Cost per opportunity is a campaign's spend divided by the opportunities it created, and cost per SQL is spend divided by its sales qualified leads, for the same campaigns and the same period. They are the earliest reliable signs of whether paid media is producing pipeline, long before deals close.
Why these two numbers
Cost per lead is quick but says little about quality. Revenue is the final answer but arrives months later. Cost per SQL and cost per opportunity sit in between: sales has looked at the lead and agreed it is worth pursuing, and the number is available within weeks of the spend.
The formulas
| Metric | Formula | Counts |
|---|---|---|
| Cost per MQL | Spend / MQLs | Contacts that reached MQL or a later stage |
| Cost per SQL | Spend / SQLs | Contacts that reached SQL or a later stage |
| Cost per opportunity | Spend / opportunities | Deals or opportunities created from the campaign |
| MQL to SQL rate | SQLs / MQLs | How well the campaign's leads pass sales review |
Count a record at the furthest stage it reached, and include later stages. Otherwise a lead that moved straight from MQL to opportunity disappears from the SQL count and the campaign looks worse than it is.
A worked example
| Campaign | Spend | MQLs | SQLs | Opportunities | Cost per MQL | Cost per SQL | Cost per opportunity |
|---|---|---|---|---|---|---|---|
| Search: category terms | $18,000 | 90 | 24 | 9 | $200 | $750 | $2,000 |
| LinkedIn: job title targeting | $22,000 | 110 | 18 | 5 | $200 | $1,222 | $4,400 |
| Search: competitor terms | $8,000 | 20 | 10 | 5 | $400 | $800 | $1,600 |
All dollar figures in the examples are illustrative. They are not results from any account.
The two largest campaigns cost the same per MQL, yet the LinkedIn campaign costs more than twice as much per opportunity. The competitor campaign looks expensive per MQL and is the cheapest way to an opportunity.
How to calculate them step by step
- Pick the periodUse the month for a quick read and the last 3 to 6 months for decisions. A single month with two opportunities is noise.
- Match spend and records to the same campaignGroup CRM records by utm_campaign and spend by the ad campaign name or ID, so both sides use the same key.
- Count by stage reachedCount each record at its furthest stage, including later ones, for the same period as the spend.
- Divide and rankDivide spend by each count, then rank campaigns by cost per opportunity. Flag any with fewer than three opportunities as not yet comparable.
- Read alongside volumeA low cost per opportunity on a tiny campaign may not hold at a larger budget. Test with a step, not a jump.
Traps that distort the numbers
- Lag. Leads created late in the month reach SQL next month, so the current month looks expensive. Compare complete periods.
- Unmatched records. Contacts with no utm_campaign fall out of every campaign. Track how many paid contacts have no campaign.
- Renamed campaigns. Renaming a campaign mid-flight splits its records across two names.
- Different definitions. If sales changes what counts as an SQL, the trend breaks. Note the date of any change.
What Pacing Console shows
On Pro, connect HubSpot, Salesforce or Pipedrive (or send any CRM by CSV or API) and Pacing Console shows contacts, MQLs, SQLs and opportunities per campaign in the pacing table, cost per MQL and cost per SQL in each campaign's window, and CRM stages by campaign with cost per stage on the Performance page. On Enterprise, the Pipeline page adds cost per opportunity, pipeline value and pipeline per $1 for this month or the last 3, 6 or 12 months. Open the live demo to see cost per stage on sample data.
Plans: CRM lifecycle stages (contacts, MQLs, SQLs and opportunities by campaign, with cost per stage) are on Pro, USD 49 per workspace per month. The Pipeline and Revenue pages, revenue forecasts, what-if scenarios and budget targets need Enterprise. See Plans, pricing and billing.
Questions people ask
What is a good cost per opportunity in B2B?
It depends on deal size and win rate. A useful check is cost per opportunity divided by win rate, which gives an approximate cost per customer to compare with customer value.
Should I use cost per SQL or cost per opportunity?
Use both. Cost per SQL is available sooner. Cost per opportunity is closer to revenue and less affected by how leniently leads are qualified.
Why does this month's cost per opportunity look high?
Leads take time to become opportunities, so the current month is always incomplete. Compare complete months, or the last three months together.
How do I count a lead that skipped the SQL stage?
Count records that reached a stage or any later stage. A lead that went straight to opportunity still counts as an SQL.
Can I see this below campaign level?
With HubSpot, Pacing Console can also keep the ad group and keyword when your landing URLs carry them, so you can see which parts of a campaign create the leads.
Related articles
Last updated 2026-11-22. Platform and CRM rules change, so check their own help for the current details.