A bad growth review starts with universal metrics. CAC, LTV, ROAS, payback, retention, NRR, conversion rate, and activation all sound precise until they are detached from the business model they are supposed to describe.
SaaS, consumer apps, marketplaces, fintech, ecommerce, usage-based infrastructure, and services-heavy businesses do not obey the same economic physics. They may all acquire customers, but the cash flows, risks, margins, retention patterns, and scaling constraints differ enough that a shared metric can distort the comparison.
In SaaS, retention and expansion can make a customer more valuable over time. A longer payback may be tolerable if the customer is durable, expands reliably, and produces high gross margin. But even inside SaaS, a self-serve SMB product and an enterprise product with heavy sales and implementation costs should not use the same growth standard.
Consumer businesses often face more terminal churn. The customer may come, monetize, and leave. In that world, LTV can still matter, but the model has less room for vague expansion optimism. The company needs sharper cohort behavior, contribution margin, and repeat-purchase evidence.
Marketplaces add liquidity. Acquisition on one side can be wasteful if the other side is not ready. A buyer acquired in a thin market is different from a buyer acquired where supply density is strong. The growth metric has to understand balance and customer count together.
Fintech adds balances, risk, regulation, funding cost, fraud, and capital constraints. A customer who looks cheap to acquire may be expensive to serve or risky to hold. Revenue can arrive through spreads, fees, interchange, assets, or usage, each with different timing and quality.
This is why a business-model metric map should come before the dashboard. The map should identify how value is created, when cash is collected, what costs scale with usage, what risks sit behind revenue, what retention means, and where capital gets trapped.
Once that map exists, the company can choose metrics with intent. SaaS may emphasize net revenue retention and net payback. Ecommerce may emphasize contribution margin after fulfillment and returns. Marketplaces may pair payback with liquidity and density. Fintech may pair growth with risk-adjusted return and cost of capital.
The mistake is treating metrics as a professional identity. A growth leader who says every business should use the same payback threshold is not being rigorous. A CFO who rejects all long-payback growth is also not being rigorous. The right answer depends on the economic engine.
The operator test: before reviewing growth performance, ask what must be true for a new customer to be economically good in this specific business. If the dashboard does not reflect that answer, the metrics are decoration.
The same business can also change metric regimes as it grows. A startup looking for a repeatable motion may care most about learning velocity and directional payback. A scaling company needs segment-level economics and capacity planning. A company approaching profitability needs tighter liquidity discipline and cleaner contribution analysis.
This is where benchmark culture gets companies into trouble. A famous payback target from one category becomes an imported rule in another. The rule sounds disciplined, but it may ignore sales cycle, contract length, gross margin, onboarding burden, usage economics, or the fact that one company is buying demand while another is building liquidity.
A business-model metric map should be reviewed whenever the company adds a new motion. Moving from self-serve to enterprise, from one-sided SaaS to marketplace dynamics, from subscription to usage-based pricing, or from software to services-heavy deployment changes what good growth means. The dashboard has to change with the motion.
The operator habit is to ask, 'what is the economic unit here?' Sometimes the unit is an account. Sometimes it is a user, transaction, seat, balance, merchant, location, workflow, or deployed customer. Until the unit is clear, the growth metric will drift between languages and create arguments that sound analytical but are really category confusion.
This also applies inside a single customer journey. Acquisition may be governed by payback. Activation may be governed by product usage. Expansion may be governed by account health, while retention depends on value realization. The company does not need one heroic metric. It needs a small set of metrics that match the economic job being done at each stage.
The mistake to avoid is metric cosplay: adopting the vocabulary of a company with different economics because it sounds sophisticated. A marketplace borrowing SaaS NRR language too casually, a SaaS company borrowing consumer ROAS habits, or a fintech product treating customer acquisition like low-risk ecommerce can all end up with dashboards that look modern and decisions that are structurally confused.
A usage-based infrastructure company is a good example. A customer can look expensive to acquire, then become excellent as workloads expand. The same customer can also create heavy cloud cost, support demand, and sales-engineering work before usage matures. A SaaS-style logo payback view will miss both the upside and the cost curve.
The cleanest review habit is to name the metric that is not allowed to decide. In one business, ROAS may be useful for campaign tuning but not for budget expansion. In another, NRR may be useful for account health but too late for acquisition decisions. Saying this out loud prevents the strongest-looking chart from taking over the meeting.
The best sign that the map is working is that it changes a decision. If a marketplace is supply-constrained, acquisition spend on buyers may be capped until density improves. If an enterprise product is implementation-constrained, demand generation may pause while onboarding capacity catches up. The metric follows the constraint, not the other way around.
This is part 2 of 10 in The Capital Allocation Theory of Growth.