Unit economics measure the contribution margin per unit: the amount a single customer, transaction, or project contributes to the business after deducting all directly attributable variable costs. For growing startups, they are primarily a communication tool: the foundation on which a loss-making company can explain why the P&L loss is not proof of a broken business model.
- Positive contribution margin per unit means: the business model works. The loss comes from overhead and growth investments, not from the unit level.
- Break-even: At what volume does the contribution margin cover fixed costs, and how far is the company from that point today? That is the first question every experienced investor asks.
- Three cost layers must be separated: unit level, fixed costs, growth investment. A company that does not draw this distinction clearly leaves investors to interpret the numbers themselves.
- Most common mistakes: confusing gross margin with contribution margin, reporting CAC (customer acquisition cost) as a blended average instead of by channel, hiding growth costs inside overhead.
Why do growing startups show a loss in their P&L?
A startup in a growth phase almost always shows a loss in its P&L. This is structural: the company is investing capital today to win more customers, projects, or transactions tomorrow. Someone reading the result without this context sees a company burning money. Someone with the context sees a company buying growth.
The problem: investors do not automatically have this context. They receive a P&L with a negative result and have to decide for themselves what it means. If the startup does not actively supply the context, the investor fills the gap themselves. Unit economics provide that context. They break the loss into its components and show which part is structurally necessary and which part is a deliberate growth decision.
Which three cost layers do investors distinguish?
A startup P&L can be broken down into three layers that say very different things about the company:
| Layer | What it measures | Investor's question |
|---|---|---|
| Unit level | Contribution margin per unit after variable costs | Does the business model work per unit? |
| Fixed cost layer | Core team, infrastructure, rent, accounting, legal and base licences: independent of volume | At what volume do the contributions cover overhead? |
| Growth layer | Marketing, sales, growth hiring, scaling-enabling IT: costs for future volume, not current | What would profit be without growth investments? |
A startup with a positive contribution margin at the unit level that still makes a loss is a fundamentally different company from one that is already negative at the unit level. The loss is the same number, but it has a different meaning.
The interactive calculator below translates these three layers into concrete numbers: contribution margin, overhead break-even, and total break-even can be calculated directly for your own business model.
Which unit is relevant for which business model?
The core principle applies to every business model: contribution margin per unit, overhead break-even, scalability. What changes is the definition of the unit and the relevant metrics. In a subscription model the unit is the customer, the key metric MRR (monthly recurring revenue). In a marketplace the unit is the transaction, the key metric GMV (gross merchandise value before platform fee). In hardware it is the device sold, with COGS (cost of goods sold) as the main cost driver. In project business it is the contract, with project margin. The table below shows the typical ranges per model:
| Model | The unit | Variable costs | Typical contribution margin |
|---|---|---|---|
| SaaS / Subscription | Customer / Contract | Hosting, support, onboarding | 65–85% |
| Marketplace | Transaction / GMV | Payment, trust & safety, fulfilment | 40–65% of take rate |
| Hardware + Software | Device + Subscription | COGS hardware, logistics, support | 20–45% (hardware), 60–75% (software component) |
| B2B Services / Project Business | Project / Contract | Direct personnel costs, subcontractors | 25–50% |
| E-Commerce | Order | COGS, fulfilment, returns | 15–40% |
Contribution margins differ significantly by model. A hardware startup with 35% contribution margin is not at a disadvantage against a SaaS company at 75%. The right benchmark is the model-specific peer group, not a universal figure. For project-based models with long durations and upfront costs, additional valuation logic applies, covered in the article Project Controlling in Scale-Ups: The KPIs Investors Expect.
Calculator: Run the three layers with your own numbers
The calculator below puts the three cost layers into concrete numbers. Select a business model, adjust the parameters to your situation, and see directly: at what unit count overhead is covered, when the company reaches break-even, and how growth investments affect the result.
Concrete example: EUR 50,000 monthly loss explained
A B2B SaaS startup with 500 active customers, EUR 200 MRR per customer, and EUR 60 in direct variable costs per customer (hosting, support, onboarding) looks like this (matching the calculator's default values above):
| Position | Month total | Per customer |
|---|---|---|
| Revenue | EUR 100,000 | EUR 200 |
| Variable costs | EUR 30,000 | EUR 60 |
| Contribution margin | EUR 70,000 | EUR 140 (70%) |
| Fixed costs / overhead | EUR 80,000 | |
| Growth investment (marketing / sales) | EUR 40,000 | |
| Net P&L | EUR -50,000 |
The company loses EUR 50,000 per month. At the unit level the model is healthy: each customer contributes EUR 140 at a 70% contribution margin, within the SaaS benchmark range. The loss originates entirely from two other sources:
- Overhead: EUR 80,000 for core team and infrastructure. Break-even at this level at 572 customers (EUR 80,000 / EUR 140).
- Growth investment: EUR 40,000 for marketing and sales. Full break-even at 858 customers (EUR 120,000 total costs / EUR 140).
At 500 customers the company loses money because it is not yet large enough and because it is actively growing. The model shows what happens at 858 customers: break-even. If growth continues and unit economics remain stable, this is a question of time, not of model validation.
The line that matters in any investor conversation is not "we will eventually be profitable". It is: "Our contribution per customer is EUR 140 at a 70% margin. Overhead break-even is at 572 customers. We are at 500 today."
How do unit economics work in project business?
The three-layer model applies to every business model. In project business, whether engineering services, Cleantech EPC (Engineering, Procurement & Construction), system integration, or defence technology, the metrics look different at each layer. The underlying logic is identical.
| Layer | Subscription / SaaS | Project business |
|---|---|---|
| Unit level | Contribution margin per unit (revenue per unit minus variable costs; in subscription models MRR minus per-unit service costs) | Project margin per contract (contract value minus direct project costs: personnel, materials, subcontractors) |
| Fixed cost layer | Core team, infrastructure, rent, accounting, legal and base licences, independent of volume | Core team on bench time, admin, rent, accounting, compliance and internal tools, independent of project volume |
| Growth layer | Marketing, sales, growth hiring, scaling-enabling IT: costs for future customers | Presales effort, proposal creation, business development, growth hiring for future capacity |
In project business, break-even at the fixed cost layer is driven by utilisation, not unit count. If an engineering team of twenty people covers its overhead costs at 75% utilisation, that is the same logic as "1,000 customers to reach overhead break-even" in a subscription model. The question investors ask is the same: at what volume or utilisation rate does the operating business cover overhead?
An important addition: in project business, revenue and profit under the German Commercial Code (HGB) only arise at client acceptance of the work (completed contract method). Costs that can be capitalised during the project are carried as work in progress on the balance sheet, offsetting the corresponding expenses in the income statement: when correctly booked, the management accounts show a neutral result throughout project execution, not a loss. Only when work-in-progress entries are missing or incorrectly booked do apparent loss-months appear, followed by a profit spike at acceptance. This accounting problem becomes an information problem for investors and banks. Investors in project companies therefore ask about the project margin per contract and the correct recognition of work in progress, not about the period result. The accounting mechanics behind this, including a checklist and worked example, are explained in the article Work-in-Progress in Project Business: Balance Sheet, Management Accounts and Revenue Recognition under HGB.
Investor communication for project businesses follows the same structure as for subscription models, with a different metric language: first show the project margin for representative contracts. Then show what utilisation covers overhead and how far away you are. Finally, show what you invest in presales and business development and what win rate and average contract value you generate from it.
What do investors read from unit economics?
The first question an experienced investor asks is not "are you profitable?". It is: "At how many units will this company break even, and how far away are you?" That is the break-even question, and unit economics provide the answer.
The mechanics are simple: the contribution margin per unit is known. The overhead is known. Dividing overhead by contribution margin per unit gives the unit count at which the company covers its overhead. Adding growth investment to overhead gives the second break-even. Any investor can do this calculation in two minutes. A company that does not communicate this proactively forces the investor to do it.
The second thing investors examine is operating leverage: is overhead growing more slowly than total contribution margin? If a company sells 50% more units and overhead rises by only 15%, it approaches break-even faster with each growth quarter. That is the scalability thesis in numbers.
If overhead and contribution grow at the same rate, break-even does not move. If overhead grows faster than contribution, the model becomes more expensive rather than more efficient at scale. Both are signals investors explicitly look for.
The third question: is the contribution margin stable as the company grows? Declining contribution margins at increasing volumes point to structural problems: pricing pressure with larger customers, rising variable costs through complexity, or a model that deteriorates with scale. Rising margins with growth are a strong signal in the opposite direction.
How do unit economics develop at scale?
A snapshot of current unit economics answers one question: does the model work today? Investors from Series B onwards regularly ask a second: will unit economics be better or worse when the company is three times the size? This question is fundamentally different from the first.
Models in which contribution margin per unit and acquisition costs remain stable or improve at higher volume have a structural advantage. Existing relationships generate repeat business, require less acquisition effort, or grow into larger contracts. Expansion revenue from existing customers carries no acquisition cost: every euro of revenue that arises without new acquisition improves unit economics at the company level. This applies to subscription models with expansion logic, to project businesses with returning clients, and to marketplaces where active users generate more transactions than new entrants.
The opposite applies to models where contribution margin per unit falls with growth, because pricing pressure, rising service costs, or declining repeat rates make the contribution of later units smaller than that of the early ones.
Acquisition costs rise at scale for three structural reasons. First, the earliest customers or contracts are the easiest to win: they are actively seeking the solution, arrive through referrals, and fit the offering well. When that pool is exhausted, customers come from more costly channels with longer decision processes. Second, acquisition channels saturate: more competitors bid on the same keywords and audiences, making paid channels more expensive even when conversion rates remain stable. Third, organisational complexity grows: longer approval chains, more handoffs between marketing and sales, and looser lead qualification extend sales cycles and raise the cost of each customer won.
At the same time, churn or repeat rates often deteriorate because fit is weaker. In project business this appears as margin pressure: when the ideal projects are taken, the company accepts lower-margin contracts to maintain utilisation. The result is the same as in a subscription model: the contribution margin per unit in the existing base is higher than that of new units. The numbers look better early than they hold up at scale.
For this reason, investors compare not only today's value against a benchmark, but older cohorts or contract vintages against newer ones. If earlier periods show better contribution margins than current ones, that is a signal: acquisition is getting more expensive, or the model is coming under margin pressure.
Conversely, improvement from period to period, even if absolute values are still below the benchmark, is a more credible signal than a strong snapshot without any discernible direction. Direction beats snapshot.
What mistakes do startups make in their presentation?
Confusing contribution margin with gross margin
Gross margin in many startup P&Ls includes costs that are not truly variable: parts of the engineering team, support costs that do not scale directly with each new unit. Contribution margin is stricter: only costs that arise directly and proportionally with each additional unit. Using gross margin as the unit economics basis overstates the unit economics. In hardware models the opposite occurs: COGS are set too low because warranty provisions, returns, and shipping costs are missing. A gross margin above 75% on a physical product is almost always a calculation error, not a competitive advantage.
Blended acquisition costs instead of channel-specific costs
Dividing all acquisition costs by all new customers produces an average that says little. In practice, organic channels (SEO, word-of-mouth, inbound) carry acquisition costs near zero. Paid channels (Google Ads, LinkedIn, events) can generate acquisition costs of EUR 2,000 to EUR 20,000 per customer. A blended CAC of EUR 800 may mean the organic channel is working well and pulling the average down. Or it may mean the paid channel is so expensive that growth is not scalable once the organic effect fades. In project business and marketplaces the same logic applies at the channel level: which contracts arrive through referrals with zero presales effort, and which require EUR 15,000 in proposal work per won contract? Experienced investors ask for the equivalent of channel-specific acquisition costs.
Presenting retention metrics without sufficient data
In early stages there is rarely sufficient data on the long-term behaviour of customers, clients, or marketplace participants. Presenting these metrics anyway means working with assumptions. A subscription startup with 6 months of data does not know how churn develops in month 18. A project company with 8 completed contracts does not know what percentage of clients will commission a second project. A marketplace with 3 months of operation has no valid data on the repeat rate of buyers or sellers. Investors know this. Better: be transparent about the data situation, show observed trends explicitly, and present projections as scenarios rather than facts.
Counting growth costs as overhead
Marketing and sales spend is not a fixed cost in the classical sense but a decision. A startup that stops growing tomorrow can reduce these costs immediately. Counting them as overhead hides the actual growth investment and makes the profitability question unnecessarily opaque. The question "what would our result be without growth investments?" should be answerable by any startup without hesitation.
Understating acquisition costs
Acquisition costs are systematically understated across business models. In subscription models, salaries of sales and marketing staff, tool costs (CRM systems, ad tech), event costs, and management time allocated to sales activities are regularly missing. In project business, presales effort is missing: proposal creation, scoping calls, demos, and the time of experienced consultants during acquisition phases. In hardware models, distribution and dealer margins as well as return and warranty costs that arise in calculating the effective acquisition cost are missing. In personnel-intensive B2B models, the advertising budget is often the smallest line item.
How to structure the unit economics argument in an investor conversation
The structure that works does not begin with the total loss. It begins at the lowest level and builds upward:
- 1Unit level first: contribution margin per unit, contribution margin as a percentage, acquisition costs and payback period.
- 2Overhead separately: what does the company cost independent of volume? At what volume does the running contribution cover this overhead?
- 3Make growth investment transparent: what is being invested today in future volume? How does this amount relate to the expected return over the next growth period?
- 4Show the break-even calculation: not as a promise, but as the mathematical consequence of the unit economics. At x units or y utilisation, the contribution margin covers overhead and growth costs.
This structure changes the dynamic in the conversation. The investor is no longer sitting in front of a loss asking when it will stop. They are sitting in front of a mechanism asking whether they trust it. That is a different question, and one that can be answered on the basis of numbers.
When good unit economics cannot rescue the argument
There are situations where the numbers look good and the company is still not an attractive investment:
- Positive contribution margin but little growth: if contract volume, customer count, or transactions stagnate despite growth investment, that points to an acquisition problem that good unit economics cannot solve.
- High churn or low repeat engagement: in subscription models a short payback period counts for little if customers leave before it amortises. In project business the same applies to a low repeat engagement rate: if every contract must be newly acquired, effective acquisition costs rise structurally.
- Contribution margin falls with growth: if higher volume means the margin per unit declines, the entire scalability promise is in question.
- Payback period too long for capital-intensive growth: if 24 months are needed to recover acquisition costs while aggressive growth is planned, a cash flow problem arises that the strength of unit economics at the unit level does not automatically solve.
How unit economics belong in ongoing board reporting
Presenting unit economics during fundraising is the first step. Investors who have committed want to see these numbers develop. That means contribution margin, acquisition cost payback, and cohort retention become part of regular board reporting, not just the pitch phase. Companies that are systematic here build investor trust over time. More on the structure of board reporting is in the article Board Reporting for Startups: What Investors Really Want to See. For project-based business models with their own reporting logic, specific requirements are addressed in the article Project Controlling in Scale-Ups.