The two lines behave differently
| Recurring revenue | Services revenue | |
|---|---|---|
| Source | Active subscription contracts | Per-engagement statements of work |
| Continues without a new sale | Yes | No |
| Scales with | Installed base | New bookings and delivery headcount |
| Gross margin | High | Low, often near breakeven by design |
| Counts toward ARR | Yes | No |
Margin is the reason investors separate them
Software margin comes from selling the same product again with almost no incremental cost. Services margin is capped by the hours it takes to deliver, so it scales with headcount. A company reporting blended revenue growth without splitting the lines can be growing its consulting arm while its software base sits still.
The same logic applies internally. Blended revenue per rep tells you nothing when one rep sells $400,000 of subscription and another sells $200,000 of subscription plus $200,000 of custom work.
Services still earn their place
Implementation work that gets a customer live faster pays for itself through the renewal. The failure mode is custom development sold to close a deal, which produces a deployment that cannot be upgraded cleanly and a support load that persists for years.
Watch services attach alongside net revenue retention rather than as a revenue target on its own. Services sold as a revenue line drift toward whatever the customer will pay for. Services sold as an adoption mechanism stay pointed at time to value.
Keeping them separate in the forecast
Split the two at opportunity level in the CRM, with distinct products and distinct close date logic. Services revenue arrives after the project starts, which is weeks or months after the software contract is signed, so the two lines have different timing even when they come from the same deal.
Forecast services from the new bookings forecast multiplied by attach rate, not from the trailing services number. A soft new business quarter shows up in services one or two quarters later, and a team forecasting services on its own history will miss that turn entirely. Build both lines inside the same sales forecast so the dependency is visible, and check forecast accuracy separately for each, since the services line has a shorter and more mechanical relationship to what already closed.
Frequently Asked Questions
Should services revenue be included in ARR?
No. ARR counts revenue contracted to repeat. A one-time implementation fee repeats only when you sell another implementation, which makes it a function of new bookings rather than of the installed base.
What services gross margin is normal in B2B SaaS?
Services carry materially lower margin than software because the cost scales with delivery hours. Many companies run services close to breakeven on purpose and treat it as an adoption investment rather than a profit line.
Does a high services attach rate hurt the business?
It depends on what the services buy. Implementation work that shortens time to value earns its keep through retention. Custom development that creates a bespoke deployment nobody can upgrade creates a support burden that shows up years later.
How do you forecast services revenue?
Forecast it off the new bookings forecast and the attach rate, not off last quarter's services number. Services revenue lags software bookings by the gap between contract signature and project start, so a soft new business quarter shows up in services one or two quarters later.
Put these metrics to work
ORM builds custom revenue forecast models that turn concepts like services revenue vs recurring revenue into prescriptive action for your team.
Schedule a Demo