Paid media
B2B paid media, judged on payback
DemandBox runs B2B paid media as a pipeline programme rather than a channel. We set the channel mix from the pipeline model, treat creative as the main targeting lever, test offers rather than audiences alone, and judge every line of spend on pipeline created and how quickly the acquisition cost comes back.
- Channels
- Search, paid social, Reddit, syndication where it earns it
- Primary lever
- Creative and offer, not audience settings
- Measured on
- Pipeline created and CAC payback
- Cadence
- Weekly review, monthly budget reset
What this covers
Most B2B paid accounts are optimised toward the cheapest conversion the platform can find, which is almost never the conversion that becomes revenue. The account looks healthy, cost per lead falls, and pipeline does not move. The fix is not better bidding. It is changing what the account is asked to produce and being willing to switch a channel off when it does not produce it.
The failure modes this exists to fix
- Budget set from last year's spend plus ten percent instead of from the pipeline required
- Capture and creation channels reported together, hiding which one is actually carrying the programme
- Offer left constant while only audience and bid get tested, when in B2B the offer decides response rate more
- Cost per lead reported as success while pipeline stays flat, because the account is optimising toward the wrong instruction
- No agreed point at which a channel gets turned off, so a losing channel survives on hope for a year
Who this is for
- B2B SaaS companies with a pipeline model, or willing to build one first
- Teams who want capture and creation reported separately
- Companies willing to have a spend line cut when the payback does not clear
Who this is not for
- Anyone judging the account on cost per lead alone
- Anyone who wants a fixed monthly lead number regardless of quality
- Companies with no offer worth converting on yet
Verdict: If the offer is not worth converting on, that gets fixed before any spend increase, because more traffic to a weak offer is a bigger loss, not a bigger result.
How the engagement runs
The paid media loop
Pipeline model
Sets the budget and the target
Offer and creative
Built and tested before scale
Channel split
Capture vs creation, reported separately
Weekly review
What moved, what gets cut
Monthly reset
Budget reallocated against payback
- 1
Days 1 to 30: model, audit and rebuild the account structure
Work backwards from the pipeline target through win rate and deal size to the pipeline paid has to create. Audit the existing account against that number, separate capture spend from creation spend, and rebuild the structure so both can be reported and judged independently.
- 2
Days 31 to 60: test the offer and scale the creative line
Run at least two offers against each other rather than testing audiences alone. Stand up a creative production and testing calendar, since creative now does more of the targeting work than the platform's audience settings.
- 3
Days 61 to 90: cut, reallocate and report against payback
Apply the pre-agreed cut points to anything not clearing them. Reallocate the freed budget to what is working. Report pipeline created and CAC payback against the original model, not against last month's platform metrics.
What you get
Included in a paid media engagement
- ✓Channel and budget allocation derived from the pipeline model
- ✓Account build or rebuild, with capture and creation reported separately
- ✓Offer testing, not just audience and bid testing
- ✓Creative concepting, production and a structured testing calendar
- ✓Landing page and conversion path work on the destination
- ✓Weekly reporting on pipeline created and payback, with explicit cut decisions
- ✓A monthly budget reset meeting tied to the pipeline model
How it is measured
CAC payback period
CAC payback (months) = customer acquisition cost / (monthly revenue per customer x gross margin)
- customer acquisition cost: fully loaded spend to acquire one customer, including media and production
- monthly revenue per customer: average recurring revenue per customer per month
- gross margin: expressed as a decimal
Example: Healthy 2026 benchmarks put payback under 18 months and pipeline coverage at 3x to 4x target, which is the bar spend gets measured against here.
| Metric | What it means here | Source |
|---|---|---|
| 19% average win rate, down from 29% | Fewer opportunities are closing, which raises the bar on pipeline quality paid has to create | ebstaPavilion |
| 106 day median sales cycle, ~8% longer YoY | Payback calculations need to assume a longer close, not last year's cycle time | salesBenchmarks |
| 3x to 4x pipeline coverage, CAC payback under 18 months, magic number above 0.75 | The reference bar spend and pipeline are measured against | causoGtm |
| 31% of reps at full quota | Context for why pipeline volume alone does not fix a close rate problem | salesBenchmarks |
What we will not do
- Report cost per lead as the headline metric
- Keep a losing channel running past its agreed cut point to protect a relationship
- Scale spend on an offer that has not been tested against an alternative
- Recommend spend increases without a pipeline model behind the number
- Blend capture and creation spend into one number that hides which is working
Objections, answered
The strongest case against this
Our cost per lead is good, so paid is working.
Cost per lead is the metric most easily improved by attracting people who will never buy. It falls when you widen the top of the funnel and it falls when you gate something popular. Neither necessarily produces pipeline. If cost per lead is down and pipeline is flat, the account is working correctly toward the wrong instruction.
The strongest case against this
Win rates are falling everywhere, so weaker paid pipeline is not our problem to fix.
The average win rate across 655,000 tracked opportunities fell from 29% to 19%, so the market is harder for everyone. That is an argument for tighter offer and audience discipline in paid, not a reason to accept it. A programme that keeps sending the same volume at a lower win rate is spending more to produce less.
The strongest case against this
We should just increase budget to hit the number faster.
Sales cycles are running about 8% longer year on year at a 106 day median, so more spend now shows up in the pipeline model later than it used to. Increasing budget without first fixing the offer or the conversion path mainly buys a bigger number of the same weak pipeline, sooner.
See the banners, UGC and video we have produced for B2B accounts in the DemandBox creative studio.
Common questions
- Which paid channels work for B2B SaaS?
- Search captures the demand that already exists and is usually the first thing to get right. Paid social and community channels create demand and carry the growth once search is saturated. Which specific platforms earn budget depends on where your buyers actually are, which the first quarter of testing establishes.
- How much should we spend?
- Work backwards from the pipeline target through win rate and deal size to the pipeline paid has to create, then to the spend that produces it at an acceptable payback. A budget set any other way is a number someone liked.
- Do you manage the creative too?
- Yes. Creative is the main targeting lever in B2B paid now, so separating media buying from creative production tends to cap the programme. Concepting, production and the testing structure are part of the engagement.
- What metrics do you report?
- Pipeline created, cost per pipeline, CAC payback, and the split between capture and creation. Clicks, impressions and cost per lead appear as diagnostics, not as the score.
- How quickly can paid produce pipeline?
- Capture channels can produce qualified conversations within weeks, because the demand is already there. Creation channels take longer and should be judged over a quarter rather than a fortnight, especially with median sales cycles now around 106 days.
- Will you tell us to spend less?
- When the numbers say so, yes. An agency paid on a percentage of spend that never recommends cutting spend is telling you something about its incentives rather than your account.
- How do you decide when to cut a channel?
- The cut point is agreed before spend starts, tied to pipeline created and payback rather than a gut call made mid quarter. Agreeing it in advance is what stops a bad channel surviving on hope.
- Do you test offers or just ad creative?
- Both, but offer testing comes first. In B2B the offer, a demo request against a diagnostic against gated research, decides response quality more than the creative wrapped around it does.
- What is a healthy CAC payback period?
- Current 2026 benchmarks put a healthy payback under 18 months, alongside pipeline coverage of 3x to 4x target and a magic number above 0.75. Those are the reference points spend gets measured against, not a promise for any specific account.
Read the thinking behind it
Want this run for you?
Tell us what you are spending and where the pipeline stalls. If this is not the right first step for you, we will say so.