Where should the next $50,000 go? A marginal-dollar test for B2B paid media
The average cost per opportunity of a campaign says little about what the next $50,000 will buy, because returns fall as you spend more. Test the margin instead: add a defined amount to one proven line for a fixed window, hold everything else steady, record the cost of the extra opportunities, and compare it with the same increment placed elsewhere. Decide the rule before you start: if the marginal cost per opportunity stays under your limit, scale; if not, move the money.
Key takeaways
- Average cost hides diminishing returns. A campaign averaging $5,000 per opportunity can cost $10,000 for each additional one.
- Google reports when a campaign loses impressions to budget and when it loses them to rank. Only the first is a case for more money.
- Test with a fixed increment, a fixed window and a written rule, and change one thing at a time.
- With low B2B volumes, treat a test as directional evidence combined with sales feedback, not as proof.
Average cost is not marginal cost
Suppose a Google campaign spends $100,000 in a quarter and creates 20 qualified opportunities, an average of $5,000 each. Add $50,000 and it creates 5 more. The extra opportunities cost $10,000 each, double the average, because the easy searches were already bought. The average looked healthy and the margin was not.
This illustration uses assumed numbers to show the logic. Your own marginal cost has to be measured, and it differs by campaign and by quarter.
| Spend | Opportunities | Average cost per opportunity | Marginal cost of the last step |
|---|---|---|---|
| $100,000 | 20 | $5,000 | |
| $150,000 | 25 | $6,000 | $10,000 |
Is the line actually budget-limited?
Before testing more money, check whether money is what limits the line. Google defines search lost impression share (budget) as the share of time your ads did not show because of insufficient budget, reported at campaign level, and search lost impression share (rank) as the share lost to poor ad rank. A campaign that loses a lot to budget while making opportunities inside your limit has room to absorb more. One that loses to rank needs better relevance, bids, ads or landing pages first.
| What you see | Likely reason | First move |
|---|---|---|
| High loss to budget, cost per opportunity within limit | Proven demand you are not reaching | Add budget in steps and run the marginal test |
| High loss to rank | Ads or pages are not competitive for these terms | Improve quality and bids before adding budget |
| Low loss on both, spend flat | You already reach most of the demand | Move money to priming or to another capture line |
| Cost per opportunity above limit | Targeting or qualification problem | Fix before any increase |
Bring your funnel numbers to a 30-minute pipeline plan session. We show where spend is leaking against your CRM data and what to fix first.
The marginal-dollar test in six steps
- Pick the line: a campaign or campaign group with at least a full quarter of qualified-opportunity data and cost per opportunity inside your limit.
- Set the increment: a fixed amount, for example 25 to 50 percent of its current spend, large enough to produce at least several opportunities in the window.
- Fix the window: eight to twelve weeks, so most leads reach the opportunity stage you are measuring.
- Write the rule before you start: scale if marginal cost per opportunity is under your limit, hold if it is between the limit and 1.25 times the limit, move the money if it is above.
- Hold everything else steady: no new creative, bidding mode or landing page changes in the test lines during the window.
- Read it by marginal cost: extra spend divided by extra qualified opportunities against the previous period, then check sales feedback on quality.
The 1.25 times threshold in step four is our rule of thumb. Set your own with finance.
Test the same increment somewhere else
A marginal result is only useful against an alternative. Run the same increment through a second line, for example priming spend on LinkedIn for your target accounts, or a new capture segment such as competitor and comparison terms. Compare the two on the metric that fits each job. Capture lines are judged on marginal cost per opportunity. Priming lines are judged on leading indicators first, as described in the long sales cycle article.
Where volume allows, use a holdout: keep a comparable region or account group out of the extra spend and compare. Where it does not, which is common in enterprise B2B, use before and after reads with the same measures and sales validation. Be honest in the report that this is directional. The jobs referred to here are the ones in our four-job 2027 budget model.
What to report to the CFO
| Item | Example format |
|---|---|
| Test question | Does adding $50,000 to non-brand intent search create opportunities within our cost limit? |
| Rule agreed in advance | Scale if marginal cost per opportunity is under the limit |
| Result | Extra spend, extra qualified opportunities, marginal cost per opportunity |
| Quality check | Sales view of the extra opportunities |
| Decision | Scale, hold or move, with the date of the next review |
Pair this with the CFO-ready budget case from our article on how to defend your marketing budget.
If you need to fund the test by stopping something else, run the spend audit first.
How we help
We design these tests inside client accounts, set the rules with the CMO and finance, and report marginal cost per qualified opportunity from the CRM. It is the work behind the pacing and reporting we run for B2B accounts, and our paid media service is the starting point if you want a test designed for your next quarter.
Frequently asked questions
How do we decide where to put additional marketing budget?
Test the margin. Add a fixed increment to a proven line for eight to twelve weeks, measure the cost of the extra qualified opportunities and compare it with the same increment placed elsewhere, using a rule written in advance.
What is marginal cost per opportunity?
The extra spend divided by the extra qualified opportunities it created. It is usually higher than the average because the easiest demand is bought first.
How do we know if a Google campaign needs more budget?
Check search lost impression share (budget) and (rank). A high loss to budget on a campaign that makes opportunities inside your cost limit suggests it can absorb more. A high loss to rank means improving relevance and ad rank first.
Can small B2B accounts run a statistically valid test?
Often not. With low opportunity counts treat the result as directional, combine it with sales feedback, and repeat it before making a large change.
Sources
Related service: Paid media. See also pricing and client results.
We run enterprise B2B demand generation for teams in Montreal, Toronto, Ottawa and the United States.
