A marketing plan can be full of good ideas and still get rejected. What usually gets a budget approved is the model behind it, the logic that connects spend to a number a CFO can defend in a board meeting.
Most marketers pitch channels. Finance teams think in models. Learning to speak both languages is what separates a budget that gets signed off from one that gets sent back for “more justification.”
6 Models CFOs Actually Use
- Percentage of revenue – budget set as a fixed share of projected sales
- Zero-based budgeting (ZBB) – every dollar justified from scratch each cycle
- Objective-based budgeting – budget built bottom-up from specific goals
- ROI-based allocation – funds flow toward the channels with proven returns
- Competitive parity – spend benchmarked against rival share of voice
- Marketing mix modeling (MMM) – statistical modeling of aggregate spend against sales outcomes
Each model answers a different question a CFO asks. Knowing which question you are actually being asked determines which model to bring into the room.
Why the Model Matters More Than the Pitch Deck
CFOs approve numbers that reconcile with a company’s financial planning process. A marketer who shows up with a wishlist of channels and a gut-feel total gets pushback almost automatically. A marketer who shows up with the same request framed inside a model finance already trusts gets a much shorter conversation.
This is also why marketing budgets keep drifting. According to Gartner’s 2026 CMO Spend Survey, average marketing budgets have settled around 7.8% of company revenue, down roughly 18% from four years earlier. Budgets are not shrinking because marketing stopped working.
They are shrinking because fewer teams can prove which model justifies the next dollar.
1. Percentage of Revenue
This is the model most finance teams default to first. Marketing gets a fixed slice, often 5% to 15% of projected revenue, and that slice moves up or down as revenue forecasts change.
It works well for stable, predictable businesses. It reconciles cleanly with the annual budget and requires almost no defense beyond “this is what similar companies spend.”
The weakness shows up during a downturn. Revenue drops, the marketing budget drops with it, and the channels that could have pulled the company out of the slump lose funding at the worst possible moment. This model treats marketing as an expense that tracks revenue rather than a lever that drives it.
2. Zero-Based Budgeting
Zero-based budgeting throws out last year’s numbers entirely. Every line item, every retainer, every subscription has to earn its place again from a blank page.
It is heavy on process. Building a case for a $30,000 events line takes real time when nothing carries over automatically. But it is the fastest way to catch the tools nobody uses anymore, the agency retainer that stopped producing results, or the campaign that kept running out of habit rather than performance.
ZBB works best in cost-cutting cycles, after a leadership change, or when a company suspects legacy waste has crept into the marketing line. It is rarely used as a full annual method because of the time cost, but running it on one or two categories a year is a low-effort way to find the drains a normal audit misses.
3. Objective-Based Budgeting
Instead of starting with a total and dividing it, objective-based budgeting starts with a goal, say, 1,000 qualified leads, and works backward to price out exactly what it takes to hit it.
This is the model early-stage companies lean on because there is no revenue history to anchor a percentage against, and no attribution data mature enough for an ROI model. The budget request reads as: to generate this pipeline, we need this content volume, this ad spend, and this headcount.
The risk is inflated goals producing inflated budgets. Objective-based requests need discipline in the goal-setting stage, or they turn into a wishlist with a spreadsheet wrapped around it.
4. ROI-Based Allocation
Once a company has enough attribution history, usually twelve months or more, budget decisions shift toward channels with proven returns. Funds move from underperforming lines to whatever is generating the best cost per acquisition or return on ad spend.
This is the model most finance teams want to see, because it directly answers the only question that matters at approval time: what does the next dollar buy? A well-built data-driven digital marketing approach makes this model possible, since ROI-based budgeting is only as good as the attribution data feeding it.
The trap is over-optimizing for what is easy to measure. Channels like brand content or PR often get starved under a pure ROI model because their impact shows up months later and rarely gets clean, last-touch credit.
5. Competitive Parity
Some budgets get set by looking sideways instead of inward. Competitive parity benchmarks spend against what rivals are estimated to invest, usually to protect or grow share of voice in a crowded market.
This model is common for challenger brands entering a category dominated by a handful of larger players. It answers a different CFO question: are we visible enough to compete, regardless of our own historical spend.
Reliable competitor spend data is hard to get outside public filings, so this model often runs on estimates from ad intelligence tools. It is directional, not precise, and works best paired with one of the other five models rather than standing alone.
6. Marketing Mix Modeling
MMM is the oldest model on this list and, thanks to cookie deprecation and consent restrictions, the one having a comeback. Instead of tracking individual users, it looks at aggregate spend and sales data over two to three years and statistically estimates how much each channel contributed.
It captures offline and brand spend that click-based attribution simply cannot see. A well-known failure pattern illustrates why this matters: a company cuts a brand campaign because a last-click dashboard shows almost no direct conversions, funds a paid search channel that looks like the hero instead, and a quarter later pipeline drops because the demand feeding paid search quietly disappeared.
The channel was working. The measurement just could not see it.
MMM needs scale to work reliably, generally a mix diverse enough and a budget large enough to show real variation over time. Smaller companies usually get more value from ROI-based or objective-based models until they reach that scale.
Comparing the Six Models
| Model | Best for | Data required | Main weakness |
| Percentage of revenue | Stable, mature businesses | Revenue history | Cuts spend right when growth is needed most |
| Zero-based | Cost audits, turnarounds | Full cost itemization | Time-intensive to run annually |
| Objective-based | Early-stage, new launches | Clear goals and unit costs | Risk of inflated asks |
| ROI-based | Mature attribution setups | 12+ months channel data | Starves hard-to-measure channels |
| Competitive parity | Challenger brands | Competitor spend estimates | Data is directional, not exact |
| Marketing mix modeling | Large, diverse budgets | 2-3 years of weekly data | Needs scale and spend variation |
How to Choose the Right One
Most finance teams do not pick a single model and stick with it forever. A common pattern is objective-based budgeting during the early years, a shift toward percentage of revenue once the business stabilizes, and a move toward ROI-based or hybrid models once attribution data matures.
Larger organizations often blend two or three models, splitting the budget so proven channels run on ROI logic while newer bets run on objective-based logic.
If your team wants to get hands-on with building these models rather than just reading about them, working through online financial modeling courses is a practical way to learn the mechanics finance teams expect to see in a budget proposal.
FAQ
Conclusion
None of these six models is inherently better than the others. Each one answers a different question finance is asking, and the marketers who get budgets approved fastest are the ones who figure out which question is on the table before they build the pitch.
Learn the model your CFO already trusts, build the request inside it, and the conversation about approval gets a lot shorter.