Inhaltsverzeichnis
Every license model in this series works well for some products and fails quietly for others. Comparing license models is less about finding a universal answer and more about picking one deliberately, then building the discipline to defend it.
Lizenzmodelle im Vergleich: Warum Struktur Improvisation schlägt
The first article in this series argued that software license models matter more than any single pricing tactic. Value-based pricing without a structure behind it eventually turns into custom pricing for every deal, which is a slow way to lose control of a business.
Every article since has run into a version of that same problem from a different angle. Perpetual vendors that let maintenance percentages drift deal by deal lose the ability to forecast renewals. SaaS vendors that underfund the service layer lose renewals to churn instead.
Usage-based and outcome-based vendors that skip real cost monitoring lose margin quietly, one transaction at a time, long before it shows up in a quarterly review. The specific failure mode changes with the model.
The underlying cause stays the same across all four software license models: a pricing decision happened ad hoc instead of by design. The company paid for it later when it tried to scale.
If you want the fuller argument for why this matters before comparing license models at all, the opening article in this series makes the case in detail.
Lizenzmodelle im Vergleich: Vier verschiedene Beziehungen zum Kunden
Software license models are not interchangeable pricing wrappers around the same product. Each one sells something structurally different to the customer, and that difference shapes everything downstream: forecasting, renewals, and how a sales team gets paid.
Das unbefristete Lizenzmodell
The perpetual license sells ownership: a fixed version, paid for once, maintained for a fee. Maintenance fees in this model typically run 18 to 22 percent of the license price annually, a structure that has stayed largely unchanged for decades.
SAP built one of the largest enterprise software companies in the world on this structure. It still fits customers who want long-term control over a system they do not want changing under them. Our deep dive on the perpetual license model covers where it still wins.
Das SaaS-Abonnementmodell
SaaS pricing models sell access and improvement instead of ownership. Salesforce proved the category could work at enterprise scale starting in 1999, and the pattern has held since.
The real cost of a SaaS subscription is not the hosting infrastructure. It is the ongoing service investment that most companies underbudget relative to its effect on renewal rates, a point our SaaS subscription article covers at length.
Das nutzungsbasierte Lizenzierungsmodell
Usage-based licensing sells consumption. Twilio and Snowflake are the two companies most often cited for proving that pricing tied directly to consumption can work at real B2B scale, removing the natural ceiling a seat-based contract runs into.
The cost is a genuinely harder forecasting problem, for the vendor and the customer alike. The usage-based licensing article in this series walks through both sides of that trade.
Das ergebnisbasierte Lizenzierungsmodell
Outcome-based licensing sells the result itself. Fin, the AI customer service agent formerly built inside Intercom, currently prices resolutions at $0.99 each, one of the clearest live examples of outcome-based pricing in the market today.
This model delivers the cleanest value story of the four, provided the definition of “outcome” is precise enough to survive a dispute. Our outcome-based licensing article goes deeper into that definitional problem.
Service ist keine Fußnote beim Vergleich von Lizenzmodellen
A word gets used too loosely across every subscription or outcome-based deal: service. Keeping the system running, implementing it for a specific customer, and supporting custom configurations are three different jobs with three different cost drivers.
ServiceNow’s reliance on a certified partner ecosystem for implementation and configuration is one of the clearest examples of keeping those layers structurally separate. Folding all three into one undifferentiated services budget is a common and expensive mistake. Our article on the service layer breaks down how to structure it instead.
Vergleich von Lizenzmodellen während eines Übergangs von unbefristeten Lizenzen zu SaaS
Moving an existing customer base from perpetual to SaaS is a company change, not a billing change. Adobe’s own SEC filings from fiscal 2013 show a real, measurable revenue decline during its Creative Cloud transition, tied directly to the shift from upfront to ratable revenue recognition.
Autodesk’s 2016 shift away from new perpetual licenses shows the same pattern regionally. Autodesk reported an APAC revenue decline of 42 percent in the first quarter after the transition began, even as new subscription revenue grew underneath it.
Both cases show a much larger long-term payoff once the dip clears, but only if the company funds product re-engineering, implementation capacity, and the patience the transition requires.
Both also show that a forced transition is exactly when a customer is most likely to leave for a competitor. That is why reaching for discounts to prevent that churn undercuts the entire structured pricing argument this series opened with.
Our full transition article walks through the planning timeline in detail.
Vergleich von Lizenzmodellen: Warum das neueste am wenigsten nachsehend ist
Outcome-based pricing has a problem specific to itself: the price per result may be fixed, but the cost of producing that result, especially with generative AI involved, is not. EY’s research on agentic AI costs puts the increase at roughly thirtyfold since 2023 for complex, tool-using workloads.
Salesforce’s own move from flat per-conversation pricing to per-action Flex Credits for Agentforce shows a major vendor adjusting in real time to exactly this problem. Of all four models compared here, this is the one where a company can look profitable on average while quietly losing money on its most complex customers.
If you take one operational habit from comparing license models across this entire series, make it this: measure cost-to-serve and margin at the individual account level, not just as a company-wide average, regardless of which model you run. Averages hide exactly the accounts that are quietly the problem.
Eine praktische Checkliste zum Vergleich von Lizenzmodellen
Before locking in a structure, or defending one you already run, work through these questions account by account, not as a company-wide average.
- Does the price per unit (license, seat, credit, or outcome) still cover the true cost to serve that specific account, including support and implementation time?
- Can you name, in writing, exactly what counts as a billable outcome or usage unit, and would that definition survive a customer dispute?
- Is your maintenance or renewal percentage the same across every deal, or has it started drifting case by case?
- Does your service layer have its own budget and its own metrics, separate from the core license or subscription line?
- If you had to move existing customers from your current model to a different one tomorrow, do you know the cash flow and services capacity impact for the first two years?
- Are you tracking cost-to-serve at the account level, or only as a portfolio-wide average that could be hiding a handful of unprofitable customers?
Lizenzmodelle im direkten Vergleich: Eine Vergleichstabelle
| Modell | What It Sells | Revenue Predictability | Forecasting Difficulty | Real-World Example |
|---|---|---|---|---|
| Unbefristete Lizenz | Ownership of a fixed version | High per deal, lumpy over time | Niedrig | SAP |
| SaaS-Abonnement | Access plus ongoing improvement | High and recurring | Moderate, tied to churn | Salesforce |
| Nutzungsbasiert | Metered consumption | Scales with customer growth | High, volume-dependent | Twilio, Snowflake |
| Ergebnisbasiert | A defined result | Tied directly to value delivered | Highest, cost side is volatile | Fin (formerly Intercom) |
The table holds a pattern worth naming directly: predictability for the vendor tends to fall as alignment with customer value rises. Comparing license models on this axis alone often explains why a company picks a hybrid rather than a pure model.
Wichtige Fakten zu Software-Lizenzmodellen
- Perpetual license maintenance fees typically run 18 to 22 percent of the license price annually.
- Salesforce’s 1999 “No Software” campaign is widely credited with establishing SaaS as a credible B2B category, an early proof point for SaaS pricing models generally.
- Twilio and Snowflake are the two companies most often cited for proving usage-based, consumption pricing at scale in B2B software.
- Fin, the AI agent built inside what used to be Intercom, prices at $0.99 per resolution, one of the clearest examples of outcome-based pricing live today.
- Adobe’s own 2013 fiscal filings confirmed a real revenue decline during its shift to Creative Cloud; Autodesk saw regional revenue declines of up to 42 percent during its own 2016 transition.
- Agentic AI task costs have risen roughly thirtyfold since 2023 according to EY, which is the specific reason outcome-based pricing needs careful, account-level cost monitoring.
Häufig gestellte Fragen zum Vergleich von Lizenzmodellen
What’s the right approach when comparing license models?
None of the four models is universally best. Perpetual licensing suits customers who want long-term ownership and control. SaaS suits products that improve continuously and depend on an ongoing service relationship.
Usage-based pricing suits products where consumption maps cleanly to value. Outcome-based pricing suits products that can define and deliver a precise, defensible result. The right choice depends on the product and the customer, not on which model is currently fashionable.
Was ist der größte Fehler, den Unternehmen bei Lizenzmodellen machen?
Letting pricing happen deal by deal instead of designing a structure upfront. This shows up differently across software license models: arbitrary discounting under perpetual and SaaS, underpriced complexity under usage-based and outcome-based pricing.
The root cause across all of them is the same lack of structural discipline.
Lohnt sich die Umstellung von Kauflizenzen auf SaaS?
It can be, as Adobe’s long-term results show, but only if the company plans for the product re-engineering, the services and partner capacity, and the multi-year cash flow dip as connected problems. Companies that treat it purely as a billing change tend to underestimate all three.
Warum erfordert die ergebnisbasierte Preisgestaltung mehr Überwachung als andere Softwarelizenzmodelle?
Because the price charged per outcome is typically fixed while the cost of producing that outcome, especially with generative AI involved, can vary enormously between customers and even between transactions for the same customer. Without account-level cost monitoring, margin can erode quietly on the most complex accounts.
What is the connection between the value meter idea and outcome-based pricing?
A product value meter would give a vendor and a customer a shared, continuously measured view of the value a product delivers. That is exactly the evidence outcome-based pricing needs to be priced and defended honestly.
Without it, price rests on a one-time projection made during the sales cycle rather than on ongoing proof. Our article on the product value meter concept explores what building one would actually take.
Comparing License Models: Pick a Structure, Then Defend It
Nine articles is a lot of ground to cover, and the temptation is to walk away looking for the one right answer among perpetual, SaaS, usage-based, and outcome-based licensing. There is not one.
There is only the discipline of picking a model deliberately, building the metrics and the service structure it requires, and defending that structure even when a big renewal or a hard transition makes a one-off exception tempting.
The companies that get this right are rarely the ones with the cleverest pricing idea. They are the ones willing to do the less glamorous work: tracking maintenance ratios, funding customer success properly, monitoring cost-to-serve account by account, and saying no to the discount that would quietly undo all of it.
If you want help auditing which of these patterns are already showing up in your own pricing, or in your next transition between software license models, Kontakt aufnehmen or see how this fits into our services.