoices module bills your own customers, for example a supplier paying a listing subscription, and payment is collected outside Nautis. The payouts in Employee management are for your staff, and Process only marks a run as paid. Neither one moves money to sellers.
- A marketplace payments provider for split payments, seller verification and payouts.
- Your own product and database for search, matching, reviews and messaging between users.
- Product analytics for match rate, cohorts and the core-action retention a16z describes.
- Trust and safety tooling for fraud checks and content moderation as volume grows.
- Accounting software and an accountant who know how to book GMV and revenue correctly.
Month two of a new city launch
Picture a home-services marketplace with five people, live in its first city and opening a second. In week one, the ops lead sees in the Supply pipeline that 40 cleaners applied but only 11 reached First listing live, and most stalled at Verified. She adds an automation rule and books short onboarding calls through a Meetings event type with a 15-minute duration.
In week two, the founder notices the top five cleaners now account for half of GMV in city one and writes that into the snapshot's Growth section, along with a plan to recruit more. In week three, a spring-cleaning push for buyers is planned in the Social calendar, and links carry UTM tags so the team can see which neighbourhood posts brought bookings. In week four, the morning briefing shows runway at seven months after the second-city spend. The founder opens a Decision Loop entry on whether to pause city two.
Where a marketplace should look elsewhere
- Your bottleneck is matching or search quality. That's product and data work. No operating system helps with ranking.
- You need sellers to log in and manage their own listings. Build that into your product. Nautis seats are for your team, and the plan sets the number: 5, 10 or 25.
- You process seller payouts. Use a payments provider built for platforms, as above.
- You're one founder still testing demand with a landing page. A spreadsheet is enough for now, and the $49 a month for Launch at the time of writing may be better spent on supply. Note that leads pipelines are limited on Launch, so check pricing before planning two of them.
- You want a long track record. Nautis launched publicly in July 2026.
FAQ
What software for marketplace startups do I need before launch?
A payments provider built for platforms, so seller payouts and identity checks are handled properly. A way to track both sides of supply and demand, and a product analytics tool once people start transacting. A company layer like Nautis helps once you have a team running supply sales, marketing and investor updates at the same time.
Should I report GMV or net revenue to investors?
Report both, and label them. GMV shows the size of the market you've built, while net revenue is what you earn from it. a16z's marketplace metrics piece states plainly that GMV does not equal revenue, and investors will check which one you mean.
Can Nautis pay out to my sellers?
No. Invoices bills your own customers and records payments made outside Nautis, and payroll payouts are for your employees. Use a payments provider that supports split payments and seller payouts.
Can I track supply and demand separately in Nautis?
Yes. Create one Leads CRM pipeline per side with its own stages and shared access, and read each pipeline's value and win counts in its overview. Liquidity metrics such as match rate come from your product data and can be logged as custom metrics in the GTM Planner.