What belongs in the numerator
Pick one definition of spend and hold it. The common choices are media cost alone, media plus content and event production, or a fully loaded figure that adds demand generation salaries and martech licenses. Each produces a different number from the same funnel, and the fully loaded version runs materially higher than the media-only version.
The definition matters less than the consistency. A cost per MQL that quietly gains a headcount line in Q3 will show a spike that no one can explain, and the team will spend the review debating the metric instead of the channels.
The scoring threshold controls the trend
Cost per MQL is the most easily manipulated metric in demand generation reporting, and usually without intent. Lower the score threshold and the MQL count rises while spend stays flat, which makes the number improve while nothing about the business changed. Raise it and the number worsens for the same reason.
Record the scoring model version and its effective date next to the metric. Any period-over-period comparison that crosses a threshold change should be labeled as such, because the two periods are measuring different populations.
Price the acceptance rate into it
An MQL that sales rejects cost exactly as much as one sales accepts. That makes cost per accepted MQL the more useful figure. Divide cost per MQL by the acceptance rate and you get the real price of a lead the sales team will work.
Carry it one stage further and the metric becomes a planning input. Cost per opportunity equals cost per MQL divided by the product of the acceptance rate and the acceptance-to-opportunity rate. That figure can be tested against allowable acquisition cost and used to back into the spend a pipeline coverage target actually requires.
Where the metric earns its place
Use cost per MQL to compare channels against each other inside a period, which is the comparison it handles well. Do not use it to defend a total budget or to grade the marketing team year over year, because too much of its movement comes from definitional drift rather than performance.
The channel view is where it pays off. A source with a low cost per MQL and a poor win rate is producing volume the sales team pays for later, and no amount of top-of-funnel efficiency makes that trade worthwhile.
Frequently Asked Questions
How do you calculate cost per MQL?
Divide total demand generation spend for the period by the number of marketing qualified leads created in that period. The decision that matters is what goes into the numerator. A media-only figure and a fully loaded figure that includes salaries and tools describe the same funnel at very different prices.
What is the difference between cost per lead and cost per MQL?
Cost per lead uses every record captured. Cost per MQL uses only the records that cleared a fit and intent threshold. The gap between the two prices your qualification rate, so a widening gap means you are paying for a growing share of leads that never qualify.
Why does cost per MQL jump without any change in spend?
Almost always because the scoring model changed. Raising the threshold cuts MQL count and inflates cost per MQL overnight. Any comparison that spans a scoring change compares two different metrics, which is why the threshold and its effective date belong on the report.
What is a good cost per MQL?
The one that survives the funnel below it. Divide cost per MQL by the product of your acceptance rate and your acceptance-to-opportunity rate to get cost per opportunity, then test that figure against allowable acquisition cost. An MQL price is only defensible once it has been priced downstream.
Put these metrics to work
ORM builds custom revenue forecast models that turn concepts like cost per mql into prescriptive action for your team.
Schedule a Demo