Start from the number you owe, not last year plus 10%
Most B2B paid media budgets are built the lazy way: take last year's spend, add a percentage, and defend it in a slide. The problem is that it anchors next year's plan to a number that was never tied to a business outcome in the first place. You end up optimizing a budget instead of a result.
A budget that scales starts from the obligation, not the history. Your company has a revenue or pipeline target. Some share of that is expected to come from paid media. That share, run backward through your funnel, is the only honest way to size spend.
The rule: every dollar in your paid media budget should trace to a pipeline or revenue commitment. If it cannot, it is discretionary, and discretionary spend is the first thing cut when the quarter gets tight.
The funnel math: working back from pipeline to spend
The core move is to invert your funnel. You know roughly what fraction of pipeline closes, what fraction of opportunities come from qualified leads, and what it costs to generate those leads. Chain those together and spend falls out of the math.
Here is a worked example. Say paid media is on the hook for $4M in new closed-won revenue this year.
- Win rate is 25%, so you need $16M in qualified pipeline to close $4M.
- Average deal size is $40K, so $16M of pipeline is 400 opportunities.
- MQL-to-opportunity conversion is 20%, so 400 opps require 2,000 MQLs.
- Blended cost per MQL from paid is $250, so 2,000 MQLs cost $500K in media.
- That implies a CAC near $5,000 per deal ($500K across 100 wins) against a $40K deal, a ratio you can grow into.
Now you have a defensible top-line number and, more importantly, a model. Change any input, your win rate improves, deal size grows, cost per MQL creeps up, and the required spend moves with it. That is the difference between a budget and a spreadsheet of wishes.
Build this model with conversion rates you can actually evidence, not the ones you hope for. If your measurement is shaky, fix that first; a model built on guessed rates is just a prettier version of last year plus 10%. Talk to us before you lock the number.
Split the budget across demand capture and demand creation
Once you have a top-line number, the first split is not by channel. It is by job. Paid media does two fundamentally different things, and conflating them is how budgets quietly underperform.
Demand capture
Capture is harvesting intent that already exists: branded and high-intent search, retargeting, review-site and comparison placements. It converts efficiently because the buyer is already looking. The catch is that it is capped by how much demand exists; you cannot capture more than the market is generating.
Demand creation
Creation is building intent that is not there yet: paid social, programmatic, content distribution, and sponsorships aimed at buyers who are not in-market today. It looks worse on last-click metrics and better on pipeline over a two-to-four-quarter horizon.
A common starting point is to weight capture more heavily early, then shift toward creation as capture saturates. Many B2B programs land somewhere around 60 to 70 percent capture and 30 to 40 percent creation, but treat that as a hypothesis to test, not a law. If your branded search is already maxed, more creation is the only way to grow.
Allocate across channels with logic, not habit
Channel allocation should follow the demand split, not the other way around. Decide how much goes to capture and creation, then choose the channels that do each job best for your buyer and your deal economics.
The questions that should drive the split:
- Where does your buyer actually research? A six-figure platform deal and a self-serve tool do not live in the same channels.
- What does each channel cost to produce pipeline, not clicks? A cheap CPL with terrible lead quality is expensive where it counts.
- How much headroom does the channel have? A channel you can double without the cost per result spiking is worth more than one at its ceiling.
- What is the sales-cycle fit? Long cycles reward channels that compound over quarters; short cycles reward fast-converting capture.
Resist spreading budget evenly to feel diversified. Concentration in channels that work beats a thin presence everywhere. A landing page and conversion layer that turns the same traffic into more opportunities effectively lowers the cost of every channel feeding it, which can change which channels are worth funding at all.
Reserve a holdback for testing
A budget with no slack cannot learn. If every dollar is committed to proven channels, you will ride them until they decay and have nothing warmed up to replace them. Scaling programs always carry a testing reserve.
A practical range is 10 to 20 percent of the budget held back for new channels, audiences, offers, and creative. Early in a program, or when proven channels are saturating, push toward the top of that range. When you have a reliable engine and a strong quarter to hit, you can run leaner.
Treat the holdback like a portfolio of small bets with explicit exit criteria. Each test gets a hypothesis, a budget cap, a timeline, and a success threshold tied to a leading indicator. Tests that clear the bar graduate into the core budget; tests that do not are killed on schedule, not kept alive out of sunk-cost loyalty.
The holdback is not spare money. It is how you find the channel that funds next year's growth before this year's channels run out of room.
Watch leading indicators and reallocate on them
Pipeline and closed revenue are lagging indicators. By the time they move, the decision that caused it is a quarter old. To manage a budget that scales, you reallocate on leading indicators that predict pipeline weeks ahead.
The signals worth watching weekly:
- Cost per qualified lead by channel, not cost per click or raw lead.
- Lead-to-opportunity conversion by source, which exposes quality problems a cheap CPL hides.
- Pipeline velocity, how fast sourced leads become opportunities.
- Marginal cost of the last dollar in each channel, which tells you whether the next dollar is still efficient.
Set thresholds in advance. When a channel's cost per qualified lead drifts above your model's assumption for two or three weeks, shift spend toward channels still inside their range. When a channel beats its target with headroom, feed it. Reallocation should be a standing weekly habit, not a quarterly fire drill, and it only works if your attribution is good enough to trust source-level numbers.
Know when to add budget and when you are hitting the ceiling
Scaling is not spending more. It is spending more where the next dollar still earns its keep. The discipline is telling the two situations apart.
Signals that a channel is ready for more budget:
- Cost per qualified lead holds steady as you raise spend week over week.
- Lead quality and lead-to-opportunity conversion stay flat or improve as volume grows.
- The channel is capacity-constrained, not demand-constrained, you are leaving impressions or auctions on the table.
Signals you are hitting diminishing returns:
- Cost per qualified lead climbs as you add spend, the classic saturation curve.
- Lead quality drops as you widen targeting to find more volume.
- Frequency rises while incremental conversions fall, you are paying to reach the same people again.
When a proven channel saturates, the answer is rarely to force more money through it. Pull from your tested holdback, promote the bets that cleared their thresholds, and open the next channel while the current one is still healthy. That is how a budget compounds instead of plateauing, and it is the whole point of building the plan from pipeline math in the first place.