Gross bookings and volume
The total money that moves across a platform, which the platform does not keep. The number marketing quotes and the number valuation cannot use.
A marketplace sits between a buyer and a seller. Gross bookings, gross merchandise value, total payment volume: whatever it is called, it is the full transaction amount, most of which belongs to the seller. Revenue is the slice the platform retains.
The gap is enormous and it is where most confusion about platform businesses comes from. Uber processed $58,022,000,000 of gross bookings in the quarter we examined and recognized $14,191,000,000 of revenue, which is 24.5%. PayPal moved $486,400,000,000 and kept 1.61%. Both are real figures about the same company and only the second one can be multiplied by a margin.
Volume still matters, for two reasons. It is the cleanest measure of whether the platform is winning share, since it is harder to flatter than revenue. And it sets the ceiling on everything else: a platform doing 3,867 million trips a quarter has a different set of available futures than one doing a tenth of that. Just do not confuse the size of the pipe with the size of the business.
Where this distinction decides a thesis: ask what the price assumes.
Take rate
Revenue divided by volume. The share of each transaction the platform keeps, and the most important single number about its bargaining power.
Take rate is where competitive position shows up as arithmetic. A platform both sides need can raise its cut; one competing on price cannot. Because the denominator is usually growing, the rate can fall while revenue rises, which is exactly the pattern a business under pressure wants you to see.
PayPal's take rate was 1.61% in the quarter we examined, down 7 basis points year over year while volume grew 10%. That combination is the report in miniature: more money moving, less kept per dollar. Uber's equivalent, revenue over gross bookings, was 24.5%, a far higher cut for a service with far fewer substitutes at the moment of use.
Mix shifts move the rate too, and they are not the same as a pricing loss. A platform growing fastest in its cheapest product will show a declining take rate without losing a single pricing negotiation. Separating mix from price is the work, and the disclosure that lets you do it is often missing.
Take rate compression as the whole argument: the PayPal report.
Transaction margin
Revenue minus the direct cost of processing the transaction. For a payments business it is closer to gross profit than revenue is, and it is the figure that should be growing.
Payments revenue carries a large variable cost: the network fees, the processing, the losses. Transaction margin removes those and leaves what the business has to cover its fixed costs and earn a return. It is the honest top line.
PayPal's transaction margin was 44.9% in the quarter we examined against 46.4% a year earlier, and in dollars it grew 1% on revenue of $8,682,000,000. A payments company whose dollar transaction margin is roughly flat is not growing, regardless of what its revenue line does, because revenue growth bought with margin is a transfer rather than a gain.
The reason to watch the dollar figure as well as the percentage is interest. A payments business holding customer balances earns interest on them, which lands in transaction margin and has nothing to do with the payments business. When rates move, that line moves, and the underlying trend is whatever is left after you remove it.
The decomposition: PayPal, and on why a margin and a volume can move in opposite directions, the one-time item.
The Rule of 40
Revenue growth plus profit margin. A shorthand for whether a company is allowed to be unprofitable, and a crude one.
The idea is that growth and profitability are substitutes. A business growing forty percent with no margin and one growing ten percent at thirty percent margins are both acceptable; a business growing ten percent at no margin is not. Add the two numbers and compare to forty.
SoFi scored 70 in the quarter we examined, from adjusted revenue growth of 40.5% and an adjusted margin of 30%. That is a genuinely strong reading, and it is also the single most quoted number in that company's investor materials, which is a reason to look harder at both inputs rather than less hard.
The weaknesses are worth stating because the metric is used as though it had none. Which margin? Adjusted EBITDA flatters almost everyone, and stock-based compensation is a real cost excluded from it. Which growth? Acquired growth counts the same as organic. And the two components are not interchangeable in the way the addition implies: growth can stop, margin generally does not resume. Treat it as a screen, never as a valuation.
What adjusted figures leave out: non-cash is not free, and the score in context in the SoFi report.
Cohort economics
Tracking each year's customers separately for the rest of their lives instead of looking at the total. The only way to see whether growth is being bought or earned.
An aggregate customer count hides everything that matters. A business adding a million customers a year while losing eight hundred thousand looks identical, in total, to one adding two hundred thousand and keeping them. A cohort view follows the group that joined in a given period and asks what happened to their number and their spending since.
The healthy pattern is revenue per cohort rising over time, which means existing customers deepen faster than they leave. SoFi reports a version of this through products per member: 24.4 million products against 15.8 million members, with 51% of new products taken by existing members. Membership grew 35% over the year, which is the aggregate; the cross-buy figure is what tells you the aggregate is not purely acquisition.
What to look for, and what to be suspicious of when it is absent. Companies with good cohort curves publish them, because the chart is the argument. Companies that publish only totals, or only the most recent cohort, or a chart with no axis labels, are usually declining to show you a curve that bends the wrong way.
Why totals mislead: growth is not a virtue, with the member data in the SoFi report.
DISCLAIMER ·
Education, not advice. Every example above is simplified to expose the mechanics. Real returns are volatile,
taxes and fees vary by country and account, and none of this considers your circumstances. 2 Comma Investor is not a
registered investment adviser or broker-dealer. Do your own research and consult a licensed professional before
acting.
Full disclosures →