Applying SaaS metrics to a platform business does not produce bad measurements. It produces measurements of the wrong thing. CAC, churn, and LTV are not imprecise when applied to platforms; they are precise about something other than what platform health depends on. The measurement framework is structurally misaligned with the business model. That misalignment is why platforms can report improving commercial metrics while the underlying markets they nominally serve fail to clear.
The problem is not aggregation, though aggregation makes it worse. Commercial metrics measure outcomes at the account level while platform health is determined at the market level, meaning whether specific participants can find acceptable counterparties within reasonable time and terms in the specific market they operate in. Most platforms fail not from lack of users but from failure to achieve liquidity in the particular micro-markets they serve, meaning the probability that a transaction occurs in a given geography, category, and time window. That failure is undetectable at the account level until it has compounded into a problem visible in the aggregates. The operator applies the standard fixes: more acquisition spend, a retention campaign, a pricing experiment. None of it moves the underlying problem, because none of it reaches the level at which the problem exists.
The unit problem
In a SaaS business, value is delivered to accounts and the account is the right unit of analysis. One customer's experience does not depend on other customers: if the product is good and the fit is sound, the relationship holds independently of who else is on the platform. CAC, churn, and LTV describe the commercial relationship accurately because the commercial relationship is the thing the business delivers. Platforms are organized differently. Value is produced through matching across a bilateral relationship, and a buyer's experience in a given segment depends entirely on whether there are sufficient sellers of the right type available in that segment, not just on the platform in aggregate.
Most platforms are collections of partially connected micro-markets defined by geography, category, time window, price band, or other constraints that are binding for any given transaction. Liquidity forms within segments and transfers across them only when matching constraints overlap sufficiently. A buyer looking for a specific type of service within a specific radius on a given schedule does not benefit from sellers outside those constraints, even if both are paying subscribers on the same platform. Standard commercial metrics aggregate across all segments as though they were a uniform customer base. The resulting number is a composite of markets at different stages of maturity, competitive conditions, and match feasibility, averaged into a figure that accurately describes none of them.
Market clearing
Fill rate (the proportion of search sessions resulting in a transaction), utilization rate (the proportion of available supply that transacts), and time-to-match are the metrics that precede commercial performance rather than follow it. They measure whether the market is clearing in a given segment, meaning whether participants can find acceptable counterparties within the constraints that define that market. Commercial outcomes are lagging: they reflect the accumulated results of whether markets cleared in the prior period. An account that churns this quarter is reporting on market conditions from several months ago. Standard commercial metrics organized around accounts describe the outputs of a process that operates at the interaction level, and that process is not visible from its outputs alone.
The causal asymmetry between SaaS and platforms is structural. In SaaS, losing some customers does not reduce the product's value for those who remain: each account's experience is independent. In platforms, when match quality drops in a segment, marginal participants leave, which removes potential counterparties, which lowers match quality further for those who remain. This reinforcing dynamic means the market can move from thin to empty faster than account-level metrics register. When interaction metrics are improving while commercial metrics lag, the platform may be building durable liquidity whose commercial returns will follow with pricing iteration. When commercial metrics look strong while interaction metrics are flat or deteriorating, the most likely explanations are mix-shifts toward already-liquid segments, multi-homing optionality sustaining nominal subscriptions, or activity maintained by incentives that do not reflect genuine market clearing.
A rigorous analyst does segment-level cohort analysis, and the segmentation principle applies to platforms as it does to any business. Proper disaggregation partially addresses the aggregation problem. But per-segment commercial metrics still cannot reveal whether segments are clearing or why, because commercial outcomes sit downstream of the interaction-level processes that produce them. The causal layer cannot be inferred from the outcome layer, which means the measurement system has to be organized differently, not just disaggregated. The right unit of analysis for diagnosing platform health is not the account or even the segment, but the interaction: whether specific participants, looking for specific counterparties, in specific local markets, are finding them.
The distortion pattern
Matching complexity is the primary driver of segment-level performance variance, and it produces a predictable distortion in aggregate CAC. In low-complexity markets, few attributes need to align for a viable interaction, so modest increases in local participation quickly improve match times and completion rates. In high-complexity markets, origin, destination, timing, equipment type, compliance requirements, and behavioral norms can each become binding constraints, requiring significantly more scale to reach comparable match quality. Early liquidity concentrates in the easiest sub-markets first, so the product early participants experience is often not the product the platform eventually delivers. Aggregate CAC appears to improve as acquisition concentrates in dense, easily-converted segments, even as the marginal cost of achieving liquidity in the harder sub-markets rises and the long tail of those markets remains thin.
In competitive markets where multi-homing is rational, the distortion extends into churn in a specific way. In SaaS, payment is closely tied to use: customers rarely pay for tools they do not use, so account retention is a reasonable proxy for competitive position. In platforms selling access to discovery and matching, participants can rationally maintain subscriptions to several venues because each covers a different slice of the network. A participant can remain a paying customer while shifting transaction volume elsewhere, leaving account retention high while competitive position weakens. Revenue-based retention captures this better than account retention, but even revenue can reflect macro cycles rather than competitive dynamics. The better diagnostic is how participants are allocating activity across venues and whether the platform's share of completed transactions is rising or falling.
LTV is distorted through a channel SaaS pricing theory does not account for. In SaaS, pricing captures a share of value delivered to the paying customer. In platforms, pricing is market design: a take rate or fee structure changes who participates on each side, which changes the match quality available to the other. Subsidies and take rate decisions are not purely financial choices; they determine whether the market clears. LTV is bounded not just by retention but by market structure and by the cost of maintaining the market's reliability, where fraud prevention, enforcement, dispute resolution, and support function as marginal costs rather than overhead, and determine whether gross growth in transaction volume produces contribution.
Liquidity before scale
Gate acquisition by segment-level interaction thresholds rather than aggregate commercial targets. Where fill rates and utilization are already acceptable, additional scale improves the match pool and lowers time-to-match. Where they are not, additional acquisition produces more participants without producing more matches, and the platform appears to be growing while the clearing problem stays the same. Density-building before scaling means narrowing scope, seeding one side, subsidizing early participation, increasing standardization, or adding operational support in thin segments. The principle is to buy liquidity rather than logos.
SaaS benchmarks regain interpretive power in this framework once segments are liquid. In mature segments where matching performance has stabilized, ratios like LTV:CAC and payback periods can be evaluated per side and per segment much as they would be in SaaS. During cold start, insisting on those thresholds before a segment has reached minimum viable liquidity is the wrong test. Aggregate metrics are largely uninformative in this phase; what matters is whether individual segment networks are achieving the density that produces self-sustaining matches. The relevant question is whether interaction metrics are converging toward the clearing threshold and whether reliability costs are moving toward a sustainable share of contribution.
Wrong instruments
CAC, churn, and LTV are not wrong metrics for platforms. They are the right metrics for a different business model, applied to one where the causal structure is different. A SaaS company delivers a product whose quality it controls directly; its commercial metrics describe that product's fit with each account. A platform builds the conditions under which participants transact, and the quality of those conditions depends on who showed up, in what proportions, and in which local markets, none of which the platform controls directly. The commercial metrics describe the outcome of those conditions after the fact, not the conditions themselves.
The operator who reaches for the product dashboard is not measuring imprecisely. The dashboard is precise about the outcome of a process it cannot show. Platform strategy requires making that process legible: measuring how interactions form in each segment, tracking whether clearing rates are converging toward liquidity, and directing growth only where the underlying market is real. That discipline is market design. The aggregate commercial metrics, read without this context, look like a verdict on the business. They are a description of its densest part.